1

George Kopasakis

NASA Glenn Research Center

Cleveland, Ohio

AeroServoElasticity - AeroPropulsoServoElasticity Fundamental Aeronautics – High Speed Project

Propulsion Control and Diagnostics (PCD) Workshop

Cleveland OH, Dec. 11 – 12, 2013

Dynamic Modeling of Supersonic Propulsion Systems for AeroPropulsoServoElasticity Analysis

Integrated APSE Model

APSE Task Objective

Develop appropriate vehicle models and controls to study the dynamic

performance of supersonic vehicles

Team:

NASA Langley – Develop the structural vehicle models

NASA Glenn – Develop the propulsion system models

Controls and Dynamics Branch (RHC): George Kopasakis, Joseph Connolly

Inlet and Nozzle Branch (RTE): David Friedlander

Multidisciplinary Design Analysis Optimization Branch (RTM): Jonathan Seidel, Jeff Chin

Antenna and Optical Systems Branch (RHA): Noulie Theofylaktos

Thermal Systems Branch (DET): Xiao-Yen Wang

University of Colorado: PhD Research Kyle Woolwine

Ohio State University: PhD Research Joseph Connolly

Both – Develop integrated model and perform

APSE studies

2

Fre

eSt

rea

m

Inle

t

Du

ct1

Fan

Spli

tte

r

Du

ctC

ore

HP

C

Du

ct3

Co

mb

ust

or

Du

ct4

0

HP

T

LPT

Du

ct7

Mix

er

DuctBypass

No

zzle

Fre

eE

nd

ShaftHP

ShaftLP

0 1 200 21 22

241 242

221 31 32 4 40 5

shaft_HPC

shaft_Fan shaft_LPT

shaft_HPT

6 7 8 9

Du

ct2

12

Fuel Fuel1

Ble

ed

s3

Bleed_HPC2

Bleed_HPC1 Bleed_LPT2

Bleed_LPT1

4142

45

49

CustBleed

Bleed_Srce2

Bleed_Srce1

Bleed_HPT2

Bleed_HPT1

Bleed_Cust

Overboard

VC

ESp

litt

er

Du

ct4

2

Duct1a200a

VCEFan VCEBypass VCENozzle VCEFreeEnd

222 42a

300 301 302 303

sha

ft_

VC

EFa

n

Variable Cycle Engine 1D Component Dynamic Model & NPSS Cycle Model

Feedback Controls GUI tool & Control Schedules

Parallel Flow Path modeling – 2D Dynamics

10-2

10-1

100

101

102

-10

0

10

20

30

40

frequency, Hz

acou

stic

wav

e ve

loci

ty s

pect

ral,

dB((

m/s

ec/H

z)

von Karman, eddy dissap.=5e-4

TF fit, eddy dissap.=5e-4

von Karman, eddy dissap.=8.5e-5

TF fit, eddy dissap.=8.5e-5

Atmospheric Turbulence Modeling

Shaker

Engines

Wing

APSE System Interfaces – analytical & by testing

Vehicle Structure – CFL3D, FUN3D & State Space

Integrated APSE

APSE

pre

ssu

re [

pa]

Nozzle Modeling – 1D & 2D for Dynamics

Note: Pictures manipulated to remove controlled information

External Compression Inlet CFD Modeling for Dynamics

0 0.5 1 1.50

0.2

0.4

0.6

0.8

1

nozzle length (m)

norm

alized a

rea &

density

nozzle area

final density

initial density

0 0.5 1 1.50

0.2

0.4

0.6

0.8

1

nozzle length (m)

norm

alized a

rea &

density

nozzle area

final density

initial density

0 0.5 1 1.50

0.2

0.4

0.6

0.8

1

nozzle length (m)

norm

alized a

rea &

density

nozzle area

final density

initial density

3

Compressor/Fan Test-Bed SIM for Derivation of Control Schedules

Variable Area Dynamic Compressor

SIM

Compressor Plug-Nozzle

SIM

Inlet Conditions As f(N)

P, T, M

𝑊

Compressor Inlet

Conditions

Vs. N

(for Plug Design)

Isolated Compressor Rig SIM - Variable Guide Vane angle as a function of rotor speed

Dynamic Engine SIM

4 Note: Pictures manipulated to remove controlled information

-- Purpose of new control design GUI is to maximize control system performance (stability vs. disturbance rejection & response time).

-- Only information needed is plant transfer function and how fast the plant is driven (plant actuation system response)

-- Robust feedback control designs obtained with couple clicks (normally in the first try) – Robustness also due to auto GUI design of 2 dominant natural controller frequencies. --- Low natural frequency for stability/robustness & for high disturbance rejection in mid frequency range

--- High natural frequency for fast response & for high disturbance rejection

-- GUI control design more tolerable to unmodeled dynamics and tighter stability margins

Feedback Control Design GUI

Step response of CMAPPS fan control system

designed using the GUI

Low natural freq. of control system High natural freq.

5

Nozzle Dynamic Characterization

Objective: Develop a dynamically accurate nozzle model to integrate with the rest of the propulsion system

Approach:

-- Develop CFD model to capture shock dynamics

-- Determine the minimum grid resolution necessary for dynamic

accuracy based on developed dynamic accuracy criteria

Nozzle CFD flow field

Nozzle frequency domain response characterization

Note: Pictures manipulated to remove controlled information

Flow

6

Integrated AeroPropulsoServoElastic (APSE) Model

Early assessments indicate that closed loop coupling of propulsion and vehicle AeroServoElastic dynamics could be significant enough, that it may not be possible to ignore.

7

Assessments of Compressor Stability Margin – realistic dynamics of stall

Test-bed to validate engine Cycle Deck designs, geometry sensitivity due to flow impedances and compressor/turbine dynamic feedback loop, and choked dynamics.

Test-bed to derive exit nozzle area control schedule and design controls to stabilize exit nozzle feedback loop

Test-bed for isolated compressor/plug system to derive compressor IGV control schedules

Perform assessments of the effect of flow distortion on dynamics

Using dynamic engine models as test-beds can substantially reduce costs by performing early design validation studies and by reducing test time of real engine hardware

Dynamic Propulsion System Modeling and Other Potential Applications

8

Future Work

Increasingly develop more detailed APSE simulations towards developing in the near future dynamic test-beds in simulation of component systems and a complete integrated vehicle model to conduct APSE performance studies such as ride quality, flight dynamics and vehicle stability, and flight efficiency.

9