SilkAir MI 185 Aircraft Accident Report
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Investigation of Aircraft Accident
SilkAir Flight MI 185
Boeing B737-300, 9V-TRF
Musi River, Palembang, Indonesia
19 December 1997
FINAL REPORT
Note:
• All times indicated in this report are based on FDR UTC time.
Local time is UTC + 7 hours.
Abstract
This report is on the accident involving the SilkAir flight MI 185, a Boeing B737-300,
which crashed into the Musi river near Palembang, South Sumatra, Indonesia, on 19
December 1997, at about 16:13 local time (09:13 UTC).
SilkAir flight MI 185 was operating as a scheduled passenger flight from Jakarta
Soekarno-Hatta International Airport to Singapore Changi Airport.
The flight departed about 15:37 local time with 97 passengers, five cabin crew and two
cockpit crew.
The airplane descended from its cruising altitude of 35,000 feet and impacted the Musi
river, near the village of Sunsang, about 30 nautical miles north-north-east of Palembang
in South Sumatra.
Visual meteorological conditions prevailed for the flight, which operated on an
instrument flight rules flight plan.
Prior to the sudden descent from 35,000 feet, the flight data recorders stopped recording
at different times. There were no mayday calls transmitted from the airplane prior or
during the descent.
All 104 persons on board did not survive the accident, and the airplane was completely
destroyed by impact forces. Except parts of the empennage that found on land, most of
the wreckage was found buried in the bottom of the Musi river.
About 73 percent by weight of wreckage was recovered, although due to the magnitude of
destruction of the airplane, the small mangled pieces precluded finding clues, evidence or
proof as to what have happened, how and why.
The safety issues in this report focused on the areas of flight operations, the flight data
recorders, the human factors and control systems malfunctions.
The investigation yielded very limited data and information to make conclusions possible.
The report is pursuant to the technical investigation conducted by the National
Transportation Safety Committee (NTSC) of Indonesia.
The investigation was conducted in accordance with the standards and recommended
practices of Annex 13 to the Convention on International Civil Aviation. In accordance
with Annex 13, the sole objective of the investigation of an accident or incident shall be
the prevention of accidents and incidents. It is not the purpose of this activity to apportion
blame or liability.
i
Table of Contents
Abstract
i
Abbreviations
v
Glossary of Terms
1
viii
FACTUAL INFORMATION
1
1.1
History of Flight
1
Synopsis
1
1.2
1.3
1.4
1.5
Injuries to Persons
Damage to Aircraft
Other Damage
Personnel Information
1
2
2
2
1.5.1
1.5.2
2
2
1.6
1.7
1.8
1.9
1.10
1.11
Pilot-In-Command (PIC)
First Officer (F/O)
Aircraft Information
3
1.6.1
1.6.2
1.6.3
3
3
4
Aircraft Data
Aircraft History
Weight and Balance
Meteorological Information
4
1.7.1
1.7.2
1.7.3
1.7.4
4
4
4
5
General Weather Condition over South Sumatra
Weather Report over Palembang and Its Surroundings
Information on Wind Directions and Strength
Weather En-route
Aids to Navigation
Communications
Aerodrome Information
Flight Recorders
5
5
5
6
1.11.1 FDR
1.11.2 FDR Data Recovery
1.11.3 CVR
6
6
7
1.12 Wreckage and Impact Information
1.12.1 Aircraft Structures
1.12.1.1
1.12.1.2
1.12.1.3
1.12.1.4
1.12.1.5
Wings
Fuselage
Empennage
Horizontal Tailplane
Vertical Tailplane
1.12.2 Power plants
1.12.2.1 Engine
1.12.2.2 Main Engine Control (MEC) / Governor
1.12.2.3 Throttle Box
1.12.3 PCU and Actuator Tear Down and Examination
1.12.3.1 Tear Down Activities
1.12.3.2 Spoiler/Flap/Slat/Thrust Reverser Actuators
1.12.3.3 Actuators Found in Non-neutral Position
1.12.4 Other Aircraft Components
1.13 Medical and Pathological Information
1.14 Fire
1.15 Survival Aspects
7
8
9
9
10
10
12
12
13
13
14
14
14
15
15
17
17
17
18
i
1.16 Tests and Research
1.16.1
1.16.2
1.16.3
1.16.4
1.16.5
1.16.6
CVR Circuit Breaker Actuation Test
Captain Seat Belt Buckle Sound Test
Voice Recognition of ATC Recording using Audio Spectral Analysis
Trajectory Studies
Flutter Studies
Flight Simulation Tests
1.17 Organizational and Management Information
1.18 Other Information
1.18.1 Air Traffic Control
1.18.2 Radar Surveillance
1.18.2.1 Radar Facilities.
1.18.2.2 Radar Data Output
1.18.2.3 Aircraft Flight Path Based on Radar
1.18.3 PIC’s Background and Training
1.18.3.1
1.18.3.2
1.18.3.3
1.18.3.4
Professional Background in RSAF
Professional Background with SilkAir.
Financial Background Information
Recent Behaviour
1.18.4 F/O’s Background and Training
1.18.4.1 Professional Background with SilkAir
1.18.4.2 Financial Background Information
1.18.4.3 Recent Behaviour
1.18.5 Relationship Between the PIC and the F/O
2
18
18
19
20
20
21
21
22
23
23
23
23
24
24
24
24
25
25
26
26
26
27
27
27
ANALYSIS
28
2.1
2.2
Introduction
Aircraft Structural and Systems Integrity
28
28
2.2.1
2.2.2
29
29
2.3
2.4
2.5
2.6
2.7
2.8
Horizontal Stabilizers and Elevators
Vertical Stabilizer and Rudder
Break Up of the Empennage
30
2.3.1
2.3.2
2.3.3
30
30
31
Results of Trajectory Studies
Results of Flutter Studies
Explanation to the Break Up of the Empennage
Power Control Units and Actuators
31
2.4.1
2.4.2
2.4.3
2.4.4
2.4.5
2.4.6
31
32
32
32
32
34
Main Rudder PCU
Standby Rudder PCU
Aileron PCU
Elevator PCU
Horizontal Stabilizer Jackscrew
Other Actuators
Powerplant
Stoppage of the CVR and FDR
34
35
2.6.1
2.6.2
35
36
CVR Stoppage
FDR Stoppage
Radio Transmission Voice Recognition
Maintenance Aspects
36
37
2.8.1
2.8.2
37
37
Aircraft Maintenance
Patch Repair
2.9 General Operational Issues
2.10 Simulated Descent Profile
2.11 High Speed Descent Issues
2.11.1 Mach Trim System and its Function
2.11.2 Emergency Descent due to Fire, Smoke or Depressurization
37
38
39
39
39
ii
2.12 General Human Performance Issues
2.13 Human Factors Aspects of the CVR and ATC Recordings
2.13.1 CVR
2.13.2 ATC Recordings
2.14 Specific Human Factors Issues
2.14.1 Personal Relationships
2.14.2 First Officer (F/O)
2.14.3 Pilot-in-Command (PIC)
2.14.3.1 Recent Behaviour
2.14.3.2 Insurance
2.14.3.3 Overall Comments on Pilot-in-Command
3
CONCLUSIONS
3.1
3.2
4
40
41
41
41
42
42
42
42
43
43
44
45
Findings
45
Engineering and Systems
Flight Operations
Human Factors
45
45
46
Final Remarks
46
RECOMMENDATIONS
Recommendations to manufacturers
General recommendation
REFERENCES
48
48
48
49
FIGURES
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
Figure 17
Figure 18
Route from Jakarta to Singapore over Sumatra, Indonesia
Sequence of events
Boeing B737-300 – Three view drawing
Flight Data Recorder (FDR)
Cockpit Voice Recorder (CVR)
Debris recovery
Sketch of the wreckage pieces found on land
Picture of wreckage recovered from the river
Sketch of fuselage skin patch repair
Sketch of empennage parts found on land
Picture of the reconstructed empennage
Impact marks at the cam feel centering unit
Pictures of horizontal stabilizer jackscrew
Diagram of debris distribution analysis
Boeing B737-300 flutter flight envelope diagram
Corrected radar data
Wiring diagram 24-58-11
Flight control surfaces location
50
52
53
54
56
57
59
60
64
65
75
76
77
80
83
84
85
86
APPENDICES
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
Transcript of the Last Portion of CVR Recording
Plots of the Last Portion of Several FDR Parameters
Transcript of ATC recording
Wreckage weight
Actuator matrix
Letters from AlliedSignal
A-1
B-1
C-1
D-1
E-1
F-1
iii
Appendix G
Appendix H
Appendix I
Appendix J
Appendix K
Appendix L
Results of flight simulation exercises
G-1
Site Acceptance Certificate of the Hughes GUARDIAN System H-1
Professional events in the Flight Crew’s career during 1997
I-1
History of FAA AD related to Boeing 737 rudder system
J-1
Boeing B737 non-normal procedures - emergency descent
K-1
Boeing B737 Alert Service Bulletin, Subject on Flight Controls –
Trailing Edge Flap and Horizontal Stabilizer Trim Systems
L-1
Appendix M Singapore Accredited Representative’s Comments on
Draft Final Report
M-1
Appendix N USA Accredited Representative’s Comments on
Draft Final Report
N-1
iv
Abbreviation
AAIB
AAIC
AC
AD
AFM
AFS
agl
APU
ASB
ASRS
ATC
ATP
ATPL
ATS
Air Accidents Investigation Branch
Aircraft Accident Investigation Commission
Advisory Circular
Airworthiness Directive
Airplane Flight Manual
Auto-Flight System
above ground level
Auxiliary Power Unit
Alert Service Bulletin
Aviation Safety Reporting System
Air Traffic Control
Airline Transport Pilot
Airline Transport Pilot License
Air Traffic Services
BASI
BEA
Bureau of Aviation Safety Investigation
Bureau Enquétes Accidents
CAAS
CAM
CB
CG
CPL
CRM
CRT
CVR
Civil Aviation Authority of Singapore
Cockpit Area Microphone
Circuit Breaker
Center of Gravity
Commercial Pilot License
Crew Resource Management
Cathode Ray Tube
Cockpit Voice Recorder
DFDAU
DME
Digital Flight Data Acquisition Unit
Distance Measuring Equipment
E&E bay
EDP
Elex
EQA
Electrical and Electronic compartment
Engine Driven Hydraulic Pump
Electronics
Equipment Quality Analysis
F
FAA
FAR
FD
FDR
FIR
FL
FTIR
F/O
GE
GPS
Fahrenheit
Federal Aviation Administration
Federal Aviation Regulations
Flight Director
Flight Data Recorder
Flight Information Region
Flight Level
Fourier Transfer Infrared
First Officer
General Electric
Global Positioning System
v
GPWS
Ground Proximity Warning System
Hg
HPa
HQ
Hz
Mercury
Hecto Pascal
Head Quarters
Hertz
ICT
IFR
ILS
IP
IRS
Industrial Computed Tomography
Instrument Flight Rules
Instrument Landing Systems
Instructor Pilot
Inertial Reference System
JKT
Jakarta
KCAS
KEAS
Kg
KIAS
kV
Knots Calibrated Airspeed
Knots Equivalent Airspeed
Kilogram
Knots Indicated Airspeed
Kilo Volt
L
LIP
LOFT
Left
Line Instructor Pilot
Line Oriented Flight Training
MAC
MCIT
M-CAB
MEC
MHz
MM
MSL
Mean Aerodynamic Chord
Ministry of Communication and Information Technology
Multipurpose Cab (simulator)
Main Engine Control
Mega Hertz
Maintenance Manual
Mean sea level
N1
N2
NG
nm
NNW
NTSB
NTSC
Engine Fan Speed
Engine Compressor Speed
Next generation
Nautical mile
north-north-west
National Transportation Safety Board
National Transportation Safety Committee
PA
PATS
PCU
PF
PIC
PLB
P/N
PNF
Psi
Public address
Play-back and Test System
Power Control Unit
Pilot flying
Pilot in Command
Palembang
Part number
Pilot non flying
Pounds per square inch
vi
PWC
PricewaterhouseCoopers
QAR
Quick access recorder
R
RA
RAPS
RSAF
Right
Radio altitude
Replay and Presentation System
Republic of Singapore Air Force
SB
SEM
SIA
SIAEC
SIN
SL
S/N
SRM
SSCVR
SSR
Service bulletin
Scanning electron microscope
Singapore International Airlines
Singapore Airlines Engineering Company
Singapore
Service letter
Serial number
Structure Repair Manual
Solid State Cockpit Voice Recorder
Secondary Surveillance Radar
TCAS
TOGA
T/R
Traffic Collision Avoidance System
Take-off / go-around
Thrust reverser
UFDR
ULB
USA
UTC
Universal Flight Data Recorder
Underwater Locator Beacon
United States of America
Universal Time Coordinated
Vd
VFR
VHF
VOR
VSV
Descent speed
Visual flight rules
Very High Frequency
Very high frequency Omni directional Range
Variable Stator Vane
XRD
X-ray diffraction
vii
Glossary of Terms
Actuator
Aileron
Auto-flight system
Bank
Blowdown limit
Catastrophic
failure condition
Command mode
Computer
simulation
Control wheel
steering mode
Cross-coupled
Crossover speed
Directional
control
E&E bay
Elevator
Empennage
Flap
A device that transforms fluid pressure into mechanical fore
An aerodynamic control surface that is attached to the rear, or
trailing edges of each wing. When commanded, the ailerons rotate
up or down in opposite directions
A system, consisting of the auto-pilot flight director system and the
auto-throttle, that provides control commands to the airplane's
ailerons, flight spoilers, pitch trim, and elevators to reduce pilot
workload and provide for smoother flight. The auto- flight system
does not provide control commands to the airplane's rudder system
The attitude of an airplane when its wings are not laterally level
The maximum amount of rudder travel available for an airplane at
a given flight condition / configuration. Rudder blowdown occurs
when the aerodynamic forces acting on the rudder become equal to
the hydraulic force available to move the rudder
A failure condition that will prevent continued safe flight and
landing
A position on the two autopilot flight control computers, that, when
engaged, allows the autopilot to control the airplane according to
the mode selected via the Mode Selector switches, which include
Altitude hold, Vertical Speed, Level Change, Vertical Navigation,
VOR Localizer, Lateral Navigation and Heading Select
The NTSB computer workstation-based flight simulation software
used flight controls, aerodynamic characteristics, and engine
models (developed by Boeing) to derive force and moment time
histories of the airplanes, which can be converted into airplane
motion
A position on the two autopilot flight control computers that, when
engaged, allows the autopilot to maneuver the airplane through the
autofilight system in response to control pressure, similar that
required for manual flight, applied by either pilot. The use of
control wheel steering does not disengage the autopilot.
The ability of the aerodynamic motion about an airplane's control
axes to constantly interact and affect each other in flight.
The speed below which the maximum roll control (full roll
authority provided by control wheel input) can no longer counter
the yaw / roll effects of a rudder deflected to its blowdown limit.
The function that is normally performed by the rudder by pilot
input or yaw damper input (also known as yaw control)
An airplane compartment that contains electrical and electronic
components.
An aerodynamic control surface to the back of the horizontal
stabilizer that moves the airplane's nose up and down to cause the
airplane to climb or descend.
The tail section of an airplane, including stabilizing and flight
control surfaces
An extendable aerodynamic surface usually located at the trailing
viii
G
Galling
Heading
Hinge moment
Hydraulic fluid
Hydraulic
pressure limiter
Hydraulic
pressure reducer
Hydraulic system
A
Hydraulic system
B
Input shaft of the
737 main rudder
PCU
Interpolation
edge of an airplane wing.
A unit of measurement. One G is equivalent to the acceleration
caused by the earth's gravity (32.174 feet/sec2)
A condition in which microscopic projections or asperities bond at
the sliding interface under very high local pressure. Subsequently,
the sliding forces fracture the bonds, tearing metal from one
surface and transferring it to the other.
The direction (expressed in degrees between 001 and 360) in which
the longitudinal axis of an airplane is point, in relation to north
The tendency of a force to produce movement about a hinge.
Specifically the tendency of the aerodynamic forces acting on a
control surface
Liquid used to transmit and distribute forces to various airplane
components that are being actuated.
A device incorporated in the design of the main rudder PCU on 737
next generation (NG) series airplanes to reduce the amount of
rudder deflection when active. It is commanded to limit hydraulic
system A pressure (using a bypass valve) as the airspeed is
increased to greater than 137 knots, and it is reset as the airspeed is
decreased to less than 139 knots.
A modification on 737-100 through -500 series airplanes to reduce
the amount of rudder authority available during those phases of
flight when large rudder deflections are not required. The pressure
reducer, added to hydraulic system A near the rudder PCU, will
lower the hydraulic pressure from 3,000 to 1,000 pounds psi on
737-300, -400, and -500 series airplanes or to 1,400 psi on 737100, and -200 series airplanes.
For 737-300, -400, and -500 series airplanes: A system that
includes an engine-drive hydraulic pump and an electrically
powered pump that provides power for the ailerons, rudder,
elevators, landing gear, normal nosewheel steering, alternate
brakes, inboard flight spoilers, left engine thrust reverser, ground
spoilers, the system A autopilot, and the autoslats through the
power transfer unit
For 737-300, -400, and -500 series airplanes: A system that
includes an engine-drive hydraulic pump and an electrically
powered pump that provides power for the ailerons, rudder,
elevators, trailing edge flaps, leading edge flaps and slats, autoslats,
normal brakes, outboard flight spoilers, right thrust reverse, yaw
damper, the system B autopilot, autobrakes, landing gear transfer
unit, and alternate nose-wheel steering (if installed).
When rudder motion is commanded, this device moves the primary
and secondary dual-concentric servo valve slides by way of the
primary and secondary internal summing levers to connect
hydraulic pressure and return circuits from hydraulic systems A
and B so that hydraulic pressure is ported to the appropriate slides
of the dual tandem actuator piston to extend or retract the main
rudder PCU piston rod
The determination, or approximation of unknown values based on
known values
ix
Kinematics
Knot
M-CAB
NG
Overtravel
Pitch control
Power control
unit (PCU)
Roll
Roll control
Rudder
Reverse rudder
response
Rudder hardover
Rudder trim
Sideload
Sideslip
Slat
Spoiler
Yaw
Yaw control
Yaw damper (in
the 737 main
rudder PCU)
A process involving fitting curves through available FDR data
(such as heading, pitch and roll), obtaining flight control time
history rates from these curves, and obtaining accelerations from
these accelerations using Newton's laws.
A velocity of one nautical mile per hour.
A Boeing multi-cabin flight simulator that can be modified to
simulate a variety of aircraft models and scenarios. It is an
engineering simulator that is capable of simulating events that are
outside of normal flight regimes, but it is not used for flight
training.
Boeing's next generation 737 series, designated as the 737-600, 700, -800, and -900 models
The ability of a device to move beyond its normal operating
position or range.
The function that is performed by the elevator by moving the
control column forward or aft, which raises or lowers the nose of
the airplane
A hydraulically powered device that moves a control surface, such
as a rudder, elevator, and aileron
Rotation of an airplane about its longitudinal axis
The function that is performed by the ailerons and flight spoilers by
moving the control wheel to the right or the left.
An aerodynamic vertical control surface that is used to make the
airplane yaw, or rotate, about its vertical axis
A rudder surface movement that is opposite to the one commanded
The sustained deflection of a rudder at its full (blowdown) travel
position
A system that allows pilots to command a steady rudder input
without maintaining foot pressure on the rudder pedals. It can be
used to compensate for the large yawing moments generated by
asymmetric thrust in an engine-out situation
The effect of lateral acceleration, typical the result of sideslip or
yaw acceleration
The lateral angle between the longitudinal axis of the airplane and
the direction of motion (flight path or relative wind). It is normally
produced by rudder forces, yawing motion resulting from
asymettrical thrust, or lateral gusts
An aerodynamic surface located on an airplane wing's leading edge
that may be extended to provide additional lift
A device located on an airplane wing's upper surface that may be
activated to provide increased drag and decreased lift
Rotation of an airplane about its vertical axis
The function that is normally performed by the rudder by pilot
input or yaw damper input, also known as directional control
A system composed of the yaw damper control switch and a yaw
damper coupler, that automatically corrects for yaw motion.
x
1
1.1
FACTUAL INFORMATION
History of Flight
Synopsis
On 19 December 1997, a SilkAir Boeing B737-300 aircraft, registration 9V-TRF, was on
a scheduled commercial international passenger flight under Instrument Flight Rules
(IFR), routing Singapore – Jakarta – Singapore.
The flight from Singapore to Jakarta operated normally. After completing a normal turnaround in Jakarta the aircraft departed Soekarno-Hatta International Airport for the return
leg.
At 08:37:13 (15:37:13 local time) the flight (MI 185) took off from Runway 25R with the
Captain as the handling pilot. The flight received clearance to climb to 35,000 feet (Flight
Level 350) and to head directly to Palembang1. At 08:47:23 the aircraft passed FL245.
Ten seconds later, the crew requested permission to proceed directly to PARDI2. The air
traffic controller instructed MI 185 to standby, to continue flying directly to Palembang
and to report when reaching FL350. At 08:53:17, MI 185 reported reaching FL350.
Subsequently, the controller cleared MI 185 to proceed directly to PARDI and to report
when abeam Palembang.
At 09:05:15.6, the cockpit voice recorder (CVR) ceased recording. According to the
Jakarta ATC transcript, at 09:10:18 the controller informed MI 185 that it was abeam
Palembang. The controller instructed the aircraft to maintain FL350 and to contact
Singapore Control when at PARDI. The crew acknowledged this call at 09:10:26. There
were no further voice transmissions from MI 185. The last readable data from the flight
data recorder (FDR) was at 09:11:27.4. Jakarta ATC radar recording showed that MI 185
was still at FL350 at 09:12:09. The next radar return, eight seconds later, indicated that
MI 185 was 400 feet below FL350 and a rapid descent followed. The last recorded radar
data at 09:12:41 showed the aircraft at FL195. The empennage of the aircraft
subsequently broke up in flight and the aircraft crashed into the Musi river delta, about 50
kilometres (30 nautical miles) north-north-east of Palembang at about 09:13. The accident
occurred in daylight and in good weather conditions.
The route map and the crash site are depicted in Figure 1. The sequence of events is
shown schematically in Figure 2.
1.2
Injuries to Persons
Injuries
Fatal
Serious
Minor/None
Total
Crew
7
0
0
7
Passengers
97
0
0
97
Others
0
0
0
0
1
Coordinates (02.52.7S, 104.39.2E)
Air Traffic Control reporting point (00.34.0S, 104.13.0E) north of Palembang in the Jakarta FIR near the
boundary with the Singapore FIR. At PARDI, flights are transferred over to Singapore ATC
2
1
Very few human remains were recovered from the crash site and only six positive
identifications were able to be made.
1.3
Damage to Aircraft
The aircraft was completely destroyed and severely fragmented on impact with the Musi
river. The wreckage had penetrated deep into the river bottom. The destruction was such
that mainly small mangled parts were recovered from the river. Parts of the rudder skin
and the outboard sections of the horizontal stabilizer were recovered on land, the furthest
was about four kilometres from the main impact site.
1.4
Other Damage
There was no damage to any other property.
1.5
1.5.1
Personnel Information
Pilot-In-Command (PIC)
Sex
Age
Date of joining SilkAir
Licence country of issue
Licence type
Licence number
Validity period of licence
Ratings
Medical certificate
Aeronautical experience
Experience on type
Last 24 hours
Last 7 days
Last 28 days
Last 90 days
Last line check
Last proficiency check
Instrument rating check
1.5.2
Male
41 years
1 March 1992
Singapore
ATPL (Airline Transport Pilot Licence)
501923
1 November 1997 to 30 April 1998
Boeing B737; Airbus A310 (not current)
First class – issued 10 October 1997
7173.3 hours
3614.7 hours
1.6 hours
20.1 hours
56.8 hours
216.7 hours
25 January 1997
7 August 1997
7 August 1997
First Officer (F/O)
Sex
Age
Date of joining SilkAir
Licence country of issue
Licence type
Licence number
Validity period of licence
Male
23 years
16 September 1996
Singapore
CPL (Commercial Pilot Licence)
503669
1 July 1997 to 30 June 1998
2
Ratings
Medical certificate
Aeronautical experience
Experience on type
Last 24 hours
Last 7 days
Last 28 days
Last 90 days
Last line check
Last proficiency check
Instrument rating check
1.6
1.6.1
Aircraft Information
Aircraft Data
Manufacturer:
Model:
Serial Number:
Registration:
Country of manufacture:
Date of manufacture:
Certificate of airworthiness:
Issued:
Certificate of registration:
Issued:
Total airframe hours:
Engines:
Engine type:
Engine No1:
Total engine hours:
Total engine cycles:
Engine No2:
Total engine hours:
Total engine cycles:
Last Major Inspection:
Last Inspection:
Hours since last inspection:
Cycles since last inspection:
Total airframe cycles:
1.6.2
Boeing B737
First class – Issued 4 June 1997
2501.7 hours
2311.8 hours
1.6 hours
21.4 hours
69.8 hours
217.6 hours
10 October 1997
15 September 1997
15 September 1997
Boeing Aircraft Company
B737-300 (see Figure 3)
28556
9V-TRF
United States of America
14 February 1997
AWC 433
14 February 1997 (one year validity)
S 153
14 February 1997
2238.6 hours
2 x CFM 56-3B2 (CFM International)
Turbofan
Serial Number: 858-480
2238
1306
Serial Number: 858-481
2238
1306
Equalised Check #3 on 9-11 December 1997
9-11 December 1997
67 hours
43 cycles
1306 cycles
Aircraft History
The aircraft was new when it was registered in the Republic of Singapore as 9V-TRF. Its
Certificate of Registration was issued on 14 February 1997. At the time of the accident,
the aircraft was operated by SilkAir Pte Ltd. It had completed all of its flight time with
SilkAir.
3
1.6.3
Weight and Balance
Maximum
61,234 kg
47,627 kg
--8 – 27 %MAC
5,710 kg
----
Take-off weight:
Zero fuel weight:
Fuel at take-off:
Dry Operating Weight:
Take-off CG position:
Cargo on board:
Total number of persons:
Number of passengers:
Number of crew:
1.7
1.7.1
Actual
51,856 kg
45,056 kg
6,800 kg
34,192 kg
18.1 %MAC
3,654 kg
104
97
7
Meteorological Information
General Weather Condition over South Sumatra
Data gathered from the weather satellite imagery at 09:00 on 19 December 1997 and
ground observation showed that South Sumatra was partly covered in some areas with
cumulus, altocumulus and cirrus clouds, except over the northern part of South Sumatra
(Singkep Island) which was covered with cumulonimbus clouds associated with
thunderstorms.
