NTSB Board MeetingAA Flight 587NTSB Board MeetingAA Flight 587

Airplane Motion and Vertical Stabilizer LoadsAirplane Motion and Vertical Stabilizer Loads

John O’CallaghanJohn O’Callaghan

-1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

0915:51 Start of second wake turbulence encounter

0915:36 Start of first wake turbulence encounter

Dis

tanc

e no

rth

of im

pact

site

, nm

Distance east of impact site, nm

31L

Bo

ein

g 7

47

Flig

ht 5

87

WIND

Location of Wake Turbulence Encounters

• FDR accelerations were typical of wake encounters

• Crew commented on wake turbulence

• Simulation indicates wake encounter

• NASA wake study supports encounter

• Wake was similar in each encounter

NTSB Board Meeting AA Flight 587

NTSB Board Meeting AA Flight 587

Effect of the Wake Encounters on the Airplane

Motion

Effect of the Wake Encounters on the Airplane

Motion• NASA study indicates nothing unusual about

wake.

• NTSB simulations determined that the effect of wake on airplane motion was minor.

• The airplane was not in or at risk of an upset.

• NASA study indicates nothing unusual about wake.

• NTSB simulations determined that the effect of wake on airplane motion was minor.

• The airplane was not in or at risk of an upset.

Control Inputs Following Start of First Wake Encounter

• First officer responded with column & large wheel inputs

• First officer did not use the rudder pedals

• Small changes in pitch and roll angles

• Airplane motion was unremarkable

09:15:34 09:15:36 09:15:38 09:15:40 09:15:42 09:15:44

100

50

0

-50

-100

09:15:34 09:15:36 09:15:38 09:15:40 09:15:42 09:15:44

3

2

1

0

-1

-2

-3

09:15:34 09:15:36 09:15:38 09:15:40 09:15:42 09:15:44

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:51

Control Inputs Following Start of Second Wake Encounter

• Start of second wake encounter

• Airplane in climbing left turn

• Controls approximately neutral

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Control Inputs Following Start of Second Wake Encounter

Time = 09:15:52

• Pedal used to help control roll • Pedal not necessary

• Large right wheel input

• Full right pedal input

• Wheel alone sufficient to control roll

• Full wheel and pedal inputs unnecessary and excessive

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:53.1

Control Inputs Following Start of Second Wake Encounter

First full alternating rudder pedal input

• Full left wheel input (78°)

• Full left pedal input

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:54.2

Control Inputs Following Start of Second Wake Encounter

Second full alternating rudder pedal input

• Full right pedal input

• Growing oscillation in column inputs

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:55.6

Control Inputs Following Start of Second Wake Encounter

• Full right pedal input maintained

• Wheel moves to large right deflection

• Large nose-down column input

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:57

Control Inputs Following Start of Second Wake Encounter

Third full alternating rudder pedal input

• Full left wheel input

• Full left pedal input

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

Time = 09:15:58.4

Control Inputs Following Start of Second Wake Encounter

Fourth full alternating rudder pedal input

• Wheel moves right

• Full right pedal input

• Vertical stabilizer separates from airplane

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

100

50

0

-50

-100

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

3

2

1

0

-1

-2

-3

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58

-10

-5

0

5

10

FWD COLUMN

AFT COLUMN

RIGHT WHEEL

RIGHT PEDAL

LEFT PEDAL

Co

lum

n, d

egre

es

LEFT WHEEL

Wh

eel,

deg

rees

Ped

al, i

nch

es

ATC Time, HH:MM:SS ESTTime

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00

-10

-5

0

5

10

NOSE RIGHT OF AIRFLOW

NOSE LEFT OF AIRFLOW

Sid

eslip

ang

le, d

egre

es

Airflow

Airflow

Sideslip Angle

Sideslip Angle Buildup Resulting From First Officer’s Control Inputs

• Airplane flew as commanded until vertical stabilizer separation

Vert

ical sta

b.

sep

ara

tion

NTSB Board Meeting AA Flight 587

NTSB Board Meeting AA Flight 587

Calculation of Vertical Stabilizer Loads

Calculation of Vertical Stabilizer Loads

• Loads dependent on airspeed, sideslip angle, and rudder deflection

• Aerodynamic loads determined by wind tunnel testing during airplane development

• No wind tunnel data available at the extreme sideslip angle corresponding to vertical stabilizer separation

• Other methods required to compute loads at time of separation

• Loads dependent on airspeed, sideslip angle, and rudder deflection

• Aerodynamic loads determined by wind tunnel testing during airplane development

• No wind tunnel data available at the extreme sideslip angle corresponding to vertical stabilizer separation

• Other methods required to compute loads at time of separation

NTSB Board Meeting AA Flight 587

NTSB Board Meeting AA Flight 587

Computational Fluid Dynamics (CFD)

Computational Fluid Dynamics (CFD)

• CFD is the use of computers to mathematically determine the aerodynamic characteristics of airplanes.

• CFD is used increasingly in the industry to supplement wind tunnel data and optimize airplane designs.

