Determining Aircraft Moments of Inertia from Flight Test Data

States.

United the in E1 protection Center copyright to Forum subject Morelli 20, Research not is SciTech and Gene January Langley AIAA Dr.

Government NASA U.S.

‐ the F of work a is

Determining Aircraft Moments

of Inertia from Flight Test Data

material This

Results

Simulation

Results

Maneuver

Approach

Test

Outline

Test

Nonlinear

Flight

‐

Introduction Modeling Flight F E1 Summary

• • • • • •

LaRC NASA Morelli E1 ‐ F modeling determined data?

be analysis (CAD) test Inertia: of

Introduction

inertia flight Used: Design of aerodynamic consuming ‐ dynamic from and time Moments Aided ‐ testing Methods moments and Aircraft accurately Nondimensional Simulation Ground Computer Costly for aircraft      LaRC Can Uses Conventional NASA Morelli power power idle idle for for 0 0 ≈ ≈ r q p p   y p p z I

 I

I I  pq 





 r r  p q



  x xz p  I I r





 z xz y I xz I I

Approach qr I



  z I pq x pr  I





y y I x I

 I

z Motion y  I   I x of z L I I







Modeling

  p  N M  Equations  r  q  Rotational LaRC Roll: Yaw: NASA Pitch: Nonlinear Morelli



r q  p 





c  p  pq r c



 c r   r q r y z p  I I c N xz xz   I I e a     q 4 q c e c a

Modeling

  N M   z y I I r q V V



Error



q q y

‐

x r I q I N   M x z  I I 





p V    q q c p c  N M

Equation

   q q o  M N   q  q o LaRC N NASA  r  Morelli variables significant axis low ‐ surfaces and negligible terms multi explanatory control terms inertia power the aerodynamics idle on automated at linear nonlinear gyroscopic

Maneuver decorrelated

with for inputs the turns for

Test

attack attack make propulsion of of to excitations descending make

Flight

angle rates perturbation angle to into surface angular nominal power Rolls nominal orthogonal High Idle Low Control LaRC     NASA Morelli Response Data Aircraft Modeling Control Surface Deflections

Maneuver

Limiters, Actuators Command Ground joystick the

Test

s +

…

from time, Desktop + :

Flight

Control Pilot Simulation Radio deg deg : , ,   Test Nonlinear LaRC Flight ‐ Inputs F E1 NASA (PTI) Test Programmed Morelli ‐ active F PTI with

Results

turns descending into

Simulation

rolls s s time, time, alternating

Nonlinear

‐

maneuver,

F

Single LaRC NASA 2 2 q r   deg / s deg / s Morelli ‐ F

Results

0.40 Error –5.68 –2.41 –5.25 Percent

Simulation

Error ± 0.0234 ± 0.0019 ± 0.0189 ± 0.0011 ± Standard 0.9643 0.0166 0.0147 Estimation –0.7517

Nonlinear

Estimate Linear

‐

F

True 0.9604 0.0176 0.0156 –0.7340 LaRC 4 5 6 c c c c NASA Parameter Morelli deg longitude, E1 deg ≤  deg ft latitude, active, altitude, PTI

Results

with turns

Test

s s

Flight

descending time, time,

E1

maneuvers, Five LaRC NASA 2 2 q r   deg / s deg / s Morelli E1

Results

3.01 3.89 Error –0.74 –5.80 Percent

Test

Error 1.002 0.732 0.495 0.107 Estimate ± ± ± ± %

Flight

Test ≤ Estimation 3.053 8.711 0.779

E1

11.036 ± Standard Flight Test uncertainty 2.964 8.776 0.750 Nonlinear 11.716 Estimate* Ground test LaRC y x z xz I I I I Ground NASA Parameter * Morelli results axis ‐ determining avoided multi be can with accurately determining for produced accurate demonstrated power data modeling and idle test at CAD flight accurately formulation or flight E1 for model developed testing and maneuvers informative test error was ‐ method ground flight simulation test produced inertia equation inertia of of piloted consuming flight ‐ nonlinear excitations versatile time ‐ F novel moments and and amplitude A moments ‐ both Findings: LaRC Large orthogonal aircraft Simple for Costly NASA   

Summary: Key

Morelli

Questions?

LaRC NASA Morelli

Slides

Backup

LaRC NASA Morelli time in xz z , I , I y contiguous , I x be I not accuracy need elements and

Solution

tensor modeling c , c , c , c for inertia maneuvers

Error

‐

for SIDPAC constants smoothed multiple using are estimation inertia from data for

Equation

solutions selected parameter function be estimation can LaRC Data Linear Nonlinear Modeling Straightforward NASA      Morelli sum the of maneuver length, s

Inputs

 Hz T    amplitude frequency, k   peak ‐ k t T  to

Multisine

2 ‐    peak k A sin  k P ,M 

, , 

1 2  minimum  k  for

Optimized

u tests input ‐ excursions dynamics flight orthogonal optimized k multi  past Input in minimal various Test design to Phases Design

Orthogonal

to LaRC (PTI) Time/Frequency Simultaneous Robust Balanced, Easy Successful NASA • • • • • • Programmed Morelli time, s zero phases optimized phases

Optimization

Phase

+ + +

Multisine

time, s LaRC 0 1 0 1 0 1 0 1 -1 -1 -1 -1 NASA Morelli are inputs conditions the

Inputs

domains trajectory and coverage flight good varied nominal frequency tests and off amplitudes and domain

Multisine

provides flight efficiency time aircraft and high other the robustness for frequency in perturbation move for or many variables not on

Optimized time

excitation content does in minimum orthogonal for input explanatory successful model to frequency mutually – maneuver design, optimized are band ‐ to

Orthogonal

LaRC Flexibility Simultaneous Decorrelates ‐ ‐ ‐ Inputs Phases balanced Wide Easy NASA • • • • Morelli