1.7.2
Weather Report over Palembang and Its Surroundings
Weather observations received from weather stations around the area of the accident were
as follows:
Station
Palembang
Pangkal Pinang
Rengat
Dabo Singkep
Jambi
1.7.3
Type of cloud
CuSc/Ac/Ci
CuSc/AcAs/Ci
Cu/As/Ci
Cb/Cu/As/Cs
Cu/Ac/Ci
Weather
Nil
Nil
Nil
Thunderstorm
Nil
Temp/Dew point
30 / 25
29 / 25
32 / 23
26 / 25
32 / 29
Information on Wind Directions and Strength
The reported wind directions and strength at 09:00 over South Sumatra were as follows:
Level
700 hPa (10,000 feet)
500 hPa (19,000 feet)
400 hPa (25,000 feet)
300 hPa (31,000 feet)
250 hPa (35,000 feet)
Direction
North-westerly
North-easterly
South-easterly
South-easterly
Easterly
Speed
15 knots
10 knots
5 knots
10-15 knots
15-20 knots
4
1.7.4
Weather En-route
Weather observed and reported from other aircraft:
•
The PIC of Sempati 134 from Jakarta to Batam flying at FL310 (approximately two
minutes ahead of MI 185) reported that due to weather conditions en-route he
requested to fly direct PARDI.
•
The PIC of Qantas 41 from Jakarta to Singapore flying at FL410, (approximately
eight minutes behind MI 185) reported that the weather was good except for two or
three isolated thunderstorms about ten miles to the east of track near Palembang. He
also reported that no turbulence was encountered during his flight except for the last
five thousand feet of the descent into Singapore.
1.8
Aids to Navigation
Navigational aids for en-route, approach and landing (VOR/DME, NDB and ILS)
covering the route of flight from Jakarta to Singapore were operating normally at the time
of the accident. The aircraft was tracking from a point near Jakarta directly to Palembang,
then to PARDI. At PARDI, flights are transferred over to Singapore ATC
The data from the Palembang radar were transmitted to Jakarta for en route air traffic
control by Jakarta ATC. These radar data were then recorded in the Jakarta ATC facility,
[Reference 1].
1.9
Communications
Up to 09:10:26, the aircraft was communicating normally with Jakarta Control on
frequency 132.70 MHz. There was no distress call received from the crew or distress
signal received from the aircraft transponder. No emergency locator beacon transmissions
were detected near the crash site. A transcript of the communications between the aircraft
and ATC is given in Appendix C.
1.10 Aerodrome Information
The Soekarno-Hatta International Airport is located about 20 kilometres west of Jakarta
(the crash site is approximately 420 kilometres north-north-west from the Jakarta airport).
The airport has two concrete parallel runways, 07R/25L (3,360 metres) and 07L/25R
(3,600 metres), with elevation between six and nine metres.
On the day of the accident both runways were checked and found in normal condition.
The MI 185 flight took off from runway 25R using standard instrument departure
Cengkareng 2 Golf.
Security checks of passengers, baggage and cargo were conducted in accordance with the
standard procedures, [Reference 1].
5
1.11 Flight Recorders
1.11.1 FDR
The aircraft was equipped with a Sunstrand Data Corporation Universal Flight Data
Recorder. The aircraft was fitted with a Teledyne common box Digital Flight Data
Acquisition Unit (DFDAU). The DFDAU formats 296 parameters into a 64 words per
second serial data stream, which is routed to the FDR.
The FDR module was recovered by divers on 24 December 1997. The FDR’s underwater
locator beacon (ULB) was found detached from the FDR at the accident site. Two of the
enclosure’s four retaining bolts were broken. The enclosure was partially opened (Figure
4.a). The partial opening exposed the tape between the two tape reels, near the write and
erase heads, see Figure 4.b and c.
The FDR module was first cleaned and then packed in a container filled with clean water
(to prevent the tape medium from drying out and becoming brittle). It was hand-carried
to the United States National Transportation Safety Board (NTSB) HQ read-out facilities
in Washington DC, USA3. The readout was accomplished by a play back hardware using
a Replay and Presentation System software (RAPS) developed by the Canadian
Transportation Safety Board, [Reference 2].
The data on the FDR tape was able to be read except from the short length of the exposed
tape mentioned above. By examining the physical position of the tape as it was found
when the casing was opened, it was determined that approximately 8 cm (3.15 inches) of
the exposed tape contained data of the accident flight. The data (about 6.3 seconds) were
not recovered because of damage to the exposed part of the tape.
The last recorded data that could be recovered was recorded at 09:11:27.4 (3,148.4 FDR
Sub-frame Reference Time). At this time, the aircraft was in a flight configuration
consistent with cruise flight at FL350. Based on the last recoverable data and the length of
the damage tape, it is estimated that the FDR could have been stopped recording at
09:11:33.7.
1.11.2 FDR Data Recovery
The FDR tape was re-examined at the FDR manufacturer’s facility in Redmond,
Washington, USA on 22 January 1998. Examination at the manufacturer revealed that in
several areas of the tape the magnetic oxide had separated from the Mylar backing.
Further attempts to retrieve information from the damaged portion were made by
Quantegy, the tape manufacturer in Alabama, USA (March 1998), Digital Instruments
(August 1998), the BEA - Bureau Enquétes-Accidents (December 1998) in Paris, France
and again by Quantegy (March 1999). The method applied at BEA utilized a Garnet
Microscope, while the test method used at the last attempt at Quantegy was destructive to
the tape. These efforts did not disclose any additional information, [Reference 3].
The relevant extracts from the FDR are contained in Appendix B.
3
This paragraph has been modified according to the suggestion in Appendix M
6
1.11.3 CVR
An AlliedSignal Aerospace Solid State Cockpit Voice Recorder (SSCVR) with a duration
of two hours was installed in the aircraft. Two channels were recorded, one channel
dedicated to the cockpit area microphone (CAM) and the other channel recorded a
combination of three other audio sources (Captain, First Officer and Observer). The last
thirty minutes of recording contained four discrete channels of the four audio sources.
The memory module was recovered, in a relatively good condition, from the river by the
dredge on 8 January 1998. The CVR’s underwater locator beacon (ULB) was found
detached from the CVR at the accident site.
The memory module was hand carried (immersed in water) to the NTSB in Washington
DC for read out and analysis (see Figure 5.a). The memory board (Figure 5.b) was
downloaded and decompressed using AlliedSignal’s Playback and Test Station (PATS)
hardware and software. The resulting download produced an excellent quality 30 minute
4 channel recording and an excellent quality two hour, one minute and 11 seconds 2
channel recording. There were no transients or identifiable signatures associated with the
end of recording. The recording included the ground operations and takeoff from Jakarta
until the CVR stopped at 09:05:15.6, [Reference 4]. It was verified that all the available
memory had been correctly downloaded and there were no missing memory blocks,
hidden or corrupted data.
A transcript of the last portion of the CVR is given in Appendix A.
1.12 Wreckage and Impact Information
Intensive searches were carried out by air in an area stretching up to nine kilometres (five
nautical miles) to the east of the crash site. Land searches were performed in an area
stretching up to five kilometres to the east of the crash site. Most of the aircraft wreckage
was found at the crash site, concentrated in an area approximately 60 metres by 80 metres
in the Musi river, which is approximately 700 metres wide and about eight metres deep at
that location, see Figure 6.
Recovery of the wreckage was extremely difficult due to the poor visibility in the water
and the fact that a lot of the wreckage had settled and got buried in the mud at the bottom
of the river. Moreover, there was also a strong tidal current flow. Wreckage recovery
during the early phases of the recovery operation was done manually by divers from the
Indonesian and Singapore navies. The divers had to search for the wreckage by touch and
use ropes to bring it to the surface. After a two-week period, dredging was employed for
the recovery operation. A clam shell dredge systematically scoured the river bed at the
crash site down to a depth of five metres below the clay surface.
The wreckage was very fragmented. The debris was deposited on a barge for the clay to
be washed away before being transported to a hangar at Palembang Airport, for cleaning
and disinfections before sorting and identification.
7
At the end of the recovery operation, including ten days of dredging, about 73 % by
weight of the aircraft structure and system/actuator components had been recovered.
Parts of the empennage were found on land. They comprised the outboard sections of the
right hand and left hand horizontal stabilizers, sections of the elevators, elevator tabs,
rudder upper balance weight and honeycomb skin pieces. These pieces were found
generally in an easterly direction from the impact site. The wreckage distribution, as
determined by a hand-held GPS (Global Positioning System) unit, is shown in Figure 7.
The right hand elevator tip was found about four kilometres to the east of the crash site.
This was the furthest location where the wreckage pieces were found. The left hand
elevator tip was found about 200 metres from the east river bank (or 700 metres from the
location of the wreckage). This was the nearest location to the crash site among the
wreckage pieces found on land. Lighter panels (composite materials) from skin of rudder
and elevators were also found on land. These were picked up by local residents and
handed over to the police and search and rescue helicopter pilots. The exact locations
where the lighter panels were originally found could not be determined.
The aircraft parts found in the river were highly fragmented and mangled on impact, see
Figure 8. As a result, all parts of the aircraft were mixed together on recovery, making
sorting and identification of the many small pieces difficult. Particular attention was paid
to finding parts of the control surfaces, actuators, electronic and avionics systems,
engines, Section 48 (rear fuselage including aft pressure bulkhead) and the empennage
(see Figure 10). The cockpit instrumental panel and circuit breakers panel were not
recovered. The wing flaps, spoilers, ailerons, and leading edge slats and their supporting
tracks, carriages and actuation systems were also accounted for. Cabin items such as seat
structures, cushions, oxygen generators, etc. were also recovered and examined. Parts and
components recovered were contaminated, stained and showing accelerated corrosion.
River water and mud samples were taken to determine their effect on the condition of the
recovered parts and components. Following cleaning, the parts were marked with the date
and method of recovery (by diver or dredge). Where possible, parts were identified, part
numbers noted and components were laid out in groups by sub-assembly, for example,
wings, landing gears, etc. Empennage parts were laid out on the hangar floor in their
approximate respective positions. Reconstruction was limited due to the degree of
destruction of the aircraft.
The recovered structure and components were put in boxes and crates (to protect them
from further damage), and labeled for shipping and storage. The small miscellaneous
pieces which remained after the sorting were also placed in boxes and labeled for storage.
The total amount of wreckage recovered was about 73 % of the aircraft empty weight
(Appendix D).
After completion of the recovery operation, on 26 January 1998 the boxes of wreckage
were transported overland from Palembang to a hangar in Curug, near Jakarta, for storage
and partial empennage reconstruction.
1.12.1 Aircraft Structures
The aircraft structure debris was examined for evidence of an in-flight fire or explosion.
No such evidence was observed, [Reference 5].
8
1.12.1.1 Wings
A large portion of the wing structure, including parts of the flight control surfaces, were
recovered from the crash site and were severely fragmented. The largest piece was a wing
panel approximately two metres long. Portions of logo light housings on both wing tips
and the wing root fittings were identified. Examination showed no evidence of preexisting corrosion or fatigue on the parts recovered.
All actuators, except one of the Krueger flap actuators, and support fittings of the leading
edge slat were recovered. The examination of the actuators indicates that the leading edge
slats and Krueger flaps were in the retracted position at impact.
The actuators, fittings, and support structure for all of the flight spoilers and all but one of
the ground spoilers were accounted for. Positions of the spoiler actuators and rods
indicate that the surfaces were retracted.
The two aileron power control units (PCUs), aileron tab rods and several pieces of aileron
hinge fittings were identified.
All of the trailing edge flap transmissions and ball screws were also identified. The
position of the ball screw trunnions indicates that the flaps were retracted.
Large fragments of the left hand and right hand landing gear beams, attachment fittings to
the wing rear spar and the landing gear actuators were recovered. The positions of the
landing gear actuators indicate that the landing gears were all in retracted position at
impact.
1.12.1.2 Fuselage
All the fuselage structure that was recovered was retrieved from the crash site in the river.
The structure of the fuselage was severely fragmented and was difficult to identify. The
pieces that easily identified were those pieces containing the SilkAir logo and lettering,
and those pieces from areas around the door frames. Various other fuselage parts that
were identified included parts of the doors and door operating mechanisms, pieces of
floor beams, seat tracks, floor and interior panels. It was very difficult to identify the
positions of these parts within the aircraft because of the severe fragmentation of the
structure.
Not all of the passenger oxygen generators were recovered. Examination of the recovered
oxygen generators revealed no evidence of activation.
A maintenance document showed that a fuselage patch repair was carried out on the
aircraft on 28 April 1997. The repair was carried out by SIA Engineering Company in
accordance with the Boeing 737 Structural Repair Manual 53-00-01, [References 6 and
7].
The patch repair was just above the floor level and forward of the left hand aft passenger
door between station 867 and station 907 and between stringer 14L and stringer 17L. The
entire patch was found intact and was still attached to the surrounding skin, see Figure 9.
9
Most of the aft pressure bulkhead ring chord (about 80 per cent) was identified. Fifty of
the 53 main stringer fitting locations were accounted for. About 15-20 per cent of the
bulkhead web was identified. It should be noted that the web pieces are thin (about 0.8
mm) and they could have been swept away by the river current.
1.12.1.3 Empennage
The purpose of reconstructing and examining portions of the empennage was to attempt
to understand the mode of failure and the structural integrity of the flight controls at the
empennage area.
A wooden structure was constructed to mount parts of the horizontal stabilizer and
elevators in their respective positions. The reconstruction facilitated documentation of the
wreckage and the deformation of the pieces that had separated in-flight. The
reconstruction work was done in a hangar in Curug near Jakarta, in March 1998. The
reconstructed empennage is shown in Figure 11.
The reconstructed tail plane was examined for evidence of an in-flight fire or explosion.
No such evidence was found or observed. However, fractures that were examined
exhibited overload characteristics. It is not known how many of the reversed loading
cycles were experienced by the structure prior to failure and separation of the outboard
stabilizer sections. The damage and deformations of the major fracture surfaces were
consistent with a rapid weakening structure from a number of excessive reverse loadings,
[Reference 5].
Two major pieces of the right hand horizontal stabilizer were found on land. These two
pieces were sections from the tip to approximately stabilizer station 249 and from station
249 to 166 (as measured along the chord-wise breaks). A portion of the leading edge
section of the horizontal stabilizer was observed to have some fasteners missing. Two nut
plates which had missing fasteners were removed and examined at the Bandung Institute
of Technology, Bandung, Indonesia. The test results showed that the fasteners were
attached at impact, [Reference 8].
The left hand horizontal stabilizer piece found on land was from the tip to stabilizer
station 221 (as measured along the inter-spar break; the rear spar extended inboard to the
next rib at station 212).
Several pieces of the elevators (right and left hand side) and control tabs were also found
on land.
1.12.1.4 Horizontal Tailplane
Left hand horizontal stabilizer (Figure 10.b) - No impact marks were found or observed
on the leading edge of this section. Both the front spar (upper and lower chords) and the
rear spar upper chord were bent up near their fracture locations. The lower chord fracture
was indicative of tensile failure but little deformation was observed. The rear spar web
was crushed upward near the fracture location. The upper skin near the fracture was bent
up. The upper skin of the leading edge and the interspar areas exhibited wrinkles. The
rear spar web of the outboard left hand horizontal stabilizer section contained diagonal
tension wrinkles, distortions and paint failures at several locations similar to the right
hand horizontal stabilizer outboard sections. As on the right hand horizontal stabilizer, the
10
damage observed on the outboard left hand horizontal stabilizer section was consistent
with high vertical and torsional loading. It is noted that the right hand horizontal stabilizer
was severely deformed in the same approximate location as the left-hand horizontal
stabilizer chord-wise break near station 221. The stabilizer front spar outboard of station
221 is a solid web, but inboard of this location, the front spar contains flanged lightening
holes.
The upper and lower trailing edge beams had generally failed at the rib locations or along
the attachment bolts to the ribs. Several pieces of trailing edge panels were found on land.
Left Hand Elevator (Figure 10.c) - The largest piece of the elevator was the portion from
the tip to elevator station 195, including the tip balance weight. The stabilizer outboard
trailing edge elevator hinge fitting (elevator station 213) remained attached to the
outboard elevator section. Deformation of fasteners at this hinge location was indicative
of the elevator and hinge moving aft and inboard relative to the left-hand horizontal
stabilizer. All of the hinge fittings were identified except for the horizontal stabilizer rib
hinge fitting located at elevator station 66. The elevator hinges and surrounding area at
elevator stations 213 and 176 contained evidence of elevator over-travel in the up and
down directions and twisting of the bearing hinge plates. The balance weights in the
balance bay and support structure were recovered from the crash site. The elevator tab
was recovered in three long pieces on land, except for the inboard hinge and tab mast
fitting which were recovered from the crash site. The tab hinge and elevator hinge fitting
at elevator station 66 were damaged resulting from twisting of the bearing and fittings.
Both control rods for the left hand elevator displayed multiple helical breaks indicative of
torsional loading. The torsional breaks were as if the trim tab had rotated clockwise
(looking forward) relative to the stabilizer and elevator. All of the attachments for the tab
control rods were accounted for. The tab lockout mechanisms and rods and the tab mast
fittings were recovered from the crash site. The elevator PCUs for the right hand and left
hand elevators were recovered. One of the units was found complete with the piston and
rod, and the rod showed an elevator position near neutral. The piston and rod on the other
unit were missing.
Right hand horizontal stabilizer (Figure 10.d) - The leading edge of the two right hand
horizontal stabilizer sections showed no impact marks. Distortion and wrinkles were
noted in the outboard areas of the horizontal stabilizer leading edge and interspar ribs and
skin.
Right hand elevator (Figure 10.e) - All of the control tab pieces were identified except for
the portion inboard of elevator station 66. The largest piece of the right hand elevator was
the portion from the tip to elevator station 176, including the tip balance weight and
elevator station 213 hinge fitting. All six elevator hinge fittings were identified. The
hinges located at elevator station 121 and outboard were found on land. The elevator
hinges and surrounding area at elevator stations 213 and 176 contained evidence of
elevator over-travel in the up and down directions and twisting of the bearings or hinge
plates. The bearing and the bearing plates of the elevator hinge at elevator station 195
were twisted as if the elevator outboard end moved upward relative to the horizontal
stabilizer. Although the lower bolt was not found, evidence indicated that the bolt was
present during the separation of the elevator. The balance weights in the balance bay and
support structure were recovered from the crash site. The right hand elevator control tab
rods were identified. Deformation of the rods at the right hand elevator showed that they
11
were bent in the downward direction. All of the attachments for the tab control rods and
lockout mechanisms were accounted for. The tab lockout mechanisms and rods and the
tab mast fittings were recovered from the crash site.
The horizontal stabilizer center section (the so-called "Texas Star") structure and the
horizontal stabilizer jackscrew were recovered from the crash site in the river. The
majority of the front and rear spars were found and identified. The horizontal stabilizer
center section has two hinge supports at the rear end about which the horizontal stabilizer
is pivoted. The front end of the Texas Star structure is connected to a ball-nut on the
jackscrew which is used to move the horizontal stabilizer (Figure 10.f).
The majority of the front and rear spars were identified. All of the stabilizer terminal
fitting installations were accounted for. Only the left hand stabilizer hinge fitting and
Section 48-support beam were recovered.
1.12.1.5 Vertical Tailplane
The recovered parts of the vertical tailplane main box and hinge fittings were severely
fragmented. All spar terminal fittings and the trailing edge rib rudder hinge fittings were
accounted for.
An estimated 50 per cent (by area) of the rudder skin honeycomb pieces was found
scattered on land. Small pieces of the rudder spar were retrieved from the river, as were
parts of the mid-spar balance weight. The rudder PCU, the lower hinge and actuator
fittings, as one unit, were recovered from the river and found to be severely damaged.
Vertical stabilizer (Figure 10.g) - All the vertical stabilizer spar terminal fittings and the
fin trailing edge hinge fittings were accounted for at the crash site, indicating that the
vertical stabilizer structure was intact at impact. Damage to the vertical stabilizer
precluded a full accounting of the front spar and interspar structure. The rudder upper
hinge showed evidence of over-travels in the left and right directions. Other hinges
exhibited twisting damage consistent with excessive rudder movement in the clockwise
direction (looking forward). However, it could not be determined when the twisting
damage was incurred.
Rudder (Figure 10.h) - The rudder hinge fittings were complete except the ones at rudder
station 194, which was found with the aft portion of the attachment missing. The rudder
auxiliary actuator fitting was not recovered. Fifty percent (by area) of the rudder
honeycomb skin panels were found on land.
1.12.2 Power plants
The aircraft was delivered on 14 February 1997 with two new CFM 56-3B-2 engines
(Serial numbers 858-480 and 858-481) manufactured by CFM International. There were
no reported problems with engine number one 90 days prior to the accident. According to
the aircraft technical log serial number D 20608 there was only one maintenance action
reported on the number two engine within a 90-day period preceding the accident. On 2
October 1997 it was reported that the ground idle fan and core speeds (N1 and N2) were
higher than those of the number one engine. The number two engine ground idle speeds
were adjusted to match those of the number one engine.
12
The engine hardware was recovered from the river. Both engines were severely damaged
by the impact. It was estimated that about 85 percent by weight of both engines was
recovered. All the major rotating components from both engines were accounted for. The
recovered engine hardware include the fan disks and blades, compressor blades and
vanes, turbine blades and vanes, variable stator vane (VSV) actuators, and main engine
control (MEC) units.
1.12.2.1 Engine
The engines were severely fragmented. The inspections indicated that the operating
conditions of the engines at the point of impact were as follows:
•
The three recovered Variable Stator Vane (VSV) actuators have impact witness marks
on the actuator rods indicating that both engines were running at high engine speed
setting at impact.
•
The high pressure compressor case fragments showed that VSV platforms were
locked in an open (high speed) position - consistent with the impact witness marks on
the actuator rods.
•
All major rotating engine hardware, i.e. fan, compressor, high pressure and lowpressure turbine disks were recovered with no indication of disk failure.
•
Disassembly and inspection of the two recovered MEC systems at the Woodward
Governor Company, Rockton, Illinois, showed that the remaining servos and valves
were in a high flow, high engine speed position.
•
The recovered high-pressure compressor blades were bent in a direction opposite to
the engine rotation, indicating that the engines were operating at the time of impact.
•
The recovered engine casings were very fragmented and did not show any evidence of
high energy uncontained rotating engine hardware.
•
No evidence was found of pre or post impact engine fire. (There was no melted or
soot-covered engine hardware recovered.)
1.12.2.2 Main Engine Control (MEC) / Governor
The disassembly and inspection at the Woodward Governor Company of the two
recovered MEC systems indicated that:
•
The recovered servos and valves were in a high flow, high engine speed position.
•
The fuel valve rotor was found in an open position suggesting a high fuel flow.
•
The governor servo piston appeared to be within approximately 0.25 mm of the
maximum fuel flow stop screw.
•
The tang from the coupler was found in a position of rotation suggesting that the
power lever was at a high angle (i.e. near the full forward stop).
•
The pump unloading and shutdown rotor was in a position of rotation indicating a
run/open position.
13
1.12.2.3 Throttle Box
The throttle box was found with the engine thrust levers in forward thrust (i.e. high
power) and the fuel shut-off levers were in the “run” position (see Section 1.12.3.3.j). It
could not be determined to which of the two engines the throttle box belonged. These
thrust lever positions are consistent with information derived from evidence found in
engine parts such as the severe rotational damage to the rotating parts of the engine.
A portion of the throttle box that was also included in the MEC had a portion of the MEC
throttle housing, power lever and shutdown lever still attached to it. The power lever was
found in the open position. It was measured that the power lever was approximately one
third of an inch (about 8 mm) away from the stop. This would be equivalent to
approximately 12 degrees from the full forward stop. However, due to the severe
deformation of this hardware, the accuracy can not be ascertained. It could not be
conclusively determined to which of the two MECs this hardware belonged.
These above findings are consistent with engine operation under power at the time of
impact. No evidence was found of pre-impact engine fire or uncontained engine failure,
[Reference 9].
1.12.3 PCU and Actuator Tear Down and Examination
Most of the actuators and power control units (PCU) were damaged at impact. Forty-nine
of the actuators recovered from the crash site were sent to Boeing Equipment Quality
Analysis (EQA) Laboratories in Seattle for tear down and examination from 28 April
1998 to 8 May 1998. The tear down and examinations were performed under the control
of representatives from AAIC (now NTSC), NTSB, FAA and Singapore MCIT.
1.12.3.1 Tear Down Activities
The activities of tear down and inspection of the actuators/servos are described in
[Reference 10].
Tear down activities were carried out as follows:
•
The shipment boxes containing the actuators were in good condition with intact seals
on them.
•
Positions of the hydraulic actuators/pistons and other parts were recorded using X-ray
i.e. 200 kV real time X-ray, 300 kV and 425 kV digital radiography and Industrial
Computed Tomography (ICT) modes.
•
Measurements of the position of relevant mechanism, sleeves etc., were made directly
on the parts or from X-ray films (using a scale factor for each film).
•
Due to deformation, the tear down had to be done by some machining operation. The
cutting operations were performed in such a way to minimize damage.
•
The tear down activities were recorded in the form of written reports, video tapes and
still pictures, [Reference 10].
•
Samples of hydraulic fluids and deposits found in the actuators and PCUs were taken
and analyzed in the Boeing Analytical Laboratory using Fourier Transform Infrared
14
(FTIR) spectroscopy, X-ray diffraction (XRD) and moisture content determination.
Analysis and characterization were also performed on the water and mud samples
from the river bottom taken on the 24 March 1998. The results of the analysis are
attached in [Reference 11].
1.12.3.2 Spoiler/Flap/Slat/Thrust Reverser Actuators
Flight spoiler actuators – four units per aircraft: (Items 7, 8, 9 & 10 in Appendix E) All
four flight spoiler actuators were recovered and examined. The actuators were found to be
in the fully retracted position (see Figure 18).
Outboard ground spoiler actuators – four units per aircraft: (Items 15, 16 & 17 in
Appendix E) Three out of the four actuators were recovered and examined. The actuators
were found to be in the retracted and locked.
Inboard ground spoiler actuators - four units per aircraft: (Items 18, 19, 20 & 21 in
Appendix E) All four actuators were recovered and examined. The actuators were found
to be in the fully extended and locked position.
