A Piloted Simulator Evaluation of Transport Aircraft Rudder Pedal Force/Feel Systems

1
A Piloted Simulator Evaluation of Transport
Aircraft Rudder Pedal Force/Feel Systems

• Eric C. Stewart
• NASA Langley Research Center
• 98th Aerospace Control and Guidance Systems
  Committee Meeting
• October 11-13, 2006
• Williamsburg, Virginia

2
Background

• American Airlines Flight 587 crashed on Nov. 12,
  2001 on Long Island, killing 265 people
• The accident was probably caused by the pilot
  over-controlling the rudder (or PIO) after an
  encounter with wake turbulence (NTSB/AAR-04/04)
• All rudder systems limit rudder (aerodynamic
  surface) travel at high speeds to protect against
  structural failure for static maneuvers such as
  cross-wind landings and engine failures
• For AA 587 it is thought that large, dynamic
  lateral-directional motions caused by rapid
  rudder pedal reversals led to structural failure
  of the vertical tail and complete loss of control
  (NTSB/AAR-04/04)
• According to several experts, certain rudder
  systems are tailor-made for over-control
  (Aviation Week April 1, 2002)

3
Background (continued)

• Two designs are commonly used (1) ratio
  changer and (2) variable stop or fixed ratio
• These two designs have vastly different pilot
  rudder pedal feel characteristics (pilot forces
  and deflections) which may cause over-control or
  PIOs
• The handling quality requirements governing
  airplane certification in the Federal Aviation
  Regulations and MIL-STD-8785 have very little to
  say about rudder pedal feel
• A literature search produced practically nothing
  relating to a systematic study of the handling
  qualities due to rudder pedal feel
  characteristics

4
Purpose of Study

• Conduct a systematic simulation study of the
  effects of pilot rudder pedal feel
  characteristics on the handling qualities of a
  transport airplane
• Results can be used to guide designers of rudder
  systems, as a basis for changing the
  certification requirements, or modifying existing
  systems

5
Langley Instrument Flight Deck (IFD) Simulator
6
Langley IFD Simulator
7
Candidate Maneuvers/Disturbances

• Operationally realistic maneuvers
• Gusts
• Wake vortex
• Engine surging
• Rudder actuator failure
• Artificial maneuvers
• Pop-up obstacles
• Arbitrary angular and/or linear displacements
• Flight condition
• High speed/altitude
• Low speed/altitude

8
Test Maneuver

• Combination realistic/artificial maneuver
• Produced most rudder pedal activity of maneuvers
  tested
• Approach in crosswind and random turbulence
• Severe lateral wind shear introduced around 125
  feet AGL
• No go-around or landing allowed
• Runway tracking at 50 feet AGL

9
Static Pedal Forces
Pedal Forces, lbs
M, force at maximum travel
B, breakout force
Slope or Stiffness
Pedal Deflection, inches
X, maximum travel
10
Pedal Feel Combinations
Central Composite Design (of Experiments)
(Numbers in cells indicate 15 actual test
combinations out of possible 125)
11
Pedal Feel Combinations
12
Lateral Wind Shear Scenarios
13
Test Subjects

• All active airline pilots operating Boeing
  equipment
• 7 males and 5 females
• 4 captains and 8 first officers
• Individual Total Hours 5,500 to 20,000, average
  11,000
• Individual Hours in command 500 to 18,500,
  average 5,000

14
Typical Time Histories(Longitudinal Parameters)
15
Typical Time Histories(Directional Parameters)
16
Response Surface Equation
Y b1 b2M b3B b4X (linear
terms) b5MB b6MX b7BX
(interaction terms) b8M2 b9B2
b10X2 (squared terms)
where
bs are constants determined from a least squares
fit
M force at maximum travel (lbs) B breakout
force (lbs) X maximum pedal travel (inches)
17
Response Surface Equation Predictions
18
Pilot Rating Contours
Maximum Travel 1.5 inches
Key
3.8
Min PR
(61,20)
(X, Y)
19
Pilot Rating Contours
Maximum Travel 2.5 inches
Key
3.2
Min PR
(80,19)
(X, Y)
20
Pilot Rating Contours
Maximum Travel 3.5 inches
Key
Min PR
2.7
(X, Y)
(98,18)
21
C-H Pilot Rating Contours
Breakout Force 26.5 lbs
Minimum is out of Range
22
PIO Tendencies (Time histories)
No Pilot induced oscillations
Pilot induced oscillations
Turbulence Induced
23
PIO Tendencies (Cross spectra)
No Pilot induced oscillations
Pilot induced oscillations
Peak Value
24
Peak Cross Spectra Contours
Max Travel 1.5 inches
Key
3.8
Min PS
(92, 25)
(X, Y)
25
Peak Cross Spectra Contours
Max Travel 2.5 inches
2.1
Key
(86, 28)
Min PS
(X, Y)
26
Peak Cross Spectra Contours
Max Travel 3.5 inches
1.5
Key
(80,30)
Min PS
(X, Y)
27
Peak Cross Spectra Contours
Breakout Force 26.5 lbs
1.5
(78, 3.4)
Key
Min PS
(X, Y)
28
Preliminary Results

• Method successfully quantified pedal feel
  characteristics
• Central Composite Design
• Response Surface Equation
• Averaged pilot ratings from line pilots gave
  consistent results for 6 or more pilots
• Results need to verified for
• Other configurations (e.g. wheel feel
  characteristics)
• Motion-base simulator
• Other maneuvers

29
Preliminary Results (concluded)

• Response Surface Equation is useful for
• Generating arbitrary contours of constant pilot
  ratings
• Revealing optimum combinations of pedal feel
  characteristics
• Peak values of cross spectra of pilot input and
  airplane response may be used to predict PIO
  tendencies
• Predicted PIO tendencies are generally consistent
  with pilot ratings
• A more complete report with added details is in
  preparation