• CFD is the use of computers to mathematically determine the aerodynamic characteristics of airplanes.

• CFD is used increasingly in the industry to supplement wind tunnel data and optimize airplane designs.

NTSB Board Meeting AA Flight 587

NTSB Board Meeting AA Flight 587

Computational Fluid Dynamics (CFD)

Computational Fluid Dynamics (CFD)

• CFD studies directed by NTSB and reviewed by NASA Langley Research Center.

• CFD studies directed by NTSB and reviewed by NASA Langley Research Center.

• CFD is the use of computers to mathematically determine the aerodynamic characteristics of airplanes.

• CFD is used increasingly in the industry to supplement wind tunnel data and optimize airplane designs.

• Airbus CFD code has demonstrated capability for solving flow problems such as flight 587 vertical stabilizer loads.

• CFD is the use of computers to mathematically determine the aerodynamic characteristics of airplanes.

• CFD is used increasingly in the industry to supplement wind tunnel data and optimize airplane designs.

• Airbus CFD code has demonstrated capability for solving flow problems such as flight 587 vertical stabilizer loads.

CFD Results: Pressure Distribution Over Vertical Stabilizer

Chordwise Distance (mm)

Pre

ssur

e C

oeffi

cien

t

Left Side Right Side

Flow Separation

CFD Results: Streamlines of Flow at High Sideslip Angle

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

0r LIN38

TAIL BENDS RIGHT

TAIL BENDS LEFT

Time

Bending Moment History During Second Wake Encounter

Limit Load

Ultimate Load (1.5 x Limit)

Ultimate Load (1.5 x Limit)

Be

nd

ing

Mo

me

nt

/ Lim

it L

oa

d

Limit Load

Base ofVertical Stabilizer

Base ofVertical Stabilizer

Bending MomentBending Moment

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

0r LIN38

TAIL BENDS RIGHT

TAIL BENDS LEFT

Time

Limit Load

Ultimate Load (1.5 x Limit)

Ultimate Load (1.5 x Limit)

Be

nd

ing

Mo

me

nt

/ Lim

it L

oa

d

Limit Load

Bending Moment History During Second Wake Encounter

Limit load: • Highest load expected in lifetime

Time

Ultimate Load (1.5 x Limit)

Ultimate Load (1.5 x Limit)

Be

nd

ing

Mo

me

nt

/ Lim

it L

oa

d

Limit Load

• Determined by conditions specified in FARs

Limit Load

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

0r LIN38

TAIL BENDS RIGHT

TAIL BENDS LEFT

Bending Moment History During Second Wake Encounter

Ultimate load: • Equal to limit load times safety factor of 1.5

Time

Limit Load

Ultimate Load (1.5 x Limit)

Ultimate Load (1.5 x Limit)

Be

nd

ing

Mo

me

nt

/ Lim

it L

oa

d

Limit Load

• Structure must not break up to ultimate load

Bending Moment History During Second Wake Encounter

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00

-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

0r LIN38

TAIL BENDS RIGHT

TAIL BENDS LEFT

09:15:50 09:15:52 09:15:54 09:15:56 09:15:58 09:16:00-2.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

2.5

TAIL BENDS RIGHT

TAIL BENDS LEFT

Time

Limit Load

Ultimate Load (1.5 x Limit)

Ultimate Load (1.5 x Limit)

Be

nd

ing

Mo

me

nt

/ Lim

it L

oa

d

Limit Load

Wind tunnel analysisWind tunnel analysis

Range of loads at separation based on wind tunnel & CFD analysis

Range of loads at separation based on wind tunnel & CFD analysis

Bending Moment History During Second Wake Encounter

NTSB Board Meeting AA Flight 587

NTSB Board Meeting AA Flight 587

ConclusionsConclusionsAirplane encountered wake turbulence twice

• Indicated by FDR, CVR, simulation, and wake analysis

First officer’s control inputs following second encounter were unnecessary and excessive

• Simulation indicates wake had minor effect on motion

• Airplane was never in an upset condition

Airplane responded to control inputs as expected until vertical stabilizer separation

• Simulation indicates large sideslip angles were the result of control inputs

Vertical stabilizer separated at a bending moment load well above ultimate load

• Determined by wind tunnel and CFD analysis

Airplane encountered wake turbulence twice

• Indicated by FDR, CVR, simulation, and wake analysis

First officer’s control inputs following second encounter were unnecessary and excessive

• Simulation indicates wake had minor effect on motion

• Airplane was never in an upset condition

Airplane responded to control inputs as expected until vertical stabilizer separation

• Simulation indicates large sideslip angles were the result of control inputs

Vertical stabilizer separated at a bending moment load well above ultimate load

• Determined by wind tunnel and CFD analysis

National Transportation Safety BoardNational Transportation Safety Board

American Airlines Flight 587Belle Harbor, New YorkNovember 12, 2001

NTSB Board MeetingOctober 26, 2004

American Airlines Flight 587Belle Harbor, New YorkNovember 12, 2001

NTSB Board MeetingOctober 26, 2004