Trailing edge flap ballscrews - eight units per aircraft: (Items 24, 25 & 26 in Appendix E)
Three out of the eight ballscrews were found and examined. The ballscrews were found
fully retracted. It is confirmed by the fact that the actuators were interconnected by a
single drive torque tube.
Leading edge flap actuators - four units per aircraft: (Items 40, 41 & 42 in Appendix E)
Three out of the four actuators were recovered and examined. Two actuators were found
to be in the retracted position and one was slightly extended (1.84 inches) with score
marks in the bottom of the bore near the retract position.
Leading edge slat actuators – six units per aircraft: (Items 43, 44, 45, 46, 47 & 48 in
Appendix E) All six actuators were recovered and examined. Five actuators were found to
be fully retracted and one was found to be slightly extended (0.5 inch).
Locking thrust reverser actuators - two units per engine (four units per aircraft): (Items
27, 28 & 29 in Appendix E) –Three out of the four actuators were recovered and
examined. Two actuators were found to be in the stowed and locked position, while the
third actuator was found in the stowed but unlocked position.
Non-locking thrust reverser actuators – four units per engine (eight units per aircraft):
(Items 32, 33, 34, 35, 36, 37 & 38 in Appendix E) (center and lower) Seven of the
actuators were recovered and examined. The actuators were found to be in the stowed
position.
Thrust reverser synchronization locks - two units per engine (four units per aircraft):
(Items 30 & 31 in Appendix E) - Two out of the four locks were recovered and examined.
The locks were found in locked positions.
1.12.3.3 Actuators Found in Non-neutral Position
During the tear down examination, the following components were found to be not in the
neutral position:
a) Main rudder power control unit (Item 1 in Appendix E) - The main rudder actuator
piston was found bent at the forward and aft ends. X-rays were taken to determine the
15
internal state of the PCU. The rudder PCU was found in a position equivalent to 3
degree left rudder deflection. The measurement was made directly from the X-ray
film.
In view of the condition of the actuator piston and the fact that the actuator is of the
piston type, its position could have been changed on impact. Its position may not be
indicative of the last position at impact.
b) The servo valve of the main rudder PCU (Item 1.a. in Appendix E) was X-rayed using
computerized tomography prior to removal. The primary slide of the main rudder
PCU servo valve was found to be in the half-rate position while the secondary slide
was in the neutral position. The primary slide was able to move freely in the
secondary slide. The secondary slide had to be removed with a force of 260 pounds.
This could have been due to deformation of the servo valve housing during impact.
No corrosion and scratch marks were observed on the primary and secondary slides.
A mapping by means of a mini Boroscope was carried out on the internal surface of
the secondary slide and the servo valve housing and this was recorded on video. No
corrosion and scratch marks were observed on the internal surfaces of the secondary
slide and the servo valve, [Reference 10].
c) Inspection showed that the yaw damper or modulating piston inside the main rudder
PCU (Item 1 in Appendix E) was in a position equivalent to a half degree of left
rudder deflection.
d) The rudder trim actuator (Item 14 in Appendix E) was found to be close to neutral.
e) The cam of the feel and centering unit (Item 22 in Appendix E) had impact marks
indicating a left rudder deflection, (Figure 12).
f) Standby rudder power control unit (Item 2 in Appendix E)- The standby rudder PCU
was found in a position equivalent to one degree right rudder deflection. The
measurement was taken directly from the X-ray film. As the actuator is of the piston
type, the position may not be indicative of the last position at impact.
g) Aileron power control units and autopilot servos (Item 3 and 4 in Appendix E)– There
are two ailerons PCUs, which are mechanically linked to a control quadrant. The
piston of aileron PCU (3529) was found in a position equivalent to 3.5 degrees left
aileron down, while the piston of aileron PCU (3509) was found in a position
equivalent to 1.2 degrees left aileron up. There are two aileron autopilot servos (Item
11.a. & 11.b. in Appendix E). Both were found to be in the disengaged positions.
Disengagement of detent pistons prevents modulating piston inputs to the PCUs.
h) Elevator power control units and autopilot servos (Items 5 and 6 in Appendix E)There are two elevators PCUs which are mechanically linked to one output torque
tube connected to the left and right elevators. One PCU piston (3559) was found to be
in an equivalent elevator position of 1.5 degree trailing edge up. The elevator PCU
(3560) was found with the piston missing and its position could not be determined.
There are two elevator autopilot servos (Item 12.a. & 12.b. in Appendix E). Both were
found to be in the disengaged position. Disengagement of detent pistons prevents
modulating pistons inputs to the PCUs.
i) Horizontal stabilizer jackscrew (Item 23 in Appendix E) – The pictures of the broken
jackscrew are shown in Figure 13. The jackscrew ball-nut was found in a horizontal
stabilizer trim position of 2.5 units, equivalent to 0.5 degree horizontal stabilizer
16
leading edge up4. During the tear down examination, the ball-nut could originally only
rotate approximately 1/8 inch around the circumference. Only when an additional
force is applied was it able to rotate an additional 3/8 inch. (Note: During cruise, as
shown by the FDR data, [Reference 2], the horizontal stabilizer was at about 4.5 units,
equivalent to 1.5 degree horizontal stabilizer leading edge down.)
j) Throttle box (Item 39 in Appendix E) - The throttle box had a power lever and a fuel
shutoff lever still attached. The position of these levers suggests that the engine was at
high power forward thrust and that the fuel shutoff lever was in the “run” position.
The analysis of the hydraulic fluid samples taken from the contained cavities showed that
the fluid was in a normal condition. There was no contamination and no moisture as
shown from the FTIR spectra, [Reference 11].
1.12.4 Other Aircraft Components
Only 370 kilograms of electrical wires, connectors and circuit boards of the aircraft were
recovered. The wires were found broken into short lengths. These were sorted according
to the aircraft system wiring, i.e. CVR, FDR, autopilot etc. The examination of the
recovered wires showed no evidence of corrosion, shorting, burning or arcing.
Other parts of the wreckage recovered from the river included the following:
•
Portions of the auxiliary power unit (APU).
•
Parts of the landing gear assemblies (oleo struts, landing gear door actuators, wheels,
brakes, tire pieces, etc).
•
Life rafts, seat frames, seat cushions (which were X-rayed for shrapnel), oxygen
bottles, oxygen masks, medical kits, cabin fittings, partitions, galley equipment and
fixtures, curtains, etc.
•
Other actuators
1.13 Medical and Pathological Information
Medical records of the flight crew showed no significant medical history. Medical
personnel, families and friends reported that both pilots were in good health. Due to the
severity of the impact and resulting fragmentation of all persons on board no autopsies
were able to be conducted. There were only six human remains identifiable.
1.14 Fire
There was no evidence found of pre-impact fire.
4
Relation between reading in units and deflection in degrees is given as follows: x units = x degrees + 3.
According to the sign convention, deflection of horizontal stabilizer leading edge up means negative, and
leading edge down means positive (see Figure 13).
17
1.15 Survival Aspects
This was not a survivable accident.
1.16 Tests and Research
1.16.1 CVR Circuit Breaker Actuation Test
Upon the completion of data readout by NTSB, the CVR was taken to AlliedSignal on 22
January 1998 for further testing. This testing was an attempt to verify if the termination of
the CVR recording was due to loss of power by the pulling of the CVR circuit breaker or
other means. The result was inconclusive. Therefore other tests had to be performed (see
Appendix F).
There were three tests conducted in a B737-300 aircraft to investigate the CVR circuit
breaker actuation sound signature.
The first test
The first test was carried out on-ground by NTSB and Boeing on 5 February 1998. The
reason for this test was to have quiet ambient condition to provide the best opportunity for
detection of circuit breaker actuation sound signature. The result showed that the CVR
cockpit area microphone did record the CVR circuit breaker actuation. Actuation of a
circuit breaker nearby gave a similar result.
The second test
The test (consisting of on-ground and in-flight tests) was conducted on 14 May 1998 and
15 May 1998 by NTSB.
The purpose of the ground test was to obtain an on-plane, on-ground CVR recording of
the CVR circuit breaker actuation, and the purpose of the flight test was to obtain an onplane, in-flight CVR recording of the CVR circuit breaker opening. In both tests the
circuit breaker was actuated manually and through the introduction of faults to the
aircraft’s wiring, i.e. short circuit and overload.
The results of these tests were compared with the accident CVR recording sound
signatures. In the short circuit tests a distinctive 400 Hz tone is recorded on one or more
of the CVR channels. No corresponding signatures could be identified on the accident
recording. The same tests found that the area microphone is able to pick up a distinctive
and identifiable snap sound that the circuit breaker makes when it is violently tripped by a
short circuit. (Note: The CVR continues to run for 250 milliseconds before it runs out of
power from the capacitor. As sounds travel about one foot per millisecond, it would take
only six milliseconds to travel the approximately six feet distance from the circuit breaker
to the area microphone. Hence the CVR was able to record the snap sound of the circuit
breaker.)
The overload tests yielded similar results as the short circuit tests except that there was a
slight time delay for the circuit breaker to trip and the snap sound was quieter but still
identifiable. No corresponding sound signatures could be found in the accident recording.
18
The last set of tests was to examine the sound signatures when the CVR circuit breaker
was manually pulled. The snap sound was identifiable on the ground without engines and
air-conditioning operating. However in the flight tests, the addition of the background
cockpit noise present during normal cruise obscures the sounds associated with the
manual in-flight pulling of the cockpit circuit breaker. No corresponding sound signatures
could be found in the accident recording.
The summaries of the results of the second tests are as follows:
•
During an overload and a short circuit, the sound of the circuit breaker popping was
loud enough to be identified on the CVR‘s area microphone channel, both on the
ground and in-flight.
•
During an overload and a short circuit, the CVR recorded unique and identifiable
sound signature on one or more of the channels, both on the ground and in-flight.
•
During the manual pull test on the ground, the sound of the circuit breaker was loud
enough to be identified on the CVR recording.
•
In cruise flight, normal cockpit background noise obscured the manual circuit breaker
pull sounds. There were no unique electronic identifying sound signature recorded on
the CVR.
The third test
The test was conducted in-flight using a B-737 SilkAir sister aircraft in Singapore on 16
October 1998 and supervised by the Indonesian AAIC (now NTSC), a FAA avionics
inspector (representing NTSB) and Singapore MCIT representatives.
In the third test, several scenarios were performed where the CVR circuit breaker in the
cockpit was manually pulled. The manual pulls were categorized as “soft”, “hard” and
“string” pull. The soft pull was by pulling the circuit breaker with minimum noise. The
hard pull was by pulling the circuit breaker normally. The string pull was by pulling on a
string that was attached to the circuit breaker. This was to simulate a short circuit causing
the circuit breaker to pop out.
All the tests were conducted with an identical AlliedSignal SSCVR 2-hours recorder as
installed in the accident aircraft.
All four channels of the CVR recordings of the above three tests were analyzed using the
same NTSB signal processing software that was used to analyze the accident CVR
recording.
Several tests were done to document the sound that were recorded on the CVR during a
soft, hard and string pull of the CVR circuit breaker. The test closely matched the data
obtained from the second test (NTSB in-flight test above).
1.16.2 Captain Seat Belt Buckle Sound Test
To further understand several “metallic snap/clunk” sounds heard during the last few
seconds of the accident recording, several seatbelt unbuckling actions were performed
19
during the in-flight test on 16 October 1998. The test was done with the PIC seatbelt
buckle only as the PIC had indicated his intention to leave the cockpit. The seatbelt
sounds recorded during this test were then compared with the sounds from the accident
recording.
The accident aircraft was equipped with headsets with “hot” boom microphones, in
addition to the cockpit area microphone. In the MI 185 recording, the “metallic snap”
sounds were picked up on the cockpit area microphone (Channel 4) of the CVR. The
sounds were also picked up by the Captain’s boom microphone (Channel 1) and the First
Officer’s boom microphone (Channel 2) of the CVR recording. The observer’s station
(Channel 3) contained no cockpit audio information probably because no microphone was
attached.
The tests showed that the metallic snap sound associated with the Captain’s right seatbelt
buckle was of sufficient intensity to be recorded in all three channels (1, 2 and 4) of the
CVR. The sound associated with the Captain’s left seatbelt buckle was picked up by the
cockpit area microphone (Channel 4) and the First Officer’s boom microphone (Channel
2) but not by the Captain’s boom microphone (Channel 1).
1.16.3 Voice Recognition of ATC Recording using Audio Spectral Analysis
According to FDR and radar data, MI 185 was still flying at FL350 for 6 minutes and 56
seconds after the CVR stopped recording. When the Jakarta ATC recording was
compared to the CVR recording, it was noted that there was one routine radio
transmission from MI 185 after the stoppage of the CVR. This was also confirmed from
the VHF keying data of the FDR recording.
To confirm who made the last MI 185 transmission, i.e. “SilkAir one eight five roger …”
at 09:10:26, a test was performed by comparing words from the last transmission in the
ATC recording with the same words from the CVR recording from an earlier part of the
flight. The First Officer made most of the air to ground radio transmissions. The flight’s
call sign of “SilkAir one eight five” was often used in these transmission.
The test results indicated that the voice print characteristics of the CVR radio
transmission matched those derived from the ATC transmission. The most notable
similarity is when the F/O said the call sign “SilkAir”.
The comparison showed that the last transmission was carried out by the F/O.
1.16.4 Trajectory Studies
As some empennage parts were found away from the main impact point, trajectory
studies were carried out by BASI [Reference 12] and by NTSB [Reference 13]. The
studies were to determine the altitude at which the parts separated from the aircraft.
Both studies, see Figure 14, were based on the theory that the trajectory of a separated
object was determined by its initial condition at the time of separation, such as mass,
speed, heading, altitude and its aerodynamic characteristics as well as the wind direction.
20
The BASI study showed that the rudder balance weight (which had the highest ballistic
coefficient – heavy with small frontal area) did not separate while the aircraft was
cruising at FL350 but at a lower altitude.
The wreckage distribution and the relative positions of the individual parts to one another
supported the finding from this study that the separation of the remaining components
took place at a lower altitude near or below 12,000 feet.
1.16.5 Flutter Studies
The NTSB-Boeing Report [Reference 14] showed the following:
•
During cruise at 35,000 feet at 0.74 Mach
The applied static loads and aerodynamic flutter margins are well within the aircraft
certified design requirements. The loads at this flight phase did not support the
separation of the empennage structure.
•
During descent from 35,000 feet to 20,000 feet
The applied static loads and aerodynamic flutter margins developed above 20,000 feet
are less than the aircraft design requirements and do not support the separation of the
empennage structure.
•
During descent from 20,000 feet to ground impact
The applied static loads are less than the design ultimate load envelope and do not
support separation of the empennage during the final descent below 20,000 feet.
The estimated descent speed exceeded the 1.2 dive speed Vd analytical flutter
clearance speed at approximately18,000 feet altitude. An onset of an empennage 22
Hz anti-symmetrical flutter mode was calculated to occur later in the descent at
approximately 5,000 feet altitude (570 equivalent airspeed in knots-KEAS). A lower
frequency 12 Hz anti-symmetrical flutter mode was calculated to occur at
approximately 3,000 feet altitude (600 KEAS), see Figure 15.
1.16.6 Flight Simulation Tests
The simulation tests were performed to explore and understand the various combination
of one or more malfunctions of flight controls, aircraft systems and power plants that
could result in an extreme descent flight trajectory as suggested by the radar data points
from the accident.
Two different simulators were used to conduct these simulation tests. The first is a
Boeing “M-Cab” simulator, which is a full motion, multi-purpose, engineering flight
simulator used to simulate various types of Boeing aircraft. In the case of B737-300, the
nonlinear mathematical software has been validated up to Mach 0.87, and extrapolated
using computational data up to Mach 0.99.
The second simulator was a Garuda Indonesia full motion airline training simulator that
replicates the B737-300 aircraft and the software is validated up to the flight operations
envelope.
21
The Boeing B737-300 aircraft has been designed to have a maximum operating cruise
speed Mmo of 0.83 Mach, and maximum dive speed Mmd of 0.89 Mach.
The results of these tests are tabulated in Appendix G.
1.17 Organizational and management information
•
SilkAir (Singapore) Pte Ltd is a subsidiary of Singapore Airlines Ltd (SIA). It
commenced operation in 1989 under the name Tradewinds. In April 1992 Tradewinds
was renamed SilkAir, [Reference 15].
•
At the time of the accident, SilkAir operated about 100 scheduled flights to about 20
business and holiday destinations in South-East Asia. SilkAir’s routes were generally
short haul regional routes, and aircraft and crew usually do not stay overnight outside
Singapore. The company operated six Boeing B737 and two Fokker F70 aircraft. In
mid-1997, SilkAir placed order for 5 new Airbus A320 and 3 new A319 aircraft to
replace the B737s and F70s.
•
SilkAir is controlled by a Board of Directors. The Board sanctions major decisions
such as choice of aircraft, annual budgets and major expenditures. The General
Manager oversees the functional operations of the four divisions within SilkAir. The
four divisions are Flight Operations, Commercial, Finance, and Engineering. The
majority of SilkAir’s senior managers are seconded from SIA5. Decisions such as
flight operations policies, personnel training and disciplinary matters are managed at
divisional level. Each divisional manager is responsible for management within the
division and coordination with other divisional managers. The divisional managers
report to the General Manager, who reports to the SilkAir Board.
•
The Fleet Manager B737 and the Fleet Manager F70 reported directly to the Flight
Operations Manager. There were approximately 60 pilots and 150 cabin crew in the
airline. Pilots were not represented by a labor union.
•
SIA Flight Operations provides advice on issues such as policy, training and
discipline. SilkAir management followed SIA procedures for the selection of pilots.
Before 1997, SilkAir depended on SIA for recruitment, screening, and selection
efforts. Ab initio pilots were interviewed and given assessments such as psychological
(aptitude) test. RSAF and airline pilots who have flying experience were required to
be interviewed only.
•
Managers in Flight Operations administer promotion practices. Criteria for promotion
to be a commander include performance, personal conduct, seniority, potential
command ability and total time of 4,400 flying hours. Criteria for promotion to be a
Line Instructor Pilot (LIP) include a minimum of 12 months in command,
instructional skills and management potential. Criteria for promotion to be an
Instructor Pilot (IP) include 12 months as a LIP and ability to function as a flight
examiner. For expatriate pilots, IP is the highest management pilot position available.
5
This sentence has been modified according to the suggestion in Appendix M.
22
•
Disciplinary processes were decentralized at SilkAir. There were two levels of
disciplinary inquiries. The first was the Divisional Inquiry and it involved only
management personnel from the work unit. The second was the Company Inquiry,
which was conducted if the employee appealed against the outcome of the Divisional
Inquiry. In a Company Inquiry, managers from the other divisions are included to
form the inquiry panel. As SilkAir is a small organization, it is natural that
disciplinary inquiries are rare6. SilkAir reported that pilots had only been involved in
two previous inquiries: a Captain was demoted to First Officer for 2 years following
an improper interaction with a stewardess; an IP was demoted following an incident
involving a rejected takeoff.
1.18 Other Information
1.18.1 Air Traffic Control
The Jakarta Air Traffic Control was responsible for the control of the MI 185 flight from
Jakarta to PARDI, a reporting point at the boundary of Jakarta FIR and Singapore FIR.
The filed flight plan route was via ATS route G579. Actual route flown after departure
was based on radar control. The Air Traffic Control Services provided by Soekarno-Hatta
Clearance Delivery, Ground Control, Tower and Jakarta Approach were normal.
The ATC transcript reveals that at 08:47:23 MI 185 contacted Jakarta Control and
reported climbing to FL350 and requesting direct PARDI. MI 185 was instructed to
standby and to report reaching FL350, which was acknowledged by the aircraft. At
08:53:17 MI 185 reported maintaining FL350 and was cleared to fly direct to PARDI and
to report abeam Palembang. At 09:10:18 Jakarta Control informed MI 185 of its position
just passing abeam Palembang, and instructed MI 185 to contact Singapore on frequency
134.4 MHz at PARDI. MI 185 acknowledged the message. This was the last
communication between Jakarta Control and MI 185. At 09:35:54 Jakarta Control
requested GA 238 to relay a message to MI 185 to contact Singapore ATC on 134.4
MHz. The aircraft was never in contact with Singapore ATC (Appendix C).
1.18.2 Radar Surveillance
1.18.2.1 Radar Facilities.
There were two radar displaying systems in operation at the Jakarta ATC Center with data
and information derived from six remote secondary surveillance radar (SSR) facilities
located at Soekarno-Hatta (Jakarta), Halim Perdana Kusumah (Jakarta), Semarang,
Palembang, Pontianak and Natuna island.
The two systems are:
1. The Thomson CSF AIRCAT 500 as primary system used by the ATC which had
been in operation since 1985 with inputs from 5 (five) SSR (less Natuna Island).
The AIRCAT 500 was capable of replaying displayed data but had no recording
capability. Being in operation for 12 years the general condition and performance
of the AIRCAT 500 was on a decline.
6
This sentence has been modified as suggested by Appendix M.
23
2. The Hughes GUARDIAN System (later known as Raytheon GUARDIAN
System) with inputs from all the above mentioned sites were in operation since 31
July 1997. It was in transition phase to replace the AIRCAT 500 to the
GUARDIAN (Appendix H). The GUARDIAN System is capable of retrieving all
recorded and displayed radar data and information.
1.18.2.2 Radar Data Output
The radar data and information recorded and displayed by the GUARDIAN system were
sent to the United States in January 1998 for examination at Hughes Raytheon ATC
laboratory at Fullerton, California. Data from Soekarno-Hatta (Jakarta) and Palembang
SSR were selected for the examination as these two radars covered the flight path of MI
185 and provided information on date, time, altitude, x/y position, x/y velocity, ground
speeds, range, azimuth and minimum visible altitude
1.18.2.3 Aircraft Flight Path Based on Radar
The radar data and information retrieved from the GUARDIAN system was further
examined and corrected, by Boeing as well as NTSB, focusing particularly on the last
twelve radar returns. Based on this final radar data, the aircraft flight path and its ground
track projection was reconstructed, see [Reference 18] and (Figure 16).
1.18.3 PIC’s Background and Training
1.18.3.1 Professional Background in RSAF
The PIC joined the Republic of Singapore Air Force (RSAF) as a pilot trainee on 14 July
1975. He obtained his ‘wings’ (fully operational) on 25 March 1977. During his RSAF
career, the PIC flew many different types of fighter and training aircraft. He held senior
flying and instructing positions. He reached the rank of Captain in 1980 and was
promoted to Major in 1989. In 1990, the PIC was selected to joint the RSAF’s Black
Knights aerobatic team. In 1991, the PIC applied to leave the RSAF under a voluntary
early release scheme. The PIC met the eligibility requirements for the early release
scheme as he was 35 years old and had at least six years in his immediate preceding rank.
His application was accepted.
The PIC’s reason to leave RSAF and join SilkAir was to keep flying and to spend more
time with his family.
The PIC obtained a US Federal Aviation Administration (FAA) Commercial Pilot
Licence on 19 November 1991 and an Air Transport Pilot Licence on 26 November 1991
in Benton Kansas. He left full-time employment in the RSAF on 29 February 1992. He
had approximately 4,100 hours flying experience at that time. The PIC served as a
squadron pilot in the RSAF on a part-time basis from 1 March 1992 to 30 April 1993. He
subsequently served in the military reserve, as a Major, in a non-flying capacity. In
January 1997, the PIC was promoted to Deputy Divisional Air Liaison Officer in his
reserve unit.
24
1.18.3.2 Professional Background with SilkAir.
The PIC formally joined SilkAir on 1 March 1992. He was initially employed as a Cadet
pilot under a training program for pilots that did not have a Boeing B737 type rating and
had no previous airline experience.
The PIC was assigned to the Airbus A310 fleet and commenced training on 30 May 1994.
He was appointed First Officer on the aircraft on 15 August 1994. When SilkAir
discontinued A310 operations, the PIC was re-qualified on the B737 in March 1995.
The PIC was selected for B737 command training on 22 October 1995. He was officially
informed of his selection on 20 November 1995. He was appointed Captain on 27 January
1996, and confirmed in that position on 27 July 19967.
SilkAir wrote to the PIC on 8 April 19978 to advise him of his selection for LIP. He
completed his training on 9 May 1997. He performed satisfactorily thereafter in this
position. On 3 July 1997, SilkAir wrote to the PIC to advise him of his de-appointment as
LIP. This was subsequently revised to 28 July 1997 following a company inquiry into an
operational incident that occurred on 24 June 1997 (see Appendix I for details)9.
The PIC had no problems with regard to his professional licence medical requirements.
His last licence renewal medical examination was on 2 December 1997.
1.18.3.3 Financial Background Information
The financial background data of the PIC was gathered to determine whether financial
factors could have affected the performance of the PIC.
At the time of the accident, the PIC operated a securities trading account in Singapore.
This account was operated from June 1990 until the time of the accident. During 1990 –
1997 the PIC traded over 10 million shares, with the value and the volume of the trading
fluctuating during this period. The PIC’s accumulated total losses from share trading
increased between 1993 and 1997, with moderate gains during 1997. There was no period
of the PIC negative net worth. The PIC’s trading activities was stopped on two occasions
due to the non-settlement of his debt, i.e. from 9 April to 15 August 1997 and again from
9 December 1997 until the time of the accident. On the morning of 19 December 1997,
the PIC promised the remisier to make a payment when he returned from his flight10.
The PIC had several loans and debts at the time of the accident. The PIC’s (and
immediate family’s) monthly income was calculated to be less (about 6%) than their
monthly expenditure at the time of the accident.
The probate document indicates that the PIC had a number of insurance policies which
provided benefits on the event of his death. Most of these policies were taken out many
years prior to the accident. In December 1997 he was required by the financial institution
7
This paragraph has been amended according to suggestion in Appendix M.
The LIP position was seen as a requirement for further promotion to instructor pilot or into management.
The position also gave a pilot additional allowance of S$ 750 monthly.
9
This paragraph has been amended according to suggestion in Appendix M.
10
Content of this paragraph has been modified according to comments in Appendix M.
8
25
granting the property loan to take a mortgage insurance policy. The PIC underwent
medical tests for the policy on 1 December 1997 and followed this with a formal
application of 5 December 1997. The PIC did not specify the commencement date for the
policy. On 12 December 1997 the insurance company informed the PIC that his
application was accepted pending payment of the insurance premium. A cheque dated 16
December 1997 was sent to the insurance company by the PIC being payment for the
premium. The commencement or the inception date of the policy was set by the insurance
company to be 19 December 1997. This information was not conveyed to the PIC. The
cheque was cleared on 22 December 1997.
An independent review of the NTSC’s findings concerning the financial background of
the PIC was undertaken by PricewaterhouseCoopers. Based on the review,
PricewaterhouseCoopers made certain recommendations to the NTSC in order for the
NTSC to refine its findings. (Note: PricewaterhouseCoopers was not a party to the
investigation, but was engaged by the NTSC for this specific task).
1.18.3.4 Recent Behaviour
The PIC’s family reported that events and activities were normal in the days before the
accident. The PIC was reported to have slept and eaten normally. There were no reported
changes in his recent behaviour. He was organizing his father’s birthday party that was
planned for 21 December 1997. No medical problems were reported or noted by airline’s
appointed medical clinics.
Work associates who observed the PIC on the day of the accident and on his most recent
flights, reported nothing odd or unusual in his behaviour.
1.18.4 F/O’s Background and Training
The F/O studied aviation at Massey University, New Zealand, between 1992 and 1993
and earned a Diploma of Aviation. As a part of this course, he acquired a Commercial
Pilot Licence. He accumulated about 185 flying hours on light aircraft during this period.
The F/O was selected for an internship programme with Garuda Indonesia airline. He
then completed a B737 type rating in Melbourne, Australia. Based in Jakarta, he worked
as an F/O for Garuda Indonesia between April 1994 and April 1996. He completed 1,309
flying hours on the B737 aircraft during this period. There were no reported operational
events or problems in the F/O’s career at Garuda Indonesia. The F/O left the airline at the
end of his internship.
1.18.4.1 Professional Background with SilkAir
The F/O joined SilkAir on 19 September 1996. He attended B737 re-qualification training
from 22 to 26 October 1996. During the base check on 26 October 1996, his performance
was rated “above average” and he was cleared for line training, which he completed on 13
November 1996. No problems were noted during his training.
The F/O underwent two further base checks on 3 March and 15 September 1997. On both
occasions his performance was rated as “above average”. He completed a line check on
10 October 1997, and no problems were noted.
26
During his career in SilkAir, the F/O was not reported to be involved in any operational
events. During the investigation, other SilkAir pilots described him as an above average
pilot with very good handling skills. It was reported that command trainees sought after
the F/O to be a support pilot during training because of his skills and good situation
awareness. He was described as someone who was professional, followed procedures, and
was willing to learn.
1.18.4.2 Financial Background Information
The F/O did not maintain any credit accounts in Singapore. At the time of the accident,
the F/O maintained his savings in a savings account. His parents reported that he owed
them an amount for initial flight training expenses which was less than the savings at that
time. There were no specified repayment terms. He was reported to be saving money to
further his flight training to qualify for an ATPL. There were no reports of any other
loans or debts. The F/O had a life insurance policy bought in 1992 and a standard SilkAir
policy11.
1.18.4.3 Recent Behaviour
The F/O last flew on 16 December 1997. The F/O’s associates and friends reported no
changes or anything unusual in the F/O’s behaviour in the weeks prior to the accident. He
was reported to have eaten and slept normally in the days prior to the accident.
On the morning of 19 December 1997, two engineers who knew the F/O socially talked
with him when the aircraft was being prepared for departure from Singapore. They made
plans to meet that night. They reported that the F/O appeared to be normal and in good
spirits.
The F/O planned to return to New Zealand in February 1998 to acquire additional pilotin-command hours and obtain his ATPL.
1.18.5 Relationship Between the PIC and the F/O
From December 1996 through December 1997, the PIC and the F/O flew 18 sectors
together across seven days (including the accident flight). The investigation found no
evidence to indicate that there had ever been any difficulties in the relationship between
the PIC and the F/O.
On the two hours of CVR information, there were no indications of any difficulties in the
relationship between the two pilots. There were several cordial discussions not related to
their work tasks at the time.
11
Amended based on information in Appendix M.
27
2
2.1
ANALYSIS
Introduction
The investigation was conducted in accordance with the standards and recommended
practices of Annex 13 to the Convention on International Civil Aviation. In accordance
with Annex 13, the sole objective of the investigation of an accident or incident shall be
the prevention of accidents and incidents. It is not the purpose of this activity to apportion
blame or liability.
This was an extremely difficult and challenging investigation due to the degree of
destruction of the wreckage, the difficulties presented by the accident site and the lack of
information from the flight recorders during the final moments of the accident sequence.
The analysis is drawn from facts compiled in the engineering and systems, operations,
and human factor aspects of the investigations.
Analysis on the engineering and systems aspect starts with the aircraft structural and
systems integrity including trajectory and flutter analysis studies. The results of tear down
and examination of the actuators are analyzed to determine the positions and conditions
of the control surfaces, engines, and other related systems. Special attention is also given
to the issues arising from previous B737 accidents, e.g. the Colorado Springs and
Pittsburgh accidents, especially on the issue of the rudder power control unit (PCU). The
stoppage of the flight recorders is analyzed based on the electrical wiring diagrams, as
well as on-ground and in-flight tests performed in the USA and in Singapore. The
maintenance aspect is analyzed in the last part of the engineering and systems section.
The investigation of the operations aspect covers general factors of the operation of the
flight such as flight crew, air traffic control, weather, navigation, communication, flight
trajectory and conditions based on CVR, FDR as well as radar data, and flight simulation
exercises. The analysis focuses on the possible system failures and its effect to the aircraft
descent trajectory as suggested from radar data points. The descent technique that arises
from emergency situation and recovery from such extreme descent is also discussed.
The investigation of the human factor aspect of the accident took into consideration all
available background data of the flight crew, and the results of the analysis in the other
areas (e.g. engineering and systems, operations, etc). The analysis covers general human
performance issues (such as medical, professional qualification, training, fatigue,
impairment, improper in-flight management, etc), the human factor aspects of the CVR
and ATC recordings, as well as specific human factor issues.
In accordance with Annex 13, a report was made to the relevant aviation security
authorities in late 1999. While the technical investigation continued, aviation security
authorities conducted a separate investigation, which is not covered in this report.
2.2
Aircraft Structural and Systems Integrity
Most of the wreckage was found and dredged from an area of about 60 m x 80 m in the
river bed. The distribution of the wreckage found in the river was documented by a sonar
28
picture and during dredging. The relatively small impact area indicated that the aircraft
impacted at a relatively steep angle.
Although approximately 73% of the aircraft by weight was recovered, only a small
amount of the wreckage were identifiable and in a condition useful for the investigation.
The inspection of the recovered wreckage suggests that the aircraft was structurally intact
until the outer sections of the horizontal stabilizers, elevators and sections of the rudder
separated from the aircraft.
The examination of the fracture surfaces and wreckage fragments showed no evidence of
pre-existing structural defects, fire or explosion in flight.
The examination of the recovered passenger oxygen generators revealed no evidence of
activation from which it is concluded that the aircraft did not experience depressurization
in flight.
One of the investigations is to determine the cause of the separation of parts of the
empennage, as described in the following paragraphs.
2.2.1
Horizontal Stabilizers and Elevators
The recovered sections of the leading edge of the horizontal stabilizer found on land
showed no surface impact marks. This suggests that the separation of the horizontal
stabilizer sections was not initiated by a foreign object impact.
The damage on the horizontal stabilizer box structure showed characteristics of the
occurrence of high vertical and torsional reverse loading. In addition, examination of the
multiple small sections of front spar, leading edge, rear spar lower chord, trailing edge
rib, and interspar ribs and skin (which were found on land) showed characteristics
consistent with numerous cycles of reverse loading.
The parts of the horizontal stabilizer center section (“Texas Star”) were recovered in the
river and were found broken due to impact, see Figure 11. In the recovered parts no
evidence was found which could be attributed to fatigue failure.
The elevator hinges and surrounding area at elevator stations 213 and 176 contained
evidence of elevator over-travel in the up and down directions and twisting of the
bearings or hinge plates. Deformation of the right hand elevator control tab rods showed
that they were bent in the downward direction. Both control rods for the left-hand
elevator displayed multiple helical breaks that indicate torsional loading.
2.2.2
Vertical Stabilizer and Rudder
All recovered parts of the vertical stabilizer box were found at the crash site. All of the
vertical stabilizer spar terminal fittings and the fin trailing edge hinge fittings were also
found at the crash site. This indicated that the vertical stabilizer structure was intact on
impact.
29
All of the rudder hinge fittings were complete except for the hinge fitting of the rudder
station 194, which was missing the aft portion of the attachment. The rudder upper hinge
showed evidence of over travels in the left and right directions. Other hinges exhibited
twisting damage consistent with excessive rudder movement in the clockwise direction
(looking forward). However, because the hinge fittings were found at the crash site, it was
difficult to determine whether the twisting damage was incurred during flight or during
impact.
An estimated 50% by area of the rudder skin honeycomb pieces were found scattered on
land. The top balance weight was found also on land. Small pieces of the rudder spar (still
attached to the hinge fittings) were recovered from the crash site, as were parts of the
mid-spar balance weight. This indicated that the rudder disintegrated partially in the air.
2.3
Break Up of the Empennage
Some parts of the empennage that separated in flight were found on land. Taking into
consideration the relative positions of these empennage parts and their location with
respect to the crash site, BASI and NTSB conducted trajectory studies (Section 1.16.4) to
determine the trajectory of the aircraft and to estimate the altitude when the break-up
could have occurred. Furthermore, to explain the reason for the empennage break-up,
NTSB and Boeing conducted the flutter study as mentioned in Section 1.16.5.
2.3.1
Results of Trajectory Studies
The studies were based on the theory that the trajectory of a separated mass will be
determined by its initial conditions at the moment of separation, taking into consideration
speed, heading, and altitude, as well as its aerodynamic characteristics. Corrections to
allow for wind speeds and directions were applied.
The results suggested that a separation had taken place at a low altitude near or below
12,000 ft., while the aircraft was flying at a high Mach number.
2.3.2
Results of Flutter Studies
Flutter studies undertaken suggested that the applied static loads and aerodynamic flutter
margins developed during the cruise and descent above 20,000 feet are below the design
and certification requirements or specifications and thus do not support the separation of
the empennage structure above that altitude.
Below 18,000 ft altitude, the estimated descent speed exceeded the 1.2 Vd and an onset of
an empennage 22 Hz anti-symmetrical flutter mode was calculated to have had occurred
at approximately 5,000 ft altitude and 570 KEAS.
At approximately 3,000 ft and 600 KEAS a lower frequency 12 Hz anti-symmetrical
flutter mode was calculated to have had occurred.
30
2.3.3
Explanation to the Break Up of the Empennage
Close examination of the wreckage (Section 1.12) supports the results of the flutter
analysis (Section 2.3.2) and the trajectory analysis (Section 2.3.1).
The above results suggest that the separation of the empennage parts could have had
occurred at an altitude near or below 12,000 ft, due to an unstable flutter as the aircraft
exceeded 1.2 Vd.
2.4
2.4.1
Power Control Units and Actuators
Main Rudder PCU
As stated in Section 1.12.3.3.a, the position of the main rudder PCU actuator piston was
measured to be in a position equivalent to a 3° left rudder deflection. The main rudder
actuator piston was found bent at the forward and aft ends. It could not be determined
whether its position reflected the last position before impact.
The yaw damper or modulating piston showed an equivalent of 0.5° left rudder
deflection. However, as the modulating piston is a hydraulic piston, the position found
may not necessarily indicate that the piston was at that position before impact.
Examination of the servo valve of the main rudder PCU showed no scratch marks on the
surface of the slides, as well as in the internal bores of the valve. Corrosion deposits were
found in the cavities of the PCU and other actuators examined. The presence of corrosion
deposits could be due to immersion in the river water, or during the period of storage after
recovery from the riverbed.
During the inspection, a force of 260 pounds (about 130 kg) was required to remove the
secondary slide from the servo valve. This gripping force was due to a deformation of the
valve housing - which was constructed from a stack of wafers. The deformation was
caused by the impact forces on the PCU actuator rod and the servo valve body.
It is noted that some issues arose during investigation of the Colorado Spring (in 1991)
and Pittsburgh (1994) accidents involving B737 aircraft. One of the issues was a
possibility of the servo valve jamming due to thermal shock resulting in rudder reversal.
In the controlled laboratory test condition [Reference 16], it was found that problems due
to thermal shock can arise. This can happen if the warm hydraulic fluid (at +77°C) rushes
into a cold-soaked servo valve (at -40°C) causing the slides to expand against the valve
housing. In such a temperature difference, a valve jamming could occur causing the
rudder to move uncommanded or in a direction opposite to the rudder pedal command
(rudder reversal). In real flight, the hydraulic temperature would not reach that high
(+77°C) a level.
The rudder PCU fitted on the aircraft complies with the applicable Airworthiness
Directives (AD) as described in Appendix J. The modifications to the PCU included one
31
that enabled circulating of hydraulic fluid to warm up the valve to prevent the thermal
shock problem.
Another possible uncommanded rudder movement is due to a corrosion caused by waste
water to the yaw damper system components in the electrical and electronic (E&E)
compartment. Such an incident occurred to a B737 in Bournemouth, UK (in 1995). An
inspection of sister B737-300 aircraft of SilkAir was performed in March 1998. This
aircraft was selected because it was fitted with similar internal configuration and forward
galley modifications, as was the accident aircraft. The inspection showed that the E&E
compartment of the sister aircraft was clean and free from any contamination. It could not
be determined if the E&E compartment of accident aircraft was in similar condition.
However, assuming that the accident aircraft was in similar condition, there would not
have been corrosion caused by toilet water leaking into the E&E compartment.
2.4.2
Standby Rudder PCU
The standby rudder PCU was found at a position of 1° right rudder deflection, which was
in conflict with the position of the main rudder PCU, which was at 3° left rudder. The
standby rudder PCU is normally unpressurized. (The standby rudder PCU is activated by
selecting a switch located on the forward overhead panel to the “Standby” position when
there is failure of normal system).
As the standby PCU was a piston type and furthermore it was unpressurized, its position
could change on impact. Its position may not be indicative of the rudder position at
impact.
2.4.3
Aileron PCU
Examination of the Aileron PCU 3529 indicated a 3.5° left aileron down position, while
the Aileron PCU 3509 indicated 1.2° left aileron up position. Both actuators were piston
type and were mechanically linked, the evidence found was contradictory, and therefore
the position of the ailerons at impact could not be determined.
2.4.4
Elevator PCU
The Elevator PCU 3559 was found to be in an elevator position of 1.5° trailing edge up,
while the Elevator PCU 3560 was found with the piston missing and therefore the
elevator position could not be determined. Both PCU actuators were mechanically linked
to an output torque tube connected to the left and right elevators.
As the elevator PCU actuators were piston type, their position could change on impact.
Therefore the position of the elevator at impact could not be determined.
2.4.5
Horizontal Stabilizer Jackscrew
During the tear-down examination, the horizontal stabilizer jackscrew was found in a
position equivalent to a horizontal stabilizer trim position of 2.5 units. It was also
observed that the ball nut could not be moved freely due to impact deformation. The
upper end counter-bore (aft portion) was ovalized by the screw-shaft being forcibly
32
extracted from the ball nut. The ball nut could be rotated 1/8 inch around the
circumference, and only when an additional force is applied was it able to rotate an
additional 3/8 inch, [Reference 10].
FDR data showed that the horizontal stabilizer position was at an average of 4.5 units
while the aircraft was cruising at 35,000 ft. At this trim setting, the horizontal stabilizer
leading edge is in a down position and corresponds with an aircraft nose-up attitude,
[Reference 17].
The last data on the FDR showed the horizontal stabilizer trim was at 4.58 units, while the
horizontal stabilizer jackscrew was found in a position equivalent to a horizontal
stabilizer trim position of 2.5 units. The difference in the trim positions indicated a
change of about 2.0 units (degrees) of the horizontal stabilizer position.
The horizontal stabilizer trim normally can be operated by three ways:
•
by the autopilot trim (e.g. to automatically correct any out-of-trim conditions);
•
by manipulating the main electrical trim switches on the pilots’ control wheels (e.g.
during manual flying); or
•
by cranking the stabilizer trim wheels (e.g. when the main electrical trim fails).
Uncontrolled movement of the horizontal stabilizers could occur (e.g. a stabilizer runaway). If such a run-away occurs, in a normal situation, it can be overcome by moving
the control column in the opposite direction.
To operate the horizontal stabilizer using the main electric trim system, the two switches
on each of the control wheels have to be operated simultaneously, one to power the
electric motor and two to release the brake and connect the clutch of the jackscrew. With
such a switch arrangement, operation of one switch or a malfunction in one circuitry
would not change the jackscrew position.
A malfunction affecting both trim switches on a control wheel could also cause a runaway. It was not possible to ascertain if such an occurrence took place. (But, at least one
Service Bulletin reported a malfunction during an approach maneuver, see Appendix L.)
However, had a run-away occurred due to a malfunction of the main electrical trim
system, it would take about 10 seconds to change from 4.5 to 2.5 units (at a rate of trim
change of 0.2 unit/sec at flaps retracted position). The trim wheel would turn
continuously. The movement of the trim wheels and the sound produced would have been
noticed by the pilots. Both pilots were trained to recognize such a condition and to take
appropriate corrective actions.
The effect of a system run-away of the horizontal stabilizer trim was simulated in the
Garuda Indonesia Training Simulator as well as Boeing M-Cab Simulator, see Appendix
G. A trim change from 4.5 to 2.5 units changed the aircraft attitude from a nose-up to a
nose-down attitude. The simulator results showed that, with such a trim change, it took 1
minute and 23 seconds to descend from 35,000 feet to 19,500 feet. However, the last five
ATC radar points showed a much faster descent of the accident aircraft, i.e. 32 seconds
from 35,000 feet to 19,500 feet. Therefore, if the simulation was correct, the change of
horizontal stabilizer trim position alone would not have resulted in the fast descent after
leaving FL350.
33
During autopilot operations, the trim operates automatically to correct changes in the
center of gravity position due to fuel use or movements of people along the aisle. Had a
run-away occurred during autopilot operations, the autopilot would disengage. The trim
wheel would also turn continuously. The warning from the autopilot disengagement and
the turning of the trim wheels would alert the pilots to this situation. Both pilots were
trained to recognize such a condition and to take appropriate corrective actions.
The manual trim wheel is available as a back-up in case there is a malfunction in the main
electrical trim system. It is possible to change the trim position using the trim wheel, but
the pilot would have to crank the wheel in a prolonged manner to effect a large trim
change. No evidence was found that the main electrical trim system had failed.
During normal operations, the main electrical trim switches on the control wheel are used
to neutralize the force on the control columns. In this mode of operation, pushing the
main electrical trim switches forward continuously would move the horizontal stabilizer
to a limit of 2.5 units when the flaps are retracted. The horizontal stabilizer trim position
was found in a 2.5 units position during the tear down examination. This matches the
forward limit of the main electrical trim.
2.4.6
Other Actuators
During the tear down examination, the following components were found to be in the
stowed or retracted position:
• Flight spoiler actuators
• Outboard ground spoiler actuators
• Inboard ground spoiler actuators
• Trailing edge flap ballscrews
• Leading edge flap actuators
• Leading edge slat actuators
• Mach trim actuator
• Thrust reverser actuators
The fact that these actuators were found in the stowed or retracted position does not
necessarily suggest that their respective systems were not activated during the descent. If
the respective systems remained in the stowed or retracted positions, they would not have
been factors contributing to the accident.
2.5
Powerplant
In the examination of the recovered engine parts, main control units and throttle box, the
following were found:
•
There was no evidence found of melting or soot on recovered engine hardware,
indicating that there was no pre or post impact engine fire.
•
All major rotating engine hardware, i.e. fan, compressor, high pressure and low
pressure turbine disks were recovered with no indications of high energy uncontained
engine failures.
34
•
Indications were found that both engines were operating at high engine rotation speed
at the point of impact.
Therefore, the engines were considered to be not a factor contributing to the accident.
2.6
2.6.1
Stoppage of the CVR and FDR
CVR Stoppage
The CVR recording ended while the aircraft was still cruising at an altitude of 35,000
feet, about seven minutes before the last radar return. Up to the CVR stoppage, the
conversation in the cockpit was consistent with normal flight operations.
The CVR stoppage could have occurred due to a malfunction of the unit itself or a loss of
power to the unit. The loss of power to the unit could be due to power interruption to the
Electronics Bus 1 that supplies power to the CVR, short circuit or overload, CVR circuit
breaker pulling or break in the wiring.
The entire two-hour recording was found normal. There were no observed anomalies
when power was transferred on the ground in Jakarta. It appeared that the recorder’s
internal energy storage capacitor was operating normally by providing continuous
recorder operation in spite of momentary aircraft electrical power interruptions,
[Reference 4].
The examination of the CVR unit performed by the manufacturer (Appendix F)
confirmed that the CVR was functioning properly. The recording had characteristics that
would be expected of a normal electrical power shutdown of the CVR. Therefore, the
stoppage of the CVR could be a result of the loss of power to the unit.
According to the aircraft wiring diagram 24-58-11 (Figure 17) the power to the CVR was
from the Electronics Bus 1 (Elex Bus 1). The Elex Bus 1 also supplies power to other
systems, such as the FDR, DME-1, TCAS, ATC-1 etc. Parameters of DME-1 and TCAS
were recorded in the FDR. Analysis of the FDR recording showed that six minutes after
the CVR stopped, the FDR was still recording TCAS and DME-1 parameters. This
indicates that the CVR stoppage was not due to power loss at Elex Bus 1.
The CVR is equipped with an energy storage capacitor. The function of this capacitor is
to provide power for 250 milliseconds after electrical power is removed from the unit
such as when the aircraft power is switched from ground power to APU generators or the
engine generators. Another function of this capacitor is to enable continued recording for
another 250 milliseconds after power loss to the unit.
Had there been an overload or short circuit, the resultant popping of the CVR circuit
breaker in the cockpit would have been recorded as a unique and identifiable sound
signature by the CVR (see Sections 1.11.3 and 1.16.1). Based on the examination of the
results of the circuit breaker pull tests, there was no such sound signature in the MI 185
CVR recording found. This indicates that there were no short circuit or overload to cause
the CVR circuit breaker to pop out.
35
The results of the CB pull tests showed that the sound signature associated with manual
pulling of the circuit breaker is obscured by the cockpit ambient noise. Hence, no
conclusion can be drawn whether the circuit breaker had been pulled manually.
A break in the wire supplying power to the CVR could also lead to CVR stoppage
without any sound being recorded on the CVR. However, from the limited quantity of
wiring recovered it could also not be determined if a break in the wiring had caused the
CVR to stop.
Thus, the cause of the CVR stoppage could not be concluded.
2.6.2
FDR Stoppage
The FDR stopped recording at 09:11:33.7, or 6 minutes and 18.1 seconds after the CVR
stoppage, and approximately 35.5 seconds before the aircraft started its descent, see
Section 1.11.1 and Figure 2. Data recorded by the FDR indicates that the flight was
normal until the FDR stoppage time. It was concluded that until the stoppage of the FDR,
there were no indications of unusual disturbance (e.g. atmospheric turbulence, clear air
turbulence, or jet stream upsets, etc.) or other events affecting the flight.
The FDR stoppage could have occurred due to a loss of power supply to the FDR, or the
malfunction of the unit itself.
The recording of the ATC radar plots during the descent of the aircraft until 19,500 ft
indicated that the aircraft ATC transponder continued operating after the FDR had
stopped recording. SilkAir stated that generally flight crews use ATC-1 flying outbound
from Singapore, and ATC-2 inbound. ATC-1 is on the same bus as the FDR, while ATC2 is powered from Elex Bus 2, i.e. a different power source. No conclusion could be
drawn as to the reasons for the CVR and FDR stoppage at different times.
The FDR was determined to be functioning normally until it stopped. The stoppage of the
FDR could not be determined from the available data.
There were no evidence found that could explain the six-minute time difference between
stoppage of the CVR and FDR.
2.7
Radio Transmission Voice Recognition
The last radio transmission on VHF from MI 185 was at 09:10:26 when it acknowledged
ATC’s call that MI 185 was abeam Palembang, to maintain FL350, and to contact
Singapore Control at PARDI, by responding “SilkAir one-eight-five roger, 134.4 before
PARDI”. Voice spectrum analysis identified that the F/O made this last radio
transmission, see Section 1.16.3. This information reveals that the F/O was in the cockpit
about 1.5 minutes prior to the descent.
36
2.8
2.8.1
Maintenance Aspects
Aircraft Maintenance
The aircraft and the engines were found to be properly maintained by SIA Engineering
Company (SIAEC) in accordance with the Maintenance Schedule ref. MI/B737-300 Issue
No.1 and approved by the Civil Aviation Authority of Singapore (CAAS). The
maintenance task cards and inspection reports were reviewed. The required checks were
carried out within the stipulated period. There were no adverse findings on any
matter/defect that would contribute to the accident.
A review was carried out of the aircraft technical logbooks, deferred defect logbooks and
in-flight log defects. It did not reveal any defects that could have affected the
airworthiness of the aircraft.
Therefore, the inspection of the maintenance records did not reveal any defects that may
have contributed to the occurrence.
2.8.2
Patch Repair
On 24 April 1997 the aircraft was damaged in a ground incident, resulting in a dent to the
fuselage skin. The patch repair, designed and performed according to Boeing Structure
Repair Manual (SRM), was carried out by SIA Engineering Company on 28 April 1997.
The patch repair was recovered from the riverbed. A damage to the upper forward corner
of the patch repair was concluded to be the result of impact. Examination of the patch
revealed no evidence of material or workmanship abnormalities.
Therefore the possibility that the patch repair was a factor contributing to the accident can
be ruled out.
2.9
General Operational Issues
Both pilots were properly trained, qualified and licensed to conduct the flight in
accordance with the Singapore Air Navigation Order. The pilots received sufficient rest
before the flight and had no medical history that would adversely affect their
performance.
The security checks of passengers, baggage and cargo were conducted in accordance with
standard procedures. The results of the checks were all normal. No dangerous goods were
carried on board.
The aircraft was dispatched and operated within the certified weight and balance limits.
At the time of the accident it was daylight. Weather observations around Palembang
indicated no adverse weather in the vicinity near that time. There were several flights in
the vicinity at different flight levels and the en-route weather experienced by these flights
was isolated in nature. Generally, flights that were affected by the weather would have
requested for deviation before the flight encountered the weather. The Sempati 134 flight
37
from Jakarta to Batam flying at FL310 approximately 2 minutes ahead of MI 185 reported
that due to weather conditions en-route he requested to fly direct PARDI. The Qantas 41
flight at FL410 about 8 minutes behind MI 185 did not report adverse weather over
Palembang except for two or three isolated thunderstorms about ten miles to the east of
track near Palembang.
All navigational aids and facilities of the Jakarta airport, and other en-route navigation
aids between Jakarta and Singapore were fully serviceable. The flight was cleared to route
from a point near Jakarta direct to Palembang, then to PARDI.
Up to 09:10:26 the Jakarta Air Traffic Controller was communicating normally with the
cockpit crew on frequency 132.70 MHz. From then on, no distress call was received from
the crew, no distress signal was received from the aircraft transponder, and no emergency
locator beacon transmissions were detected from the crash site.
2.10 Simulated Descent Profile
The last five ATC radar points recorded represent the flight trajectory of the aircraft from
the cruise altitude 35,000 feet to approximately 19,500 feet. Each point consisted of data
relating to time, altitude and geographical coordinates.
Simulator tests and computer simulation fly-out studies were done to determine failures
or combination of failures of the flight control and autopilot systems that could result in
the extreme descent trajectory. Aircraft flight data were not available for the time period
after the stoppage of the FDR. The initial condition for these tests and studies was cruise
configuration at 35,000 feet based on the last known FDR data. The altitude range for the
simulations was from 35,000 feet to approximately 19,500 feet.
The results of these simulation studies (Appendix G) are summarized as follows:
•
Any single failure of the primary flight controls such as hard-over or jamming of
aileron, rudder or elevator did not result in a descent time history similar to that of the
last ATC radar points. In simulations of these flight control failure conditions the
aircraft could be recovered to normal flight manually.
•
Any single failure of the secondary flight controls such as hard over or jamming of
yaw damper, or runaway of the stabilizer trim would not result in a descent time
history similar to that of the last ATC radar points. In simulations of these flight
control failure conditions the aircraft could be recovered to normal flight manually.
•
Manipulation of the primary flight controls without horizontal stabilizer trim would
result in a descent time history similar to that of the last ATC radar points. But this
required large control column input forces and the aircraft was subjected to a loading
exceeding 2 G. However, if the control column input forces were relaxed, in the
simulations the aircraft would recover from the steep descent due to its inherent
stability.
•
Among other possibilities, a combination of changing the stabilizer trim from about
4.5 to 2.5 units and an aileron input could result in a descent time history similar to
38
that of the last ATC radar points. This simulated descent trajectory would result in the
aircraft entering an accelerating spiral and being subjected to a loading of less than 2
G. Furthermore, the aircraft would continue in the spiral even when the control forces
were relaxed. This would result in a descent at a speed exceeding 1.2 Vd, in
agreement with the analysis on the break up of the empennage as discussed in Section
2.3.
2.11 High Speed Descent Issues
2.11.1 Mach Trim System and its Function
The aircraft was equipped with a Mach Trim system to provide stability at the higher
operating speeds, i.e. higher Mach numbers. Mach trim is automatically accomplished
above Mach 0.615. When the Mach Trim system is operative it will normally compensate
for trim changes by adjusting the elevator with respect to the stabilizer, as the speed
increases. With the Mach Trim inoperative, the aircraft could exhibit a nose down
tendency ("Mach Tuck") as speed increases. However, the expected control forces to
overcome the “Mach Tuck” are light. Additionally when the speed exceeds the maximum
limit, audible overspeed warnings are activated.
Since the aircraft was cruising at subsonic speed (Mach 0.74) and trimmed for level
flight, the aircraft will eventually return to the trimmed condition after a minor speed
disturbance.
For the aircraft to dive, a significant disturbance resulting in an increasing speed must
have taken place. Such a disturbance could be initiated by changing aircraft elevator or
stabilizer trim. Should the airspeed increase to the point where it becomes transonic, and
as the lift resultant moves aft and local supersonic flow develops, the nose-down pitching
moment could be sufficiently large that the aircraft becomes speed unstable, i.e.
continuing speed increase of the aircraft. Once the aircraft is in a transonic dive, a
recovery from the dive becomes more difficult because of an increase in control column
forces, due to the aircraft’s increasing nose down pitching moment, as well as a large
reduction of elevator effectiveness due to the formation of shock induced air flow
separation in front of the elevator.
It is possible to recover from a transonic dive by timely action of the pilot, by reducing
thrust and deploying the speed brakes. Should the pilot not initiate a prompt recovery
action, the recovery becomes more difficult.
During the tear down examination, the mach trim was found in the fully retracted
position. The fact that this actuator was found in the retracted position may not
necessarily indicate that the mach trim system is a factor contributing to the accident.
2.11.2 Emergency Descent due to Fire, Smoke or Depressurization
An emergency descent is necessary when there is a rapid cabin depressurization or when
a fire or smoke occurs in flight. The procedure is to simultaneously retard the thrust
39
levers, deploy the speed brakes and initiate the descent12 (Appendix K). Some forward
stabilizer trim is applied to attain a dive that will accelerate the aircraft towards the
maximum speed limit. Once the maximum speed is reached aircraft is re-trimmed to
maintain the speed. This facilitates a limit on maximum rate of descent to the minimum
safe altitude.
The last pilot radio transmission about two and a half minutes before the descent sounded
normal and there was no mention of any in-flight fire or smoke. Furthermore,
examination of the wreckage showed no evidence of in-flight fire or explosion.
Examination of the recovered oxygen generators showed that they were not activated.
This indicated that there was no rapid depressurization at high altitude.
Based on the above findings, there was no indication of an emergency descent due to fire,
smoke or rapid depressurization.
2.12 General Human Performance Issues
This section analyses the general human performance issues such as medical, professional
qualification, training, fatigue, impairment, improper in-flight management, etc.
•
The flight-crew's medical background and recent activities were examined. All
medical files reviewed showed no significant medical history. Medical personnel,
family, and friends reported both pilots to be in good health. Neither pilot flew in the
two days before the crash. Family and friends reported routine activities and rest
during that time period. Evidence from the CVR and the conduct of the flight (based
on FDR and ATC communications) indicated that neither pilot experienced any
medical difficulties in flight. Further, there was no evidence found of incapacitation or
impairment. Therefore, the investigation concludes that until the stoppage of the flight
recorders there was no evidence found to indicate that the performance of either pilot
was adversely affected by any medical, psychological or physiological condition.
•
Both pilots held valid airman licenses and medical certification. They had received all
required training, including unusual attitude recovery training. At the time of the crash
both pilots were within duty-time limits. Peers, instructors, and supervisors described
both pilots as competent pilots with good airplane handling skills. It was concluded
that both pilots were properly trained, licensed and qualified to conduct the flight.
•
The relationship between the PIC and the F/O was examined. There were no reports
of any conflict or difficulties between the pilots prior to the occurrence and before the
day of the crash. Based on the available recorded data of the CVR, there was no
evidence of any conflict or difficulties between the pilots during the approach and
landing into Jakarta, on the ground at Jakarta, and during the accident flight. The
infrequent non-flight related conversations between the pilots were also cordial. It was
concluded that the investigation did not find any evidence of difficulties in the
relationship between the two pilots either during or before the accident flight.
12
Amended based on suggestion in Appendix N.
40
•
During their career in SilkAir, the PIC and the F/O received training in recovery from
unusual flight attitudes. The PIC was a member of the RSAF “Black Knights”
aerobatic team in the 1970s. The F/O was reported to be sought after as co-pilot by
other SilkAir pilots who were undergoing command upgrade training. This training
typically includes scenarios which require handling of system failures or other
abnormal flight situations. Therefore both pilots were familiar with recovery from
unusual flight attitudes.
2.13 Human Factors Aspects of the CVR and ATC Recordings
2.13.1 CVR
(a) The conversations and sounds recorded by the CVR before it stopped were
examined. The CVR transcript (Appendix A) showed that at 09:04:57 the PIC
indicated his intention to go to the passenger cabin, "go back for a while …. finish
your plate….". At 09:05:02 the PIC offered water to the F/O, and at about the same
time, several metallic snapping sounds were recorded. Thirteen seconds later, at
09:05:15.6 the CVR ceased recording. Analysis of the recording indicated that the
metallic snapping sounds were made by a seatbelt buckle striking the floor. (See
Section 1.16.2)
(b) During the period recorded by the CVR, all door openings or closings were related to
pre-departure activities, in-flight meal service and normal pilot-cabin crew
interaction. In the four minutes following the last meal service, there were no sounds
associated with cockpit door opening or closing. After takeoff from Jakarta,
conversations within the flight deck were between pilot-to-pilot, pilot-to-flight
attendants, and normal pilot-to-ATC radio communications. During the flight, except
for cabin attendants serving meals and drinks to the pilots, there were no indications
of any other person(s) in the cockpit. It is concluded that after the last meal service
and until the stoppage of the CVR, the recording did not reveal any indications that
person(s) other than the flight crew and cabin attendants attending to their duties
were in the cockpit.
(c) Analysis of the CVR stoppage indicated that the failure of the CVR could not have
been caused by a short circuit or overload. This is because either occurrence would
have resulted in the CVR recording a “pop” sound which was heard on the test
recording but not on the accident recording.
The CVR in-flight tests could not identify the sound of the CVR circuit breaker being
manually pulled as the ambient noise obscured the sound made. The accident tape did
not contain any identifiable sound attributable to manual pulling of the CVR circuit
breaker. It was not possible to determine from the CVR tests if there was a pulling
out of the CVR circuit breaker.
2.13.2 ATC Recordings
The data transcribed from the ATC communications recording of the air-to-ground
conversation indicates that at 09:10:26, or 5 minutes and 10.4 seconds after the CVR
stoppage, the F/O acknowledged the “abeam Palembang” call from the ATC. The F/O
41
was positively identified by voice analysis examination. This confirms that the F/O was in
the cockpit when the aircraft was abeam Palembang. However, it is not possible to
conclude whether the PIC was in the cockpit at the time. It was also not possible to
determine events or persons present in the cockpit from the time of the last transmission
to ATC.
The absence of a distress call could suggest that the pilots were preoccupied with the
handling of an urgent situation. However, it is not possible to conclude on the reason for
the absence of a distress call.
2.14 Specific Human Factors Issues
In this section, the specific, personal, financial backgrounds and recent behaviour of the
PIC and the F/O are examined.
2.14.1 Personal Relationships
Evidence obtained from family and friends of both the PIC and F/O reported no recent
changes or difficulties in personal relationships.
It was concluded there was no evidence that either pilot was experiencing difficulties in
any personal relationships.
2.14.2 First Officer (F/O)
The investigation into the F/O's personal and professional history revealed no unusual
issues. No records of incidents or unusual events were found, and no career setbacks or
difficulties were experienced. Financial records showed no evidence of financial
problems. Interviews with family, close friends and relations seem to indicate that the F/O
was a well-balanced and well-adjusted person, and keen on his job, and planning to
advance his a flying career. There were no reports on recent changes in his behaviour.
2.14.3 Pilot-in-Command (PIC)
The investigation into the personal and professional career revealed that the PIC was
considered to have been a good pilot, making his transition from a military pilot to
commercial pilot smoothly. His career at SilkAir showed that he was well accepted and
given higher responsibilities. He was considered to be a leader among the Singaporean
pilot community in SilkAir.
During his professional career at SilkAir, he was involved in a few work-related events,
which were in general considered minor operational incidents by the management.
However in one particular event, for non-technical reasons the PIC infringed a standard
operating procedure, i.e. with the intention to preserve a conversation between the PIC
and his copilot, the PIC pulled out the CVR circuit breaker, but the PIC reset the circuit
breaker in its original position before the flight. This was considered a serious incident by
the management, and the PIC was relieved of his LIP appointment. The PIC was known
to have tried through existing company procedures to reverse the management decision.
Although there were some indications of the PIC being upset by the outcome of the
42
events, the magnitude of the psychological impact on the PIC performance could not be
determined.
The PIC’s financial history was investigated for the period from 1990-1997. Based on the
data available to the NTSC it was noted that the PIC’s accumulated losses in share trading
increased between 1993 and 1997, with moderate gains in 1997. There was no period of
negative net worth. His trading activity was stopped on two occasions due to nonsettlement of his share trading debt.
In the independent review of the NTSC’s findings concerning financial background of the
PIC by PricewaterhouseCoopers, the available information in regard to the assets,
liabilities, income, expenditure, and share trading were incorporated into a net worth
analysis. This analysis reveals the following:
•
•
•
The PIC's net worth (known assets less known liabilities) deteriorated over the period
of 1994 to 1996, however it grew marginally during 1997.
Between 31 December 1994 and 31 December 1996 his known assets declined, while
his known liabilities increased resulting in a net decline in the value of the PIC's
known net worth.
Between 31 December 1996 and 19 December 1997, the PIC’s known net worth
increased again.
Analysis of the net monthly income available for discretionary and general out-of-pocket
expenses, based on a monthly combined gross income (including the PIC and his wife’s
salaries), indicated that the PIC had a relatively minor monthly cash shortfall at around
the time of the accident. This should be considered in light of the PIC’s positive net worth
at the time of the accident.
2.14.3.1 Recent Behaviour
The PIC’s recent behaviour was analysed from statements made by family members,
friends and peers during interviews. The PIC’s family reported no recent changes in his
behaviour. Work associates who observed the PIC on the day of the accident and on his
most recent flights, reported nothing odd or unusual in his behaviour.
2.14.3.2 Insurance
Based on the data available, it was found that at the time of the accident, the PIC had a
number of life insurance policies. The majority of these were taken up earlier in his life.
The most recent policy was a mortgage policy which was required by the financial
institution from which he took the loan for his house in line with normal practice for
property purchases in Singapore. The PIC applied for the mortgage policy on 27
November 1997. The insurance company approved the policy on 12 December 1997
pending payment of the first premium. The PIC submitted a cheque dated 16 December
1997 for the first premium payment. The commencement or the inception date of the
policy was set by the insurance company to be 19 December 1997. This information was
not conveyed to the PIC. The cheque was cleared on 22 December 1997. From the data
available to the NTSC there was no evidence to indicate if this mortgage policy has any
relevance to the accident.
43
2.14.3.3 Overall Comments on the Pilot-in-Command
NTSC concluded that the combination of financial situation and his work related events
could be stressors on the PIC. However, NTSC could not determine the magnitude of
these stressors and its impact on the PIC’s behaviour.
44
3
3.1
CONCLUSIONS
Findings
Engineering and Systems
•
There was no evidence found of in-flight fire or explosion.
•
From flutter analysis and wreckage distribution study, the empennage break-up could
have occurred in the range between 5,000 and 12,000 feet altitude.
•
Examination of engine wreckage indicated that the conditions of the engines at
impact were not inconsistent with high engine rotation speed. No indications were
found of in-flight high energy uncontained engine failures. Therefore, the engines
were considered to be not a factor contributing to the accident.
•
Examination of the actuators of flight and ground spoilers, trailing and leading edge
flaps, as well as engine thrust reversers indicate retracted or stowed positions of the
respective systems.
•
Examination of the main rudder power control unit (including the servo-valve), the
yaw damper modulating piston, the rudder trim actuator, the rudder trim and feel
centering unit, the standby rudder PCU, the aileron PCUs, the elevator PCUs, and the
horizontal stabilizer jack-screw components, revealed no indications or evidence of
pre-impact malfunctions.
•
Examination of the 370 kg of recovered electrical wires, connectors and circuit
boards showed no indication or evidence of corrosion, shorting, burning or arcing in
these wires or parts.
•
The CVR stopped recording at 09:05:15.6 and the FDR stopped recording at
09:11:33.7. The examination of the CVR and FDR showed no malfunction of the
units. The stoppages could be attributed to a loss of power supply to the units.
However, there were no indications or evidence found to conclude on the reason for
the stoppages due to the loss of power. The cause of the CVR and FDR stoppages
and the reason for the time difference between the stoppages could not be concluded.
•
The inspection of the aircraft maintenance records did not reveal any defects or
anomalies that could have affected the airworthiness of the aircraft or that may have
been a factor contributing to the accident.
•
The horizontal stabilizer trim was found to be in the 2.5 units position which matched
the forward limit of the manual electrical trim.
Flight Operations
•
Weather and Air Traffic Control were not factors contributing to the accident.
•
Audio spectral analyses on Air Traffic Control communications and the accident
CVR indicate that the last communication from the MI 185 at 09:10:26, occurring at
a position approximately abeam Palembang was performed by the F/O.
•
The examination of the flight deck noise and sounds concludes that the metallic snap
recorded on the CVR was made by a seatbelt buckle hitting against a metal surface.
45
•
Based on flight simulations, it was observed that the simulated descent trajectory
resulting from any single failure of flight control or autopilot system would not
match the radar data.
•
Based on the same flight simulations, it was also observed that the trajectory shown
by the radar data could have been, among other possibilities, the result of the
combination of lateral and longitudinal inputs together with the horizontal stabilizer
trim input to its forward manual electrical trim limit of 2.5 units.
Human Factors
•
Both pilots were properly trained, licensed, and qualified to conduct the flight.
•
There was no evidence found to indicate that the performance of either pilot was
adversely affected by any medical or physiological condition.
•
Interviews with respective superiors, colleagues, friends and family revealed no
evidence that both the flight crew members had changed their normal behaviour prior
to the accident.
•
There was no evidence found to indicate that there were any difficulties in the
relationship between the two pilots either during or before the accident flight; or had
been experiencing noteworthy difficulties in any personal relationships (family and
friends).
•
Until the stoppage of the CVR, the pilots conducted the flight in a normal manner
and conformed to all requirements and standard operating procedures.
•
Although a flight attendant had been in the cockpit previously, after the last meal
service and until the stoppage of the CVR there was no indication that anyone else
was in the cockpit other than the two pilots.
•
In the final seconds of the CVR recording the PIC voiced his intention to leave the
flight deck, however there were no indications or evidence that he had left.
•
Interviews and records showed that in 1997 the PIC had experienced a number of
flight operations related events, one of which resulted in his being relieved of his LIP
position.
•
The PIC was involved in stock-trading activities, but no conclusions could be made
indicating that these activities had influenced his performance as a pilot.
•
From the data available to the NTSC there was no evidence found to indicate if the
mortgage policy taken out by the PIC in connection with his housing loan has any
relevance to the accident.
3.2
•
Final Remarks
The NTSC investigation into the MI 185 accident was a very extensive, exhaustive
and complex investigation to find out what happened, how it happened, and why it
happened. It was an extremely difficult investigation due to the degree of destruction
of the aircraft resulting in highly fragmented wreckage, the difficulties presented by
the accident site and the lack of information from the flight recorders during the final
moments of the accident sequence.
46
•
The NTSC accident investigation team members and participating organizations have
done the investigation in a thorough manner and to the best of their conscience,
knowledge and professional expertise, taking into consideration all available data and
information recovered and gathered during the investigation.
•
Given the limited data and information from the wreckage and flight recorders, the
NTSC is unable to find the reasons for the departure of the aircraft from its cruising
level of FL350 and the reasons for the stoppage of the flight recorders.
•
The NTSC has to conclude that the technical investigation has yielded no evidence to
explain the cause of the accident.
47
4
RECOMMENDATION
Recommendations to manufacturers
•
It is recommended that the ICAO FLIREC Panel undertake a comprehensive review
and analysis of flight data recorders and cockpit voice recorders systems design
philosophy be undertaken by aircraft and equipment manufacturers. The purpose of
the review and analysis would be to identify and rectify latent factors associated with
stoppage of the recorders in flight, and if needed, to propose improvements to ensure
recording until time of occurrence.
•
It is recommended that, to facilitate the recovery of flight recorders after impact into
water, a review of the flight recorders design philosophy be undertaken by the
equipment manufacturers to ensure that the underwater locator beacons (ULB) are
fitted to the flight recorders in such a manner that the ULB would not be separated
from the recorders in an accident.
•
It is recommended that the ICAO FLIREC Panel recommend aircraft and equipment
manufacturers to include recording of actual displays as observed by pilots in
particular for CRT type of display panels.
•
It is recommended that a review of the flight crew training syllabi be undertaken by
aircraft manufacturers to include recovery from high speed flight upsets beyond the
normal flight envelope. The purpose of developing the additional training is to
enhance pilot awareness on the possibility of unexpected hazardous flight situations.
•
It is recommended that a review of aircraft auto-flight systems be undertaken by
aircraft and equipment manufacturers to provide all passenger aircraft with auto flight
systems that could prevent an aircraft from flying beyond the high speed limit of its
flight envelope. It is also recommended that such auto flight systems limit the rate of
descent of the aircraft to a certain value that operationally safe.
General recommendation
•
It is recommended that a regional investigation framework for co-operation in aircraft
accident investigations be established to enable fast mobilization of resources and
coordination of activities to support those states that do not have the resources and
facilities to do investigations on their own.
48
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
Air Traffic Services (ATS) Report
Flight Data Recorder, NTSB Report DCA98RA013, 21 June 1999
FDR BEA Garnet Microscope
Cockpit Voice Recorder, NTSB Report DCA98RA013, 21 June 1999
Structures Group Notes – Phase Two, 18 March 1998
Aircraft Technical Log
Summary of Records of Fuselage Skin Repair
Reports on Fasteners, Metallurgical Laboratory – Bandung Institute of Technology,
Bandung - Indonesia, 1998
9. Powerplant Investigation, General Electric (GE) Report No R98AEB122, 30 April
1998
10. Notes on the Actuators Teardown Activities
11. Chemical Analysis of SilkAir Investigation Sample, Analytical Engineering Report,
No. 9-5576-WP-98-195, 27 August 1998
12. Wreckage Trajectory Studies, BASI Report, 16 April 1998
13. NTSB Report DCA98RA013, 27 April 1999
14. Structures Analysis of Empennage Failure, NTSB Report DCA-98-RA-013, 2 May
1999
15. SilkAir Flight Administration Manual + Organization Diagram
16. B737 Rudder Conference, April 1997
17. B737 Training Manual on Stabilizer Trim Unit
18. Cassandra Johnson, Recorded Radar Study, Specialist’s Report of Investigation,
NTSB Report DCA98RA013, 2 November 1998.
49
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix A
Transcript of the Last Portion of CVR Recording
(CVR time is FDR UTC time – 2 seconds)
A-1
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0829:23 {25:29}
CAM-1
checked.
0829:24 {25:30}
CAM-2
checked, flaps?
0829:25 {25:31}
CAM-1
five degrees, green light.
0829:28 {25:34}
CAM-2
five green, rudder, aileron stab trim?
0829:29 {25:35}
CAM-1
zero, zero and four point seven units checked.
0829:35 {25:41}
CAM-2
zero zero four point seven checked, takeoff briefing.
0829:37 {25:43}
CAM-1
complete.
0829:38 {25:44}
CAM-2
it's complete to the line.
0829:40 {25:46}
CAM-1
okay.
0829:50 {25:56}
CAM
((sound of door opening)).
0829:51 {25:57}
CAM-1
all set.
0829:52 {25:58}
CAM-5
yeah just give me one one five seconds to put on the infant
seatbelt you know.
DCA98RA013
A-2
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0829:57 {26:03}
CAM-1
never mind lah you can take quite some time you can take
up to 3 seconds.
0829:59 {26:05}
CAM-5
okay.
0830:00 {26:06}
CAM-1
(sound of laugh) five seconds ( sound of laugh) * * give me
five seconds to put -.
0830:03 {26:09}
CAM
((sound of door closing)).
>0830:44 {26:50}
GND
Silk Air one eight five continue taxi on alpha contact tower
one one eight point seven five.
>0830:48 {26:54}
RDO-2
one one eight seven five Silk Air one eight five.
>0831:34 {27:40}
RDO-2
Tower selamat sore Silk Air one eight five on alpha.
>0831:38 {27:44}
TWR
Silk Air one eight five number two for departure two five
right.
>0831:41 {27:47}
RDO-2
Silk Air one eight five.
>0834:00 {30:06}
TWR
Silk Air one eight five line up and wait two five right.
>0834:03 {30:09}
RDO-2
line up and wait Silk Air one eight five.
DCA98RA013
A-3
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0834:06 {30:12}
CAM-1
sit the girls thanks, below the line.
0834:09 {30:15}
PA-2
cabin crew take off positions please.
0834:17 {30:23}
CAM-2
cabin announcement complete, engine start switches?
0834:20 {30:26}
CAM-1
on.
0834:21 {30:27}
CAM-2
transponder?
0834:22 {30:28}
CAM-1
TA only.
0834:23 {30:29}
CAM-2
strobe lights?
0834:25 {30:31}
CAM-1
on.
0834:26 {30:32}
CAM-2
holding at takeoff clearance.
0834:28 {30:34}
CAM-1
thank you.
0835:22 {31:28}
CAM-1
*.
0835:25 {31:31}
CAM-2
sorry?
DCA98RA013
A-4
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0835:28 {31:34}
CAM-1
this guy is also going papa lima bravo.
0835:30 {31:36}
CAM-2
yeah.
0835:32 {31:38}
CAM-1
Sempati is turning right.
0835:37 {31:43}
CAM-1
we're waiting for him to cross before he can let us go.
0835:40 {31:46}
CAM-2
yeah.
0835:55 {32:01}
CAM-2
they turned around that triple seven fast.
0835:57 {32:03}
CAM-1
yeah.
0835:59
CAM-2
{32:05}
that SQ one.
>0836:30
CAM-1
{32:38}
come on.
>0836:35 {32:41}
TWR
Silk Air one eight five cancel SID after airborne turn right
direct papa lima bravo cleared for takeoff.
>0836:41 {32:47}
RDO-2
airborne right turn papa lima bravo cleared for takeoff Silk
Air one eight five.
0836:44 {32:50}
CAM-2
takeoff clearance.
DCA98RA013
A-5
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0836:45 {32:51}
CAM-1
obtained.
0836:47 {32:53}
CAM-2
before takeoff checklist's complete.
0836:48 {32:54}
CAM
((sound of increasing engine noise)).
0836:50 {32:56}
CAM-1
ninety one five.
0836:55 {33:01}
CAM-2
ninety one five thrust is set.
0836:58 {33:04}
CAM-2
eighty knots.
0836:59 {33:05}
CAM-1
my control.
0837:00 {33:06}
CAM-2
you have control.
0837:12 {33:18}
CAM-2
V-one rotate.
0837:14 {33:20}
CAM-2
V-two.
0837:17 {33:23}
CAM-2
positive climb.
0837:18 {33:24}
CAM-1
gear up thanks.
DCA98RA013
A-6
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0837:28 {33:34}
CAM-1
heading select right turn three three zero yeah.
0837:32 {33:38}
CAM-2
it's all clear.
0837:33 {33:39}
CAM-1
okay.
0837:36 {33:42}
CAM-1
N-one, two ten, flaps one.
>0837:48 {33:54}
TWR
Silk Air one eight five contact departure one one nine seven
five.
>0837:52 {33:58}
RDO-2
one one nine seven five Silk Air one eight five terima kasih
pak.
>0837:55 {34:01}
TWR
*.
>0837:58 {34:04}
RDO-2
ah arrival Silk Air one eight five airborne one thousand six
hundred.
>0838:09 {34:14}
RDO-2
Jakarta approach Silk Air one eight five.
0838:10 {34:16}
CAM-1
flaps up.
DCA98RA013
>0838:12 {34:18}
DEP
Silk Air one eight five identified on departure climb to three
five zero turn right heading three four zero report passing
one five zero.
A-7
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
>0838:21 {34:27}
RDO-2
climb three five zero right turn heading three four zero Silk
Air one eight five roger.
0838:27 {34:33}
CAM-2
flaps are up.
0838:32 {34:38}
CAM-1
V-Nav thanks.
0838:47 {34:53}
CAM-2
TCAS twenty.
0838:48 {34:54}
CAM-1
okay it's set.
0839:15 {35:21}
CAM-1
((sound of humming)).
0839:21 {35:27}
CAM-1
request high speed thanks.
>0839:24 {35:30}
RDO-2
Silk Air one eight five request high speed climb.
>0839:27 {35:33}
DEP
Silk Air one eight five approved.
>0839:29 {35:35}
RDO-2
thank you.
0839:30 {35:36}
CAM-1
okay delete th- start switches off, seatbelt sign off, after
takeoff checklist thanks.
DCA98RA013
A-8
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0839:43 {35:49}
CAM-2
transponder TA-RA twenty miles, air-con and pres is set
climbing start switches off, landing gear up and off, flaps up
no lights, landing lights on until ten thousand, fasten belt's
off, after takeoff complete.
0839:56 {36:02}
CAM-1
thank you.
0840:50 {36:56}
CAM-1
V-Nav heading select.
0840:51 {36:57}
CAM-2
checks.
0840:52 {36:58}
CAM-1
auto-pilot A engaged.
0840:59 {37:05}
CAM-1
we probably get direct Pardi eh probably.
>0841:15 {37:21}
DEP
Silk Air one eight five take up heading three three zero.
>0841:20 {37:26}
RDO-2
heading three three zero Silk Air one eight five.
0841:31 {37:37}
CAM-1
ten thousand checked one zero one three and ah.
0841:38 {37:44}
CAM-2
set.
0841:53 {37:59}
CAM-1
forty miles.
DCA98RA013
A-9
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0843:20 {39:26}
CAM-1
*.
0843:21 {39:27}
CAM-2
yeah.
0843:22 {39:28}
CAM-1
* for the next two days *.
>0843:35 {39:41}
RDO-2
Silk Air one eight five pass one five zero.
>0843:38 {39:44}
DEP
control one two four three five.
0843:40 {39:46}
CAM-1
I'll be off here.
>0843:43 {39:49}
DEP
Silk Air one eight five contact one two four three five.
>0843:46 {39:52}
RDO-2
twenty four three five Silk Air one eight five.
>0843:53 {40:00}
RDO-2
ah Jakarta control Silk Air one eight five climbing three five
zero.
>0844:00 {40:06}
CTRL
Silk Air one eight five maintain heading climb three five zero
report passing two four zero.
0844:05 {40:11}
CAM-2
want direct Pardi.
0844:06 {40:12}
CAM-1
huh.
DCA98RA013
A-10
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0844:06 {40:12}
CAM-2
direct Pardi.
0844:10 {40:16}
CAM-1
call two four zero.
>0844:10 {40:17}
RDO-2
flight level three five zero wilco Silk Air one eight five
request direct Pardi.
>0844:15 {40:21}
CTRL
all right stand-by.
0844:29 {40:35}
CAM-1
I'll be off the air for a while.
DCA98RA013
A-11
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0844:37 {40:43}
PA-1
good afternoon ladies and gentleman this is your Captain
my name is Tsu Wai Ming on the flight deck this afternoon
with me is first officer Duncan Ward we'd like to welcome
you aboard and ah we are now climbing through nineteen
thousand feet we'll be cruising today at thirty five thousand
heading towards the north west tracking initially towards the
eastern cost of Sumatra towards the town of Palembang
before turning right towards Singapore flight time one hour
twenty minutes you can expect ah to arrive at Singapore at
about six o'clock in the evening Singapore time which is one
hour ahead of Jakarta time, time in Singapore is now four
forty five in the afternoon this is about five minutes ahead
of schedule. weather conditions clear skies out of Jakarta
very hot afternoon and at the moment we are still in good
weather however toward Singapore we can expect a bit of
showers thunderstorm towards the southern part of
Singapore arrival at Singapore should be fine with a
temperature of about twenty eight degrees Celsius the
seatbelt sign is now off feel free to move around the cabin
however while seated for your own safety have your
seatbelt fastened sit back and relax enjoy the services
provided today on Silk Air one eight five and I'll get back to
you just before our descent into Singapore with a updated
weather forecast thank you.
0846:04 {42:10}
CAM-1
I'm back with you.
0846:05 {42:11}
CAM-2
okay.
>0847:01 {43:08}
RDO-2
Silk Air one eight five passing two four zero.
DCA98RA013
A-12
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
>0847:07 {43:13}
CTRL
Silk Air one eight five contact Jakarta upper one three two
decimal seven.
>0847:13 {43:19}
RDO-2
one three two seven Silk Air one eight five.
>0847:21 {43:28}
RDO-2
Jakarta Silk Air one eight seven climbing passing two five
five two four five correction.
>0847:30 {43:36}
CTRH
Silk Air one eight five confirm.
>0847:31 {43:38}
RDO-2
affirm Silk Air one eight five climbing three five zero
requesting direct Pardi.
>0847:38 {43:44}
CTRH
one eight five stand-by direct Pardi direct papa lima bravo
report three five zero.
>0847:43 {43:49}
RDO-2
direct Palembang wilco Silk Air one eight five.
0847:50 {43:56}
CAM-1
*.
0847:52 {43:58}
CAM-5
* would you like to have some sandwich.
0847:55 {44:01}
CAM-1
drinks ah.
0847:56 {44:02}
CAM-5
*.
DCA98RA013
A-13
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0847:57 {44:03}
CAM-1
tau hueh chui. (( soya drink))
0847:57 {44:03}
CAM-5
tau hueh chui.
0847:59 {44:05}
CAM-2
I'll have a ice lemon tea.
0848:00 {44:06}
CAM-5
ice lemon tea do you want the sandwich too.
0848:03 {44:09}
CAM-2
what kind.
0848:04 {44:10}
CAM-5
we have egg mayonnaise and chicken *.
0848:08 {44:14}
CAM-2
just a couple thanks nice clear day.
0848:12 {44:18}
CAM
((sound of door closing)).
0848:16 {44:22}
CAM-1
yeah.
0848:33 {44:39}
CAM-1
((sound of singing).
0848:49 {44:55}
CAM-1
some water you want?
0848:51 {44:57}
CAM-2
ah fine thanks.
DCA98RA013
A-14
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0849:48 {45:54}
CAM-1
just go level change and get up.
0849:50 {45:56}
CAM-2
yup.
0849:51
CAM-1
{45:57}
so we can go direct Pardi.
0850:17 {46:23}
CAM-2
thirty for thirty five.
0850:52 {46:58}
CAM-1
on speaker.
0852:18 {48:24}
CAM-1
a thousand to three five zero.
0852:40 {48:46}
CAM-1
*.
0852:49 {48:55}
CAM
((sound of altitude alert tone)).
0853:08 {49:14}
CAM-1
*.
>0853:15 {49:22}
RDO-2
Silk Air one eight five maintaining three five zero.
>0853:20 {49:26}
CTRH
silk one eight five maintain three five zero cleared direct to
Pardi report abeam papa lima bravo.
>0853:25 {49:31}
RDO-2
three five zero direct Pardi wilco Silk Air one eight five.
DCA98RA013
A-15
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0853:51 {49:57}
CAM-1
that's him behind us.
0853:52 {49:58}
CAM-2
yup.
0853:53 {49:59}
CAM-1
very fast.
0857:25 {53:31}
CAM-1
he'll be ahead of us arriving in Singapore.
0857:28 {53:34}
CAM-2
yeah.
0857:28 {53:34}
CAM-1
he is he is speeding, shit.
0857:35 {53:41}
CAM-1
at least point eight *.
0857:52 {53:58}
CAM-2
he'll be above the weather as well.
0900:48 {56:54}
CAM
((sound of door opening)).
0900:51 {56:57}
CAM-5
tau huey chui.
0900:56 {57:02}
CAM-1
thanks.
0901:01 {57:07}
CAM-5
I was so busy I keep two pieces of sandwich for him then
this coming in as well ( sound of laugh).
DCA98RA013
A-16
INTRA-COCKPIT COMMUNICATION
TIME and
SOURCE
CONTENT
AIR-GROUND COMMUNICATION
TIME and
SOURCE
CONTENT
0901:12 {57:18}
CAM
((sound of door closing)).
0904:09 {00:15}
CAM
((sound of rustling papers )).
0904:55 {01:01}
CAM-1
go back for a while, finish your plate.
0904:56 {01:02}
CAM-2
I am.
0905:00 {01:06}
CAM-1
some water.
0905:01 {01:07}
CAM((sound of several metallic snap)).
0905:03 {01:09}
CAM
((sound of snap)).
0905:05 {01:11}
CAM-2
no thanks.
0905:13.6 {01:19.6
C
(end of recording)
DCA98RA013
A-17
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix B
FDR Plots
B-1
B-2
B-3
B-4
B-5
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix C
Transcript of ATC Recording
(ATC time is FDR UTC time + 37 seconds)
C-1
TAPE TRANSCRIPT OF TRAFFIC COMMUNICATION WITH
JAKARTA ATC (CLEARANCE DELIVERY (CLR), TOWER (TWR),
GROUND CONTROL (GND), APPROACH CONTROL (APP) AND
AREA CONTROL (ACC)) ON DECEMBER 19,1997
From
To
Time
Communications
Doubtful
Words
Obs.
SLK 185 CLR
08:17:50 Delivery clearance selamat sore Silk Air 185
CLR
SLK 185 08:17:54 Good afternoon Silk Air 185 Soekarno Hatta delivery go
ahead
SLK 185 CLR
08:17:59 Silk Air 185 D 11 requesting FL 350 to Singapore
CLR
SLK 185 08:17:14 Silk Air 185 350 is already occupied do you accept FL 390
clearance
SLK 185 CLR
08:17:21 Negative FL 310 please
CLR
SLK 185 08:17:27 Roger Silk 185 you are cleared to Singapore via G 579 FL
310 squawk 2344 expect Cengkareng 2 G departure for
runway 25 R
SLK 185 CLR
08:18:38 Roger Singapore G 579 FL 310 squawk 2344
Cengkareng 2 G 25 R Silk Air 185
CLR
SLK 185 08:18:47 Silk Air 185 that is correct when ready to start and push
back contact ground 121.6
SLK 185 CLR
08:18:52 121.6 Silk Air 185 good day
SLK 185 GND
08:22:51 Soetta ground Silk Air 185 selamat siang
SLK 185 GND
08:23:00 Soetta ground Silk Air 185
GND
SLK 185 08:23:04 Silk Air 185 go ahead
SLK 185 GND
GND
08:23:07 Selamat sore Silk Air 185 D 11 request pushback and
start
SLK 185 08:23:10 Silk Air 185 cleared to pushback and start facing Hotel
SLK 185 GND
GND
SLK 185 08:26:54 Silk Air 185 are you able level 350
SLK 185 GND
GND
08:23:15 Facing Hotel Silk Air 185
08:26:58 Silk Air 185 affirmative
SLK 185 08:27:01 Silk Air 185 are you able level 350
SLK 185 GND
08:27:03 350 thank you Silk Air 185
SLK 185 GND
08:27:40 Silk Air 185 taxi
GND
SLK 185 08:27:43 OK runway 25 R via Hotel and Alpha
SLK 185 GND
08:27:47 OK 25 R Silk Air 185
GND
SLK 185 08:29:20 Silk Air 185 you are no.2 after aircraft Fokker 28 out Lima
joining Alpha for runway 25R
SLK 185 GND
08:29:27 Silk Air 185 copied
GND
SLK 185 08:31:21 Silk Air 185 continue taxi on A contact tower 118.75
SLK 185 GND
08:31:26 118.75 Silk Air 185
C-2
From
To
SLK 185 TWR
TWR
Communications
Doubtful
Words
Obs.
08:32:12 Tower selamat sore Silk Air 185 On Alpha
SLK 185 08:32:15 Silk Air 185 you are no.2 departure 25 R
SLK 185 TWR
TWR
Time
08:32:18 Silk Air 185
SLK 185 08:34:38 Silk Air 185 line up and wait
SLK 185 TWR
08:34:41 Line up and wait Silk Air 185
TWR
SLK 185 08:37:13 Silk Air 185 cancel SID after airborne turn right direct PLB
clear for take off
SLK 185 TWR
08:37:18 Airborne right turn PLB clear for take off Silk Air 185
TWR
SLK 185 08:38:26 Silk Air 185 contact departure 119.75
SLK 185 TWR
TWR
08:38:30 119.75 Silk Air 185 terima kasih pak
SLK 185 08:38:33 Sama – sama
SLK 185 APP
08:38:36 Arrival Silk Air 185 airborne one thousand six hundred
SLK 185 APP
08:38:47 Jakarta approach Silk Air 185
APP
SLK 185 08:38:50 Silk Air 185 identified on departure climb to 350 turn right
heading 340 report passing 150
SLK 185 APP
08:38:59 Climb to 350 right turn heading 340 Silk Air 185 roger
SLK 185 APP
APP
SLK 185 08:40:06 Silk Air 185 approved
SLK 185 APP
APP
08:40:09 Thank you
SLK 185 08:41:53 Silk Air 185 pick up heading 330
SLK 185 APP
APP
08:40:03 Silk Air 185 request high speed climb
08:41:58 330 Silk Air 185
SLK 185 08:44:15 Silk Air 185 contact control 124.35
SLK 185 APP
08:44:21 Silk Air 185 contact 124.35
SLK185
08:44:21 124.35 Silk Air 185
APP
SLK 185 ACC
ACC
08:48:00 Jakarta Control SLK 185 climbing 255 2344
SLK 185 08:48:08 SLK 185 confirm?
SLK 185 ACC
08:48:11 Affirm SLK 185 climbing 350 requesting direct Pardi
ACC
GLK 185 08:48:17 SLK 185 stand by direct Pardi, direct to PLB report 350
SLK185
ACC
08:48:21 Direct Palembang wilco SLK 185
C-3
From
To
Time
Communications
SIA 157
ACC
ACC
SIA 157
Selamat siang SIA 157 report the present heading
SIA 157
ACC
SIA 157 we cleared to PLB now maintaining heading 330
ACC
SIA 157
SIA 157
ACC
SIA 157 for avoiding traffic left heading 325 climb to level
390
Climb on heading 325 FL 390 SIA 157
SSR 134 ACC
ACC
SSR 134
SSR 134 ACC
ACC
SSR 134
Doubtful
Words
Obs.
08:49:55 Jakarta Control SIA 157 selamat siang
08:51:40 Jakarta…..Jakarta control this is SSR 134
SSR 134 climb and maintain 310 proceed to PLB
Climb and maintain 310, request direct to Pardi at present
due to weather
Stand by for direct Pardi due to traffic
SSR 134 ACC
Roger SSR 134
SLK 185 ACC
08:53:54 SLK 185 maintaining 350
ACC
SLK 185 08:53:58 SLK 185 maintain 350 clear direct to Pardi report abeam
PLB
SLK 185 ACC
08:54:04 350 direct Pardi, wilco SLK 185
SIA 157
ACC
ACC
SIA 157
Go ahead SIA 157
ACC
SIA 157
SIA 157
ACC
Okay after passing 350 clear direct Pardi report reaching
390 after passing 350 sir
Roger after passing FL 350 cleared direct Pardi, SIA 157
call you reaching 390
QFA 41
ACC
ACC
QFA 41
QFA 41 turn left heading 350 climb to FL 410
QFA 41
ACC
Heading 350, QFA 41
QFA 41
ACC
QFA 41, how long on this heading ?
ACC
QFA 41
QFA 41
ACC
Roger 3 minutes Sir, for avoiding traffic, your traffic now
about 2 o'clock 18 miles 265 climbing to 350
Confirm it's about 2 o'clock?
ACC
QFA 41
Until 3 minutes heading
08:54:10 SIA 157 Jakarta Control
08:57:26 Jakarta Control, QFA 41 is heading 360 on climb FL 410
C-4
From
To
Time
Communications
QFA 41
ACC
ACC
QFA 41
Negative, your traffic is 11 o'clock, sorry 11 o'clock
QFA 41
ACC
Roger, QFA 41.
ACC
QFA 41
QFA 41
ACC
SIA 157
ACC
ACC
SIA 157
SIA 157
ACC
Report abeam PLB, SIA 157.
ACC
QFA 41
QFA 41 you're passing the traffic, now direct to PLB.
QFA 41
ACC
Direct PLB, QFA 41.
Doubtful
Words
Obs.
08:58:15 Confirm the traffic is now 2 o'clock ?
08:58:55 QFA 41 left now heading 330, continue climb to FL410.
Roger, turning 330, continue climb to 410, QFA 41.
08:59:20 Jakarta Control SIA 157 reaching FL390 request direct PARDI.
Cleared direct PARDI report abeam PLB.
SSR 134 ACC
09:00:20 Jakarta control SSR 134 reaching 310 request direct –
PARDI
ACC
SSR 134
SSR 134 roger, maintain 310 cleared direct PARDI, report
abeam PLB.
SSR 134 ACC
Maintain 310 cleared direct PARDI report abeam PLB
abeam SSR 134.
ACC
QFA 41
QFA 41
ACC
QFA 41…..
ACC
QFA 41
QFA 41
ACC
Cleared direct to PARDI report abeam PLB……..cleareddirect PARDI report reaching 410.
Cleared direct to PARDI, call you reaching 410, QFA 41
SIA 157
ACC
ACC
SIA 157
SIA 157
ACC
09:09:40 QFA 41, Jakarta Control……
09:10:33 Jakarta Control SIA 157 abeam PLB/10 FL390 and
estimate PARDI at 27.
Copied 27, maintain 390 at PARDI contact Singapore
134.4
Maintain 390, PARDI 134.4 SIA 157 Selamat siang.
C-5
From
To
Time
Communications
Doubtful
Words
Obs.
ACC
SLK 185 09:10:55 SLK 185 you just passing abeam PLB, maintain 350 at
PARDI contact Singapore 134.4
SLK 185 ACC
09:11:03 SLK 185 roger 134.4 before PARDI.
ACC
SSR 134 09:11:25 SSR 134 maintain 310 just leaving PLB, now contact
Singapore 134.4 at PARDI SSR 134 at PARDI.
SSR 134 ACC
134.4 at PARDI SSR 134.
QFA 41
ACC
ACC
QFA 41
QFA 41 maintaining 410 direct PARDI, report abeam PLB.
QFA 41
ACC
Roger, report abeam PLB, direct PARDI, QFA 41.
09:12:34 QFA 41 maintaining FL410.
09:13:33 (sound off………..Husssssssss……………for a few
second)
ACC
QFA 41
QFA 41
ACC
ACC
QFA 41
Affirmative, Singapore Radar
QFA 41
ACC
Roger good day …….
GIA 238 ACC
ACC
GIA 238
GIA 238 ACC
09:16:38 QFA41 you position abeam PLB, maintain 410, at PARDI
contact Singapore 134.4
Roger 134.4 confirm………..?
09:20:01 Jakarta Control good afternoon GIA 238 passing 250.
GIA 238 climb and maintain 280, proceed PLB
Proceed PLB maintain 280, GIA 238
SIA 157
ACC
ACC
SIA 157
SIA 157 Go-ahead ………
SIA 157
ACC
SIA 157 any objection to descend before Pardi ?
ACC
SIA 157
SIA 157 say again your last message
SIA 157
ACC
Is there any objection if we descend before Pardi?
ACC
SIA 157
SIA 157
ACC
Affirmed, Jakarta no objection, contact Singapore for
traffic below
Roger, thank you, SIA 157
09:20:31 Jakarta Control SIA 157
C-6
From
ACC
To
GIA 238
GIA 238 ACC
ACC
Time
Communications
Doubtful
Words
Obs.
09:36:26 GIA 238 Jakarta …….
Go-ahead, GIA 238
GIA 238
09:36:31 Please relay to SLK 185 …….SLK 185 to contact
Singapore 134.4 sir
GIA 238 SLK 185
Roger, GIA 238 . ……..SLK 185, GIA 238. ………..
C-7
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix D
Wreckage Weight
D-1
SILKAIR WRECKAGE WEIGHT
Box
Number
Box Size
1
Full
2
3
Description
Full Box
Weight (kg)
Empty Box
Weight (kg)
Weight of
Contents (kg)
Misc. Cabin and Interior Items/Cargo Liners/Galley - Box #1
690
130
560
Full
Misc. Fragments – Bits and Pieces - Box #1
1670
130
1540
Full
Misc. Fragments – Bits and Pieces - Box #2
1460
130
1330
4
Full
Misc. Fragments – Bits and Pieces - Box #3
1250
130
1120
5
Full
Misc. Fragments – Bits and Pieces - Box #4
1220
130
1090
6
Full
Misc. Fragments – Bits and Pieces - Box #5
1350
130
1220
7
Full
Misc. Fragments – Bits and Pieces - Box #6
1070
130
940
8
Full
Misc. Fragments – Bits and Pieces - Box #7
880
130
750
9
Full
Misc. Fragments – Bits and Pieces - Box #9 (Note: No Box #.8)
960
130
830
10
Full
Misc. Fuselage Parts and Fragments - Box #1
600
130
470
11
Full
Wing Skin Parts Upper and Lower - Box #1
1260
130
1130
12
Full
Wing Control Surfaces and Associated Parts - Box #1
720
130
590
13
Full
Wing Skin Parts Upper and Lower - Box #2
1200
130
1070
14
Full
Fuselage Skin Parts with Identification Markings, Door Parts - Box #1
580
130
450
15
Full
Landing Gear Parts Wheel Hub and Associated Parts LPT No.2, Exhaust
Sleeve - Box #1
830
130
700
16
Full
Aft Fuselage Skin Section 46, 48 - Box #1
450
130
320
17
Full
Misc. Fuselage Part and Fragments - Box #2
660
130
530
18
Half
Pumps; Valves; Motors/APU + Associated Parts/Fit Control Comps.
700
110
590
19
Half
Flap Tracks Carriages Transmission and Associated Parts – Box #1
700
110
590
20
Half
Thrust Reverser Parts- Box #1
500
110
390
21
Half
No.1 Engine
810
110
700
22
Half
No.2 Engine
1030
110
920
23
Half
Connectors Circuit Boards/Wires – Box #1
480
110
370
24
Quarter
Control Actuation – Box #1
70
50
20
25
Quarter
Control Actuation – Box #2
80
50
30
26
Quarter
Control Actuation – Box #3
150
50
100
27
Quarter
Control Actuation – Box #4
120
50
70
28
Quarter
Control Actuation – Box #5
110
50
60
29
Quarter
Control Actuation – Box #6
110
50
60
30
Half (tall)
Engine Parts
750
120
630
31
Half (tall)
Landing Gear, APU engine diffuser
560
120
440
32
Half
Landing Gear Parts/Support Beam/Pylon/Mid Spar Fitting/A.C. Comps.
830
110
720
33
Half
Aircraft Documents, Aircraft Manuals
N.A.
N.A.
N.A.
34
Full
Parts from Empennage/Rudder, Elevator – Box #1
520
130
390
35
Full
Texas Star/Pressure Bulkhead/Fin Spar ends/Patch
570
130
440
36
Quarter Sm
Engine Part
350
40
310
Right Hand Horizontal Stabiliser Tip
160
110
50
Wheel Assembly
190
40
150
37
Flat
38
Quarter Sm
39
Flat
Right Hand Horizontal Stabiliser Piece
190
110
80
40
Flat
Left Hand Horizontal Stabiliser Tip
220
110
110
41
Half (tall)
Misc. Fragments Bits and Pieces – Box #10
410
120
290
26460
4310
22150
Total Weights (kg)
Total Weights (lb)
58212
9482
48730
Total Weight Excluding Cabin and Interior Items – Box #1 (kg)
25770
4180
21590
Total Weight Excluding Cabin and Interior Items – Box #1 (lb)
56694
9196
47498
Manufacturing Empty Weight PQ973 (no seats, galleys, lav) (lb)
64924
Percentage of Aircraft Recovered
73.2%
D-2
Appendix E
Actuator Matrix
E-1
No.
Actuator/Part Name
1a Main Rudder PCU Piston
1b Main Rudder Servo Valve
Measured
Position
1.79"; 2.21"
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
1.97"
0.2"; Extend
Primary:
Neutral;
Primary: 1/2 rate;
Secondary:
Secondary: Neutral Neutral
**
2 Standby Rudder PCU Piston
3.1"
~ 3.17"
3 Aileron PCU - Piston - 3529
1.39" extend cyl.;
2.07" retract cyl
1.72"
4 Aileron PCU - Piston - 3509
1.83" extend cyl;
1.60" retract cyl
1.72"
Follows
command input Last powered
(always engaged) position
3 deg left
Primary:
Follows
Neutral;
command input Secondary:
(always engaged) Neutral
N/A
Remarks
Main and Standby tied to
same torque tube
(therefore Main and
Standby can not
disagree)
** When unit is physically
disconnected from input
linkage, bias spring
moves Primary to 1/2
PCU rate (unit was
physically disconnected)
Main and Standby tied to
same torque tube
(therefore Main and
Standby can not
disagree)
Follows
Last
command input commanded
0.07"; Retract (always engaged) position
1 deg right
3.5 deg left
Follows
Last
aileron down
0.33" retract; command input commanded or right aileron
0.35" retract
(always engaged) position
up
Rod bent
1.2 deg left
Follows
Last
aileron up or
0.11" extend; command input commanded right aileron
0.12" extend (always engaged) position
down
Rod bent
E-2
No.
Actuator/Part Name
Measured
Position
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
5 Elevator PCU - Piston - 3559
1.59" extend cyl;
1.87" retract cyl
1.72"
0.13" retract;
0.15" retract
6 Elevator PCU - Piston - 3560
Piston missing
1.72"
N/A
7 Flight Spoiler Actuator - 8312A
0.410" **
.4-.5"
Full Retract
8 Flight Spoiler Actuator - 8339A
0.410" **
.4-.5"
Full Retract
9 Flight Spoiler Actuator - 8271A
0.465" **
.4-.5"
Full Retract
10 Flight Spoiler Actuator - 8379A
0.410" **
.4-.5"
Full Retract
11a Aileron Autopilot Servo - A7072
11b Aileron Autopilot Servo - A7017
.103" off center
Centered
Centered
Centered
Centered
Centered
Follows
Last
command input commanded 1.5 deg trailing
(always engaged) position
edge up
Not
Follows
Last
measurable
command input commanded (piston
(always engaged) position
missing)
Retracted/Do
N/A - No servos wn
Down
Retracted/Do
N/A - No servos wn
Down
Retracted/Do
Down
N/A - No servos wn
Retracted/Do
N/A - No servos wn
Down
Disengaged *
Disengaged
*
No input
command to
Ail PCUs
Disengaged *
Disengaged
*
No input
command to
Ail PCUs
Remarks
** Measured Cap to Rod
end
** Measured Cap to Rod
end
** Measured Cap to Rod
end
** Measured Cap to Rod
end
* Disengagement of
detent pistons prevents
mod piston inputs to
PCU
* Disengagement of
detent pistons prevents
mod piston inputs to
PCU
E-3
No.
Actuator/Part Name
Measured
Position
12a Elevator Autopilot Servo - A7076 .25" off center
12b Elevator Autopilot Servo - A6995 .20" off center
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
Centered
Centered
Centered
Centered
Disengaged *
Disengaged *
Follows
command input
(on demand
system)
13 Mach Trim
Fully retracted
Centered
Full Retract
14 Rudder Trim
1.4"
~ 1.44"
~ 0.04" retract ~ 0.06" retract
Follows
command input
Locked/Retrac (on demand
ted
system)
Follows
command input
Locked/Retrac (on demand
ted
system)
Follows
command input
Locked/Retrac (on demand
ted
system)
15 Outboard Ground Spoiler Actuator 0.0" (Flush) **
~ 0"
16 Outboard Ground Spoiler Actuator 0" **
~ 0"
17 Outboard Ground Spoiler Actuator 0" **
~ 0"
Disengaged
*
No input
command to
Elev PCUs
Disengaged
*
No input
command to
Elev PCUs
Remarks
* Disengagement of
detent pistons prevents
mod piston inputs to
PCU
* Disengagement of
detent pistons prevents
mod piston inputs to
PCU
Last
commanded
position
Last
commanded
position
4.2 degrees
elevator trailing Provides input into
edge up
Elevator PCUs
Min right
rudder trim
Provides input into
Rudder PCUs
Piston
retracted
Down
** rod end to end cap
Piston
retracted
Down
** rod end to end cap
Piston
retracted
Down
** rod end to end cap
E-4
No.
Actuator/Part Name
Measured
Position
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
Inboard Ground Spoiler Actuator
18 (A)
22 Feel & Centering Unit
N/A
Full Extended/
Locked
~ 9.26" **
~ 9.25"
Full Extended/
Locked
~ 9.26" **
~ 9.25"
Full Extended/
Locked
Inboard Ground Spoiler Actuator
21 (D)
N/A
~ 9.25"
Inboard Ground Spoiler Actuator
20 (C)
Full Extended/
Locked
~ 9.26" **
Inboard Ground Spoiler Actuator
19 (B)
Follows
command input
(on demand
system)
Follows
command input
(on demand
system)
Follows
command input
(on demand
system)
Follows
command input
(on demand
system)
~ 9.26" **
~ 9.25"
** Impact marks on
CAM
Centered
Remarks
Piston
extended
Down/Locked ** in total length
Piston
extended
Down/Locked ** in total length
Piston
extended
Down/Locked ** in total length
Piston
extended
N/A
Down/Locked ** in total length
** Measured rudder left
Rudder Left
based on impact marks
E-5
No.
Actuator/Part Name
Measured
Position
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
23 Stab Jackscrew
2.5 units *
3 units **
Cruise
Follows
Condition (*1) command input
24 T E Flap Ballscrew
Fully retracted
Fully
retracted
Fully Retracted N/A
25 T E Flap Ballscrew
Fully retracted
Fully
retracted
Fully Retracted N/A
26 T E Flap Ballscrew
Fully retracted
Fully
retracted
27 Locking T/R Actuator (LA)
Stowed
Stowed
Fully Retracted N/A
Locked/Stowe
d
N/A
Stowed
Unlocked/Stow
ed
N/A
28 Locking T/R Actuator (LB)
Stowed
Last
commanded
position
Last
commanded
position
Last
commanded
position
0.5 deg
Leading Edge
Up
Fully retracted
Fully retracted
Last
commanded
position
Fully retracted
Stowed
Stowed
Remarks
* These units are pilot
units on the control
stand; ** 0 deg
Horizontal Stab; (*1)
Verify Cruise Condition;
Check FDR data; Note:
2.5 units correspond to
Airplane Nose down
Electric Limit
Flap Ballscrew position 2
or 7; outboard flap,
inboard track
Flap Ballscrew position 4
or 5; inboard flap,
inboard track
Flap Ballscrew position
1, 2, 7 or 8; Either
outboard flap, either
track
Stowed
Stowed
Unlocked actuator piston
was approximately 1/8"
extended
E-6
No.
Actuator/Part Name
Measured
Position
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
29
30 Sync Lock T/R (LC)
31 Sync Lock T/R (Z)
32 Center Non-Lock T/R
33 Center Non-Lock T/R
34 Center Non-Lock T/R
Stowed
Stowed
Locked/Stowe
d
N/A
Stowed
Stowed
Locked
N/A
Locked
N/A
Stowed
Stowed
Locked
N/A
Locked
N/A
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
35 Lower Non-Lock T/R (A)
36 Lower Non-Lock T/R (B)
37 Lower Non-Lock T/R (C)
38 Lower Non-Lock T/R (D)
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
Stowed
Stowed
Stowed
N/A
Stowed
Stowed
N/A
Stowed
Stowed
Locking T/R Actuator (LC)
42 LE Flap Act - 0095
Stowed
Stowed
Stowed
High power fwd
thrust; Fuel Shut off
Lever in run
Idle/minimu
position
m fuel flow N/A
Extend = Flap
Extended 1.84"
Retracted Down
Extend = Flap
Retracted
Retracted Down
Extend = Flap
Retracted
Retracted Down
43 LE Slat Act - YL768 #1
Fully retracted
39 Throttle Box
40 LE Flap Act - 0271
41 LE Flap Act - 0745
Retracted
Retracted
Follows
Command input
No Locking
Mechanism
No Locking
Mechanism
No Locking
Mechanism
Lock stud, no
damage, 1 finger
broken
Last
commanded
position
N/A
Slightly
Retracted
Extended
Retracted
Retracted
Retracted
Retracted
Retracted
Slats retracted
Remarks
Broken Sync shaft
Piston Stowed; Rod
broken & extended 4"
Thrust Control rack
apparently pulled out by
impact forces (missing)
Marks in bottom of bore
near retract position
E-7
No.
Actuator/Part Name
Measured
Position
Electrical,
Actuator
Hidraulic,
Nominal/
Expected Equivalent
Extend or Mechanical
Neutral
Unpowere Surface
Retract
control input
Position
d State
Position
Direction Engaged or
Disengaged
44 LE Slat Act - YL509
Inner Piston
Extended .5"
Retracted
Retracted
45 LE Slat Act - B-401 #2
Fully retracted
Retracted
Retracted
46 LE Slat Act - B-401 #1
Fully retracted
Retracted
Retracted
47 LE Slat Act - YL689
Fully retracted
Retracted
Retracted
48 LE Slat Act - YL768 #2
Fully retracted
Retracted
Retracted
No damage to
fingers, Lock
stud deformed
Not torn down
Lock stud,
fingers intact
Lock stud,
fingers intact
Lock stud,
fingers intact
Retracted
Slats retracted
Retracted
Slats retracted
Retracted
Slats retracted
Retracted
Slats retracted
Retracted
Slats retracted
Remarks
E-8
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix F
Letters from AlliedSignal
F-1
F-2
F-3
F-4
F-5
F-6
F-7
Follow-up Explanation
From
Greg Francois
June 23, 2000
I had told the professor that from the information that
was provided to us, there was no evidence on the SSCVR
recording itself that the breaker had been pulled. I
explained the way that the SSCVR worked (with the
capacitive “battery”), such that the SSCVR recorder would
STILL record (to the Crash Survivable Memory) after power
had been removed for a minimum of 200 milliseconds.
I
said that there would then have been plenty of time for
the audio indication (if any) from the CVR breaker to be
recorded by the SSCVR (I had personally done this in
demonstrations and ground tests).
There was no evidence on the recording itself. The only
reason I know this is that the team came to Redmond (I
know someone from the NTSB was there, but I can’t remember
who) and we did play back the recording and listened to
the last portion (and even captured the last few hundred
milliseconds of the recording on an oscilloscope). There
was no indication of any change whatsoever in cockpit
audio signal (no sound of the breaker, no muffling of the
area microphone, no change in the audio waveform, etc.).
What I did tell the Professor was that there should be a
TEST conducted to attempt to duplicate the cockpit
conditions and verify if the pulling of the breaker (the
audio indication that is) could or could not be heard. It
was my understanding that a test was conducted by Boeing,
but I never heard the results.
What I think I also said was that the most probable cause
of the recorder going off was indeed power being removed.
I think that I may have also said that another way to
remove power to the recorder (other than pulling the
breaker in the cockpit) was to cut the power wire (I’m
pretty sure that this was also testing) or pull the
recorder out of the mounting tray.
Sorry that this is causing so much trouble. I just knew
that the area microphone used on the SSCVR was pretty
sensitive (especially within a few feet) and that the
breaker wasn't too far away from it (on the overheard
F-8
panel) and that I would have THOUGHT that if the breaker
had been pulled, an AUDIO indication on the SSCVR
recording itself would have been likely and that this
should be verified (or refuted) by actual testing (ground
and flight tests), which again I believe were conducted.
F-9
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix G
Results of Flight Simulation Exercises
G-1
1. First Simulation Test
This simulation test was conducted on 23 January 1998 at Boeing M-Cab engineering
flight simulator, Seattle - USA. The M-Cab is a full motion, multi-purpose flight
simulator which is used to simulate various types of Boeing aircraft. In the case of B737300, the nonlinear mathematical software has been validated using flight test data up to
Mach 0.89, and extrapolated using aerodynamics data based on transonic wind tunnel
data from 0.89 to 0.99 mach.
The objective of the test was to explore and understand the various possible combination
of one or more malfunctions of flight controls, aircraft systems and power plants that
would result in the extreme descent flight trajectory as suggested by the radar plots.
There were 21 scenarios performed as follows:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Autopilot (A/P) ON, Auto Throttle (A/T) ON and One engine FAIL
A/P OFF, A/T OFF and One engine FAIL
A/P ON, Yaw Damper (Y/D) FAIL and A/P was disconnected after 80 seconds
Repeat of scenario 3 with A/P OFF
A/P ON, Rudder INPUT simultaneously A/P disconnected, aircraft rolled to 120
degrees bank. Manual recovery.
A/P OFF, Rudder INPUT for 4 seconds (aircraft rolled right round) and manual
recovery
A/P OFF, Rudder INPUT for 2 seconds (aircraft bank 85 degrees) and manual
recovery
A/P ON, A/P hard over, A/P disconnected and manual recovery
A/P ON, A/P disconnected and push-over from level flight. The G’s forces
recorded at about 0.2 G.
A/P OFF, Repeat push-over (¼ column) and G’s forces recorded at about 0.5 G
until the aircraft descended to 30,500 feet.
A/P OFF, One Engine FAIL just before the last radar point at 35000 ft
A/P OFF, A/T ON (This session was aborted due to engine inadvertently left failed
from the previous session.)
A/P OFF, A/T ON (to give max asymmetry with one Engine fail)
A/P ON, A/T ON and Y/D hard-over
A/P OFF, Yaw damper hard-over (manual recovery after 5 seconds )
A/P ON and A/T OFF, Rudder input (the aircraft descended to 33,700 feet at 270
knots with 90 degrees bank)
A/P OFF and A/T OFF, Rudder input
A/P ON and A/T ON, A/P hard over in lateral axis
A/P OFF and A/T OFF, control column pushed and held at 0.75 position down
until speed reached Mach 1 (high control column force required )
A/P OFF and A/T OFF (try to show the locations of the radar points)
A/P OFF and A/T OFF. Rudder Input (aircraft allowed to spiral for much longer
period than so far experienced).
Note: Scenarios 1 to 10 were conducted without simulator motion.
The results from the test indicated that none of the scenarios matched the flight trajectory
as suggested by the radar points.
G-2
2. Second Simulation Test
The simulation test was conducted on 12 May 1998 at the same Boeing M-Cab facility,
using corrected data provided by NTSB and Boeing.
The objective of the second test was to demonstrate aircraft responses to specific inputs
with appropriate pilot actions to recover. Pilots were to apply appropriate recovery
techniques after approximately 4 seconds per JAR ACJ 25.1329 sec 5.3.1 & 5.3.2.
The initial flight conditions of the simulation were based on approximate MI-185 cruise
conditions: weight = 110,000 lbs., CG = 18% MAC, altitude = 35,000 feet, airspeed =
Mach 0.74 and flaps up.
There were 20 scenarios performed in two sessions as follows:
Session One
1.
2.
3.
4.
5.
A/P ON, one engine FAIL
A/P ON, one engine FAIL, recovery demonstrated after A/P disconnected
A/P ON, Yaw Damper FAIL
Yaw Damper FAIL with A/P disconnected after approximately 50 seconds
A/P hard over (A/P disconnected almost immediately due to aircraft monitoring
system)
6. Attempts to replicate A/P hard over using manual inputs to simulate the monitors not
functioning on the aircraft, A/P OFF
7. A/P ON. Full right pedal input, recovery demonstrated after A/P disconnected
8. A/P OFF. Full pedal input recovery demonstration
9. A/P ON. Full pedal input. Recovery after aircraft inverted with A/P disconnected.
10. Backdrive through radar points, demonstration 1
11. Backdrive through radar points, demonstration 2
12. Manual attempts to fly through radar points.
Session Two
1.
2.
3.
4.
5.
6.
7.
8.
Manual attempt to fly through radar points with full thrust and full pedal input.
Manual attempt to fly through radar points
Full right pedal input recovery demonstration
Full right pedal input and attempt to simulate an A/P type of response
Full right pedal input recovery demonstration
Manual attempt to fly through radar points
Manual attempt to fly through radar points
Full right pedal input with no pilot inputs to recover.
The results of the second simulation test were summarized as follows:
a) The descent trajectory resulting from any single failure would not fly through all the
radar points.
b) Any single failure could be effectively recovered by the pilot.
G-3
c) The manual maneuvering of the aircraft by rudder/ailerons/elevators inputs would
result in flying through some but not all of the radar points. The G load recorded was
in excess of 2 G.
3. Third Simulation Test
The test was conducted on 17 February 1999 and 6 March 1999 by NTSC at the Garuda
Indonesia B737-300 Training Flight Simulator facility in Jakarta. The test on 6 March
was a repeat from the 17 February test with corrected weight & balance data.
The objective of the test was to verify certain findings by the Engineering and Operation
groups regarding the horizontal stabilizer settings. The horizontal stabilizer screw jack
was found at a position corresponding to a trim setting of 2.5 units. This coincides with
the forward manual electric trim limit. However, FDR data shows that the horizontal
stabilizer trim setting during cruise was at 4.61 units.
Based on radar data, a time factor of 32 seconds to descend from FL350 (35000 feet) to
approximately FL195 (19500 feet) was targeted for this test. Radar points were not
considered as they were not available in this simulator.
The test scenarios were:
1.
2.
3.
4.
5.
6.
Runaway horizontal stabilizer trim.
Rudder hard over.
Aileron hard over.
A combination of rudder hard over and runaway stabilizer trim.
A combination of aileron hard over and runaway stabilizer trim.
A combination of sustained manual inputs of rudder and/or aileron plus manipulation
of elevator (push and pull).
The results of scenarios 1 to 5 showed that the descent time did not match the target. In
scenario 6, the descent time matched the target. However, the g-factor could not be
ascertained for all the scenarios.
4. Computer Simulation Fly-out Study
NTSB independently performed a simulation fly-out study of the MI-185 descent
trajectory. The study was conducted using NTSB B737-300 NT Workstation based
simulation. All simulations used the following cruise flight conditions: weight = 109,920
lbs., CG = 18% MAC, altitude = 35,000 feet, airspeed = Mach 0.74 and flaps up.
The simulation study used similar scenarios as in the second simulation test at the Boeing
M-Cab facility.
From this study, NTSB reported on 6 May 1999 that airplane response to pitch control
failures, rudder control failures and autopilot roll failures has been investigated. Pitch
control failures and autopilot failures do not match the available radar data. Yaw Damper
failures do not match the radar data. A rudder hard over failure will not match the radar
G-4
data unless accompanied by adverse pilot action. Several scenarios have been identified
in which active pilot control will produce a match with the radar data.
5. Fourth Simulation Test
This test was conducted on 13 July 1999 at the same Boeing M-Cab simulator facility as
mentioned above. It incorporated scenarios from the second and third simulation tests.
The objective of the test was to review and reaffirm the results of the three simulation
tests conducted so far, particularly with the effect of the manual electric horizontal
stabilizer trim being at the full forward limit. Furthermore, emphasis was given to the
value of load factor, airspeed, pitch angle and rate of descent.
The scenarios performed were as follows:
Case #4
Case #5
Case #6
Case #7
Case #8
Case #9
Case #11
Case #12
Case #14
Case #15
Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Healthy aircraft, Pilot Control Inputs: Wheel and Column
Full Rudder Input, Pilot Control Inputs: Wheel and Column
Full Rudder Input, Pilot Control Inputs: Wheel and Column
Full Rudder Input, Pilot Control Inputs: Wheel and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel and Column
Full Rudder Input, Pilot Control Inputs: Wheel and Column
Full Rudder Input, Pilot Control Inputs: Wheel and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel only
Case #16
Case #17
Case #18
Case #21
Case #23
Stab Trim Runaway, Pilot Control Inputs: Wheel, Pedal and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel, Pedal and Column
Healthy aircraft, Pilot Control Inputs: Wheel and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel, Pedal and Column
Stab Trim Runaway, Pilot Control Inputs: Wheel, Pedal and Column
Case #30 Right Aileron Hard over (trailing edge up), Pilot Control Inputs: Wheel, Pedal
and Column
Case #31 Right Aileron Hard over (trailing edge up), Pilot Control Inputs: Wheel, Pedal
and Column
Case #32 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Case #33 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Case #34 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Case #35 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Case #36 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
Case #37 Healthy aircraft, Pilot Control Inputs: Wheel, Pedal and Column
The results from the test indicated that for the scenarios that were flown with the
Horizontal Stabilizer Trim in the normal cruise position (4.61 units) would generally
produce a load factor higher than 2 G during the descent.
The scenarios flown with the Horizontal Stabilizer Trim in the manual electric forward
limit (2.5 units) would generally produce a load factor lower than 2 G during the descent.
G-5
Appendix H
Site Acceptance Certificate of
The Hughes GUARDIAN System
H-1
H-2
H-3
H-4
H-5
National
Transportation
Safety
Committee
Aircraft Accident Report- Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix I
Professional Events in the Flight Crew’s Career during
1997
I-1
Professional Events in the Pilot-in-Command’s Career During 1997
18 February
3 March
8 April
10 May
3 June
6 June
19 June
24 June
25 June
3 July
7 July
15 July
28 July
1 August
PIC attends base check. Performance rated as ‘above average’.
PIC was flying pilot on a flight from Singapore to Manado,
Indonesia. During approach to runway 36, the crew conducted a goaround manoeuvre.
The PIC was selected as a line instructor pilot (LIP).
The PIC commenced line operations as an LIP.
The first officer on the 3 March flight was interviewed by the B737
Fleet Manager about the Manado incident.
The B737 Fleet Manager interviewed and obtained written
statements from the PIC and the first officer on the 3 March flight,
and wrote an initial report. He recommended that further
investigation was required.
The Flight Operation Manager re-interviewed both pilots involved in
the Manado incident (3 March).
The PIC and the first officer on the 3 March Manado flight were
rostered to fly together. While preparing for the flight in the
flightdeck, the PIC had asked the first officer some questions about
the Manado incident.He then pulled the circuit breaker for the
cockpit voice recorder (CVR) in order to preserve the previous
conversation between the two pilots as evidence for the ongoing
investigation into the Manado incident. The PIC decided to reset the
circuit breaker before take off. The flight continued uneventfully.
Both pilots on the 24 June flight reported the CVR circuit breaker
incident to management. The Flight Operations Manager decided to
conduct a Divisional Inquiry into the incident.
The PIC was handed a letter summarizing the findings of the inquiry
panel into the circuit breaker incident. The PIC was also informed
that the company had decided to remove him from his LIP position
immediately, and he was reprimanded for his actions. Attached to the
letter was a form that asked the PIC whether or not he accepted the
company’s decision.
The PIC returned the form attached to the 3 July letter, and stated
that he did not accept the company’s decision. As a result, a
Company Inquiry into the go-around and circuit breaker incidents
was held.
The PIC was informed by letter that the Company Inquiry would
take place on 28 July.
The Company Inquiry into the go-around and circuit breaker
incidents took place.
The panel recommended that the original penalty stands.
The PIC was informed of the Company Inquiry’s decision by the
Flight Operations Manager. During this meeting, the PIC indicated
that he did not accept the decision and he inquired about further
I-2
7 August
19 August
16 December
19 December
avenues of appeal.
The PIC completed a base check. Performance was rated ‘above
average’.
The PIC had a meeting with the SilkAir General Manager about the
inquiries.
The PIC last flight assignment was on this date
Accident
I-3
Professional Events in the First Officer’s Career During 1997
3 March
4 June
14 Sept.
15 Sept.
10 October
16 December
19 December
The F/O completed base check. Performance rated as “above
average”.
The F/O was found fit during his last medical examination.
The F/O completed his most recent recurrent training (included
recovery from unusual attitude) and was found to be satisfactory.
The F/O completed base check. Performance rated as “above
average”.
The F/O completed a line check, and no problems were noted.
The F/O last flight assignment was on this date
Accident
I-4
Appendix J
History of FAA AD Related to
Boeing 737 Rudder System
J-1
J-2
J-3
J-4
J-5
J-6
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix K
Boeing B737 Non-normal Procedures –
Emergency Descent
K-1
K-2
K-3
Appendix L
Boeing B737 Alert Service Bulletin, Subject on Flight
Controls – Trailing Edge Flap and Horizontal
Stabilizer Trim
L-1
L-2
L-3
L-4
L-5
National
Transportation
Safety
Committee
Aircraft Accident Report- Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix M
Singapore Accredited Representative’s Comments on
Draft Final Report
Note:
For ease of reference by readers, NTSC’s comments (in
courier) are printed onto the attached comments by the
Singapore Accredited Representative.
M-1
M-2
Attachment to MCIT/CA/MI185
dated 8 December 2000
COMMENTS ON DRAFT FINAL REPORT
1.
For the benefit of most readers with little knowledge of aircraft accident
investigations, the aim and nature of the technical investigation should be explained
upfront in the final report and executive summary. For example, paragraph 3.1 of
Annex 13 states that the sole objective of the investigation of an accident shall be the
prevention of future accidents. Paragraph 5.4.1 of Annex 13 and ICAO Manual of
Aircraft Accident and Incident Investigation Part I – Organisation and Planning (Doc
9756) state that any judicial or administrative proceedings to apportion blame or
liability should be separate from any investigation conducted under the provisions of
Annex 13.
2.
The reason for informing the aviation security authorities should be explained in
the final report as well as in its executive summary to avoid any possible
misunderstanding by the public. It should explain that the NTSC had informed the
relevant aviation security authorities in 1999 based on paragraph 5.11 of Annex 131. As
such, there is a separate investigation conducted by the aviation security authorities, i.e.
the police. It should also explain that the final report of this technical investigation does
not go in depth into the pilots’ background as it is the subject of the separate
investigation by the aviation security authorities.
NTSC’S COMMENTS:
The aviation security authorities were notified
taking into account the findings of the Human
Performance & Factors Group in July 1999.
3.
The Singapore Police has commented that two statements in the report relating
to the financial position of the pilot are inaccurate:
Page 25 paragraph 1.18.3.3
The statement “During 1990-1997 the PIC traded over 10 million shares, where
the value and the volume of the trading increased significantly every year”
should be corrected. The value and volume of the PIC’s share trading did not
increase every year but fluctuated during this period. Also, the two dates from
which the PIC’s trading activities were suspended should read as 9 April 1997
(not 15 April 1997) and 9 December 1997 (not 4 December 1997).
Page 43 para 2.14.3
The statement “The data available also showed that his loans and debts were
greater than his realizable assets” is incorrect. The pilot's realizable assets
were higher than his loans and debts.
1
Para 5.11 of Annex 13 states that “If, in the course of an investigation it becomes known, or it is
suspected that an act of unlawful interference was involved, the investigator-in-charge shall immediately
initiate action to ensure that the aviation security authorities of the State(s) concerned are informed.
M-3
Attachment to MCIT/CA/MI185
dated 8 December 2000
4.
The Singapore Police also recommends that the report include mention of the
capital gains made by the pilot over the years through the sales of his houses. In
addition, it should also mention that the First Officer had another insurance bought in
1992.
NTSC’S COMMENTS:
The First Officer’s insurance policy bought in 1992
was not relevant to the accident.
5.
We suggest that the fact that only a relatively small amount of wreckage was
recovered be reflected in the report. This will illustrate clearly the difficulties faced by
the investigation team and enable readers to better appreciate the NTSC’s tests (and
hence its findings) as well as its constraints due to the extent of damage to the wreckage
and the non-recovery of many critical parts of the aircraft. This would also give readers
an understanding of why comprehensive tests using the recovered wreckage could not
be conducted. As such, we propose the following:
Page 2 paragraph 1.3
There should be a more detailed description of the wreckage recovered. The
report should also elaborate that the state in which the wreckage was found
severely limited the extent to which meaningful reconstruction could be
undertaken.
NTSC’S COMMENTS:
The report has stated that the wreckage was very
fragmented. Section 1.12 contains details of
wreckage recovered.
Page 7 & 8 paragraph 1.12
The report should state expressly that the cockpit and the circuit breaker panel
were not recovered from the wreckage.
Page 9 paragraph 1.12.1.2 & Page 28 para 2.2
The report states that “examination of the recovered passenger oxygen
generators revealed no evidence of activation from which it concluded that the
aircraft did not experience depressurization in flight”.
M-4
Attachment to MCIT/CA/MI185
dated 8 December 2000
We suggest to insert a line before it, “Not all of the passenger oxygen
generators were recovered.”
Page 17 paragraph 1.12.4
The report states that “Approximately 370 kg of electrical wires, connectors and
circuit boards of the aircraft were recovered”.
We suggest that the word “Approximately” be replaced by “Only” to avoid any
misimpression that a very substantial part of the aircraft was recovered.
Page 6 Section 1.11.1
6.
We propose to change the third paragraph to read as “The FDR module was first
cleaned and then packed in a container filled with clean water (to prevent the tape
medium from drying out and becoming brittle). It was hand-carried to the United
States National Transportation Safety Board (NTSB) HQ’s readout facility in
Washington D.C., USA.” This is to avoid readers having the misimpression that the
FDR tape was damaged because the FDR module was carried to the NTSB immersed in
river water.
Page 22 Section 1.17
7.
In the third paragraph fifth sentence, it is incorrectly stated that "All managers
are seconded from SIA." SilkAir's Engineering Manager (now titled Senior Manager
Engineering) is a SilkAir employee and not on secondment from SIA. We suggest
changing the sentence to “The majority of SilkAir's senior managers are seconded from
SIA.”
Page 23 Section 1.17
8.
We suggest changing the existing sentence “Disciplinary inquiries are rare.” to
read as “As SilkAir is a small organisation, it is natural that disciplinary inquiries are
rare.”
Page 25 Section 1.18.3.2
9.
In the third paragraph it is incorrectly stated that the PIC was appointed Captain
on 26 January 1996. He was appointed on 27 January 1996. Also, SilkAir wishes to
add that while the PIC was selected for command training on 22 October 1995, he was
officially informed of his selection only on 20 November 1995. We suggest amending
the third paragraph to read as “The PIC was selected for B737 command training on 22
October 1995. He was officially informed of his selection on 20 November 1995. He
M-5
Attachment to MCIT/CA/MI185
dated 8 December 2000
was appointed Captain on 27 January 1996, and confirmed in that position on 27 July
1996.”
Page 25 Section 1.18.3.2
10.
In the fourth paragraph, SilkAir wishes to advise that the PIC was written to and
advised that he had been selected for LIP training on 8 April 1997. Also, while SilkAir
wrote to the PIC to advise him of his de-appointment as LIP on 3 July 1997, this was
subsequently revised to 28 July 1997 following a company inquiry. We propose
amending the fourth paragraph to read as “SilkAir wrote to the PIC on 8 April 1997 to
advise him of his selection for LIP. He completed his training on 9 May 1997. He
performed satisfactorily thereafter in this position. On 3 July 1997, SilkAir wrote to the
PIC to advise him of his de-appointment as LIP. This was subsequently revised to 28
July 1997 following a company inquiry into an operational incident which occurred on
24 June (see Appendix 1 for details).”
Page 38 Section 2.9
11.
In the earlier Section 1.7.4 (page 5) second bullet point, it was mentioned that
Qantas 41 reported that "the weather was good except for two or three isolated
thunderstorms about ten miles to the east of track near Palembang." On page 38
Section 2.9 the sentence about Qantas 41 stated that “Qantas 41 …. did not report
adverse weather over Palembang.” To be consistent with Section 1.7.4, we suggest
that the sentence on Qantas 41 in Section 2.9 on page 38 be changed to “The Qantas 41
…. did not report adverse weather over Palembang except for two or three isolated
thunderstorms about ten miles to the east of track near Palembang.”
Page 43 Section 2.14.5
12.
The last paragraph is intended to cover the whole of the preceding Sections
2.14.1 to 2.14.5 and not just the topic of “insurance”. It may be misleading to put this
paragraph immediately after Section 2.14.5 without another heading. We suggest
adding a new heading “2.14.6 Overall Comments on Section 2.14 ” before this
paragraph.
Appendix I - Professional Events in the PIC's Career during 1997
13.
In the item against the date of 1 August in this Appendix, it is stated that (on 1
August) the PIC met with the Flight Operations Manager and indicated then that he did
not accept the decision of the company's inquiry and asked about further avenues of
appeal. SilkAir wishes to state that both the Flight Operations Manager and the Fleet
Manager B737 who met with the PIC on that day cannot recall him saying that he did
not accept the decision. They do, however, recall that he was not happy with it and
M-6
Attachment to MCIT/CA/MI185
dated 8 December 2000
asked if a meeting with the General Manager to appeal his case was appropriate. We
propose to amend the second sentence of this item to read as “During the meeting, the
PIC was not happy with the decision and asked if a meeting with the General Manager
to appeal his case was appropriate.”
Recommendation
14.
We propose the addition of the following recommendation:
“To facilitate the recovery of flight recorders after impact into water, it is recommended
that a review of the flight recorders design philosophy be undertaken by the equipment
manufacturers to ensure that the underwater locator beacons (ULB) are fitted to the
flight recorders in such a manner that the ULB would not be separated from the
recorders in an accident.”
M-7
National
Transportation
Safety
Committee
Aircraft Accident Report - Appendix
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Appendix N
USA Accredited Representative’s Comments on
Draft Final Report
Note: - For ease of reference by readers, NTSC’s comments (in
courier) are printed onto the attached comments by the USA
Accredited Representative.
- Legend :
Times New Roman, Italics – Draft Final Report
Times New Roman, Bold – NTSB’s Comments
Times New Roman, Normal – NTSB’s Comments
Courier New, Italics – NTSC’s Comments
N-1
N-2
N-3
Comments on Draft Final Report of Aircraft Accident
Submitted by the Accredited Representative of the
United States National Transportation Safety Board
SUMMARY
Introduction
As the state of Design and Manufacture of the Boeing 737 airplane, a United
States Accredited Representative and advisors1 participated in all aspects of the Republic
of Indonesia’s National Transportation Safety Committee (NTSC) investigation into the
December 19, 1997, crash of SilkAir flight MI185 in the Musi River, Palembang,
Indonesia. On October 17, 2000, the Safety Board received the NTSC’s draft Final
Report. These comments are submitted pursuant to Section 6.9 of Annex 13 to the
Convention on International Civil Aviation, which provides that the State conducting the
investigation “shall either amend the draft Final Report to include the substance of the
comments received, or append the comments to the Final Report.”
Review of the draft Final Report revealed that several sections require correction,
clarification, or the inclusion of additional information. Of greatest concern are the
statements in the draft Final Report that the “NTSC is unable to find the reasons for the
departure of the aircraft from its cruising level of FL350 and the reasons for the stoppage
of the flight recorders” and the “investigation has yielded no evidence to explain the
cause of the accident.”
A significant amount of pertinent factual information developed during the 3-year
investigation is either not discussed in the draft Final Report or not fully considered in
analyzing the cause of the accident. In particular, the draft Final Report does not take into
account all of the investigative findings of the Human Performance Group (HPG), which
were documented in a report produced July 30, 1999, and identified as version 6.0. This
version of the HPG report was the only version that was developed through consensus
agreement among the group members, which included representatives from the
Indonesian AAIC (who served as Group Chairman), the Singapore CAA, the United
States NTSB, and the Australian BASI.2 Relevant content from version 6.0 of the HPG
1
Advisors to the U.S. Accredited Representative included representatives from the Federal Aviation
Administration, Boeing Commercial Airplane Group, United Technologies, and Pratt and Whitney.
NTSC’S COMMENTS:
The NTSC understood that the advisors to the US Accredited
Representative included representatives from General Electric as
the aircraft had GE CFM-56 engines. NTSC was not aware that the
representatives included Pratt and Whitney’s.
2
The HPG provided version 6.0 of its report to the NTSC; however, the document that was designated by
the NTSC as the final HPG report (without consensus agreement from the HPG members) omits a
N-4
report is provided later in this document in connection with specific comments on
individual sections in the draft Final Report. Among other things, version 6.0 of the
report contains comprehensive information about the flight crewmembers, including
information about their professional, personal, and financial backgrounds. For example,
substantial information was developed indicating that the captain’s professional and
financial situations had undergone negative changes in the months preceding the accident.
It is disappointing that much of this information was either omitted from the draft final
report or was not fully analyzed.
NTSC’S COMMENTS:
The Australian BASI (now ATSB) advisor is not an
official
representative.
The
Singapore
Accredited
Representative is from the Ministry of Communications
and Information Technology and his advisors are from
the
Civil
Aviation
Authority
of
Singapore,
SIA
Engineering Company and SilkAir.
The final draft is based on information that had since
overtaken the HPG report version 6.0, for example the
PricewaterhouseCoopers
audit
report.
Hence,
sole
reliance on version 6.0 would lead to inaccurate
conclusions.
All group reports have been analysed and integrated
into the final report.
Specific details, such the
pilots’
personal
details
etc.,
were
not
deemed
appropriate for inclusion in the final report.
Following this summary, this document suggests specific corrections,
clarifications, and/or additions for each section of concern. This summary provides an
overview of the primary areas of concern and offers an explanation for the accident that is
consistent with all of the evidence. As further discussed in this summary, when all of the
investigative evidence is considered, it leads to the conclusions that: 1) no airplanerelated mechanical malfunctions or failures caused or contributed to the accident, and 2)
the accident can be explained by intentional pilot action. Specifically, a) the accident
airplane’s flight profile is consistent with sustained manual nose-down flight control
inputs; b) the evidence suggests that the cockpit voice recorder (CVR) was intentionally
disconnected; c) recovery of the airplane was possible but not attempted; and d) it is more
likely that the nose-down flight control inputs were made by the captain than by the first
officer.
significant amount of information that was included in version 6.0.
N-5
NTSC’S COMMENTS:
The NTSC is aware that others may draw different
conclusions from the same set of facts. With
reference to the conclusions referred to in the
statement “when all of the investigative evidence is
considered, it leads to the conclusions …”, the NTSC
has a set of different conclusions based on the
evidence available.
1. No airplane-related mechanical malfunctions or failures caused or contributed to
the accident.
The investigation examined the aircraft structures, flight control systems, and
powerplants extensively, and the results are presented in the NTSC draft Final Report. As
stated in the conclusions in the draft Final Report, there was no evidence of any preimpact mechanical malfunctions or failures. Further, the pilots did not report any
problems with the airplane or make any distress calls to air traffic controllers throughout
the duration of the flight, as would be expected if they had experienced a mechanical
problem. Finally, engineering simulations of flightpath data (derived from pre-upset
DFDR data, recorded radar information, and wreckage locations) were conducted to
determine the motion of the airplane from the time it departed cruise flight until the end
of recorded data indicated. As noted in the NTSC draft Final Report, analysis of these
simulation results indicated that no single mechanical failure of the airplane structure or
flight control systems would have resulted in movement of the airplane through recorded
radar data points. Further, there was no evidence of any combination of systems failures.
Therefore, the evidence supports a conclusion that no airplane-related mechanical
malfunctions or failures caused or contributed to the accident.
NTSC’S COMMENTS:
The evidence available does not rule out that there
were no airplane-related mechanical malfunctions or
failures as only 73% of the wreckage was recovered,
most of which was highly fragmented. It is for this
reason that NTSC has not made any conclusions on this.
NTSC has taken the consistent position throughout the
final report that conclusions must be backed by
evidence.
N-6
2. The accident can be explained by intentional pilot action.
a)
The accident airplane’s flight profile is consistent with sustained
nose-down manual flight control inputs.
The engineering simulations just discussed indicated that manual manipulation of
the primary flight controls in multiple axes would result in a descent time history that was
similar to the last recorded radar points. Without the use of horizontal stabilizer trim, this
would require control column forces greater than 50 pounds and large control column
inputs; if those forces were relaxed, the airplane would have initiated a return to a noseup attitude due to its inherent stability. However, the simulations indicated that a
combination of either control column inputs and/or changing the stabilizer trim from
about 4.5 to 2.5 units nose-down trim (which would have “unloaded” the high control
forces) in conjunction with aileron inputs, would result in a descent time history similar
to that of the last recorded radar points. It is important to note that the physical evidence
indicated that the horizontal stabilizer trim was set at the maximum airplane nose-down
main electric trim limit (2.5 units) at the time of impact.
Therefore, on the basis of the engineering simulations, it is very likely from the
time it departed from cruise flight until the end of the recorded data, that the airplane was
responding to sustained flight control inputs from the cockpit.
NTSC’S COMMENTS:
There is no evidence to conclude that there was manual
intervention. After the stoppage of the FDR, which
occurred before the airplane started its descent, there
was no data available on the inputs made to flight
control surfaces or engine thrust levers up to the
point of impact.
b) The evidence suggests that the CVR was intentionally disconnected.
The NTSC draft Final Report states that no reason could be found for the stoppage
of the flight recorders and recommends that “a comprehensive review and analysis of
[FDR and CVR] systems design philosophy by undertaken . . . to identify and rectify
latent factors associated with the stoppage of the recorders in flight.”
This
recommendation implies that the NTSC believes the flight recorders stopped because of
mechanical malfunction. However, this implied conclusion is not supported by the
evidence. Rather, the evidence suggests that the CVR was intentionally disconnected.
There is also no evidence to indicate that the digital flight data recorder (DFDR) stopped
as a result of mechanical malfunction.
N-7
NTSC’S COMMENTS:
Although it may be possible that the CVR was
intentionally disconnected, there is no evidence to
suggest as such.
The evidence only showed that the
stoppage was not caused by an over load or short
circuit.
The first indication of an anomaly in the flight occurred at 09:05:15.6, when the
CVR ceased recording. As further discussed later in this document in detailed comments
on section 2.6.1 of the draft Final Report, evidence (including the sound signature at the
end of the recording) indicated that the stoppage of the CVR was consistent with the
removal of power going to the unit through activation (“pulling”) of the circuit breaker,
rather than the CVR stopping as a result of a mechanical malfunction or a short circuit or
other electrical condition.3 Further, the evidence from the last recorded minutes on the
CVR indicates that during the 4 minutes that elapsed after the last meal service and
before the recording stopped, only the captain and first officer were present in the cockpit.
The HPG determined that the captain’s statement at 0904:55, “go back for a while, finish
your plate,” indicated that he was leaving the cockpit and told the first officer to finish
eating. In addition, the CVR also recorded sounds that were consistent with seat
movement and removal of a seat belt just before the captain offered the first officer water
at 0905:00. This sequence of events is consistent with the captain preparing to leave the
cockpit.
The circuit breaker panel located directly behind the captain’s seat contains the
circuit breakers for both the CVR and FDR. It was determined that the cockpit door did
not open before the CVR ceased recording at 0905:15.6, thus it is evident that the captain
would have been in the best position to manually pull the CVR circuit breaker at the time
that it stopped. (It should be noted that the captain had pulled a CVR circuit breaker on a
previous occasion.4)
The DFDR stopped recording approximately 6 minutes after the CVR stopped
recording. There was no evidence of any malfunction of the DFDR until the moment it
stopped recording. Examination of other aircraft systems and the review of the air traffic
control radar tapes revealed that the DFDR is powered through the same electrical bus
(Electronics Bus 1) as ATC-1 (one of the airplane’s two radar transponders) and the
Mach trim actuator. The radar transponder (which was likely ATC-1 during the accident
flight)5 continued to operate and return data for a short time after the DFDR stopped. In
3
The DFDR gave no indication of any other electrical problems associated with the cessation of the CVR
or electrical problems preceding the subsequent cessation of the DFDR.
4
The NTSC draft Final Report acknowledges this incident in section 2.14.3, which describes the incident as
follows: “for non-technical reasons the PIC infringed a standard operating procedure, i.e., with the intention
to preserve a conversation between the PIC and his copilot, the PIC pulled out the CVR circuit breaker, but
the PIC reset the circuit breaker in its original position before the flight.”
5
The radar transponder can also be powered through ATC-2, which operates off of Electronics Bus 2.
However, it was reported to NTSB investigators by SilkAir instructor pilots that SilkAir operating
procedure (as documented in company procedures appended to the 737 operating manual) called for the
N-8
addition, the Mach trim actuator was found at its high speed (not cruise speed) setting,
indicating that it was powered and operational during the airplane’s high-speed dive. It
can be concluded that the absence of a malfunction of the DFDR up to the point at which
it stopped, combined with the fact that the transponder continued to transmit and the
Mach trim actuator continued to operate after the DFDR had stopped, indicates that the
stoppage was not due to a loss of power to Electronics Bus 1. However, the stoppage
could be explained by someone manually pulling the circuit breaker.
NTSC’S COMMENTS:
As mentioned above, although it may be possible that
the CVR was intentionally disconnected, there is no
evidence to suggest as such.
The NTSC draft Final Report suggests that the cessation of the CVR and DFDR
could in each case be explained by a broken wire. Although this is technically correct,
the probability of two such unrelated wire breaks occurring several minutes apart and
affecting only the CVR and DFDR is so highly improbable that it cannot be considered a
realistic possibility.
NTSC’S COMMENTS:
•
•
The draft final report stated that “A break in the
wire supplying power to the CVR could also lead to
CVR stoppage without any sound being recorded.
However, from the limited quantity of wiring
recovered, it could not be determined if a break in
the wiring had caused the CVR to stop.” The report
did not conclude or rule out that the stoppage of
the CVR was due to a broken wire.
The draft final report did not make any statement
linking FDR stoppage with wire breakage as implied
by NTSB.
pilots to select ATC-1 when the captain was flying and ATC-2 when the first officer was flying unless one
was inoperative. Because the captain was flying during the accident flight, pursuant to this procedure,
ATC-1 would have been selected.
N-9
c) Recovery of the airplane was possible but not attempted.
The NTSC draft Final Report contains a recommendation that flight crews be
trained in “recovery from high speed flight upsets beyond the normal flight envelope…to
enhance pilot awareness on the possibility of unexpected hazardous flight situations.”
This recommendation implies that the NTSC has concluded that the accident may have
been caused by an unexpected unusual flight upset and that the flight crew was not
properly trained to recover from such an upset. However, such a conclusion is not
supported by the evidence.
NTSC’S COMMENTS:
The evidence showed that the airplane had exceeded its
flight envelope during its high-speed transonic
descent. The recommendation does not in any way imply
that NTSC has concluded that the accident may have
been caused by an unexpected unusual flight upset.
This recommendation is to generate crew awareness of
the narrow margin between the normal high-speed flight
regime and the limits of the flight envelope, and the
hazards of exceeding the normal flight envelope.
Regardless of the reason for the airplane’s departure from cruise flight, it could
have been easily recovered using conventional techniques that both pilots had received
training for and that were within the capabilities of both pilots. Further, there was ample
time for the pilots to take such corrective action to return the airplane to a straight and
level attitude and flight. Both pilots had training in unusual attitudes, and the captain was
an accomplished fighter pilot adept at aerobatic maneuvers as evidenced by his
membership on the RSAF flight demonstration team, the “Black Knights.” It is apparent
that, had the pilot attempted to recover by initiating immediate corrective action using
standard flight control inputs and techniques, the airplane would have recovered to a
straight and level attitude with a minimum loss of altitude.
NTSC COMMENTS:
•
•
•
The NTSB statement is based on the assumption that the
recovery was possible if recovery action was taken
immediately (eg JAR ACJ 25.1329 gives a norm of 4
secs). Without such immediate recovery action the
airplane could exceed the normal flight envelope.
Airframe manufacturers have not specified to airline
crew recovery techniques for flight outside the normal
envelope.
There was evidence indicating that the Captain had the
intention to leave the cockpit.
There was no evidence that the recovery of the airplane
was not attempted.
The fact that during impact the
engine was in high power and that the undercarriage was
N-10
retracted do not preclude the possibility of pilot
attempts to recover from the steep dive by using flight
controls and varying engine thrust.
As previously mentioned, the simulations indicate that from the time it departed
cruise flight, a sustained nose-down flight control input was necessary to maneuver the
airplane through the recorded radar points. Additionally, the impact damage to the engine
was consistent with a higher-than-cruise power setting. (Without pilot input, the
autothrottle system would have reduced engine power to idle when the descent began;
therefore, the high power setting must have been input by the pilot.) Further, there was
no evidence that any other measures were taken (such as deploying aerodynamic drag
devices on the airplane) to slow the airplane’s speed. The wing leading-edge devices and
trailing-edge flaps, the “speed brakes” (in-flight spoilers) and the landing gear were found
to have been in a position that was consistent with cruise flight.
NTSC’S COMMENTS:
There was no FDR recording to provide any data as to
what actually happened from the time the recording
stopped until impact.
The simulation results, in combination with the physical evidence of a high engine
power setting, a horizontal stabilizer trim setting positioned for maximum nose-down
attitude, and the absence of any indication of an attempt to reduce the airplane’s speed,
are clearly inconsistent with an attempt to a recover from a dive and return to cruise
flight, and strongly suggest the maneuver was intentional.
NTSC’S RESPONSE:
As mentioned above, there was no FDR recording to
provide any data as to what actually happened from the
time the recording stopped until impact.
d) It is more likely that the nose-down flight control inputs were made by
the captain than by the first officer.
The HPG evaluated the professional, personal, and financial backgrounds of the
flight crew of flight MI185. The HPG findings are discussed in more detail in comments
on individual sections of the draft Final Report later in this document. In summary, the
HPG investigation revealed that both pilots were trained in accordance with applicable
company and civil aviation authority regulations and were competent to promptly
recognize, address, and manage an unanticipated in-flight situation using all resources
available to them; there was no evidence to indicate that the performance of either pilot
was adversely affected by any medical or physiological condition existing before the
accident; there was no evidence to indicate that there were any difficulties in the
relationship between the two pilots before or during the accident flight; and there was no
N-11
evidence that either pilot was experiencing any significant difficulties in personal
relationships involving family and friends.
Further, with respect to the first officer, the evidence developed by the HPG
revealed the first officer was not experiencing any professional setbacks or difficulties at
the time of the accident, nor was he experiencing any financial difficulties. Also, there
was no evidence that he was experiencing any behavioral changes before the accident.
However, the investigation of the captain’s background developed evidence that
revealed he had experienced multiple work-related difficulties, particularly during the
6-month period before the accident. Additionally, the investigation found that the captain
was experiencing significant financial difficulties about the time of the accident, and there
were indications that the captain's behavior and lifestyle had changed before the accident.
NTSC’S COMMENTS:
The information obtained after HPG report version 6.0
showed that the captain had no significant financial
difficulties at the time of the accident.
It is not possible to determine with certainty which pilot made the manual flight
control inputs. However, when the HPG findings are considered in the context of all the
other investigative findings, they lead to the conclusion that the airplane departed cruise
flight as a result of an intentional maneuver requiring sustained manual flight control
inputs that were most likely performed by the captain.
NTSC’S COMMENTS:
There is no evidence to conclude that either of the
pilots made any manual flight control inputs.
In summary, the investigative findings strongly support the conclusions that no
airplane-related mechanical malfunctions or failures caused or contributed to the accident,
and the accident can be explained by intentional pilot action.
The remainder of this document sets forth detailed comments on individual
sections in the draft Final Report.
N-12
1.
1.1
FACTUAL INFORMATION
History of Flight
Sequence of Events
On 19 December 1997, a SilkAir Boeing B737-300 aircraft, registration 9V-TRF, was on
a scheduled commercial international passenger flight under Instrument Flight Rules
(IFR), routing Singapore – Jakarta – Singapore.
The flight from Singapore to Jakarta operated normally. After completing a normal turnaround in Jakarta the aircraft departed Soekarno-Hatta International Airport for the
return leg.
At 08:37:13 (15:37:13 local time) the flight (MI 185) took off from Runway 25R with the
Captain as the handling pilot. The flight received clearance to climb to 35,000 feet
(Flight Level 350) and to head directly to Palembang6. At 08:47:23 the aircraft passed
FL245. Ten seconds later, the crew requested permission to proceed directly to PARDI7.
The air traffic controller instructed MI 185 to standby, to continue flying directly to
Palembang and to report when reaching FL350. At 08:53:17, MI 185 reported reaching
FL350. Subsequently, the controller cleared MI 185 to proceed directly to PARDI and to
report when abeam Palembang.
At 09:05:15.6, the cockpit voice recorder (CVR) ceased recording. According to the
Jakarta ATC transcript, at 09:10:18 the controller informed MI 185 that it was abeam
Palembang. The controller instructed the aircraft to maintain FL350 and to contact
Singapore Control when at PARDI. The crew acknowledged this call at 09:10:26. There
were no further voice transmissions from MI 185. The last readable data from the flight
data recorder (FDR) was at 09:11:27.4. Jakarta ATC radar recording showed that MI
185 was still at FL350 at 09:12:09. The next radar return, eight seconds later, indicated
that MI 185 was 400 feet below FL350 and a rapid descent followed. The last recorded
radar data at 09:12:41 showed the aircraft at FL195. The empennage of the aircraft
subsequently broke up in flight and the aircraft crashed into the Musi River delta, about
28 kilometres north east of Palembang. The accident occurred in daylight and in good
weather condition.
The route map and the crash site are depicted in Figures 1.a to c. The sequence of events
is shown schematically in Figure 2.
6
Coordinates (02.52.7S, 104.39.2E)
Air Traffic Control reporting point (00.34.0S, 104.13.0E) north of Palembang in the Jakarta FIR near
the boundary with the Singapore FIR. At PARDI, flights are transferred over to Singapore ATC
7
N-13
Section 1.1 in the NTSC’s draft Final Report does not present all factual
information necessary to portray a more complete picture of the flight
crewmembers’ interaction in the cockpit shortly after departure from Jakarta. It is
strongly suggested that this section be revised to include information from the HPG
Final report (dated June 8, 2000), which provides a time history of the captain’s
movements in the cockpit in relation to the time that the CVR stopped recording.
Once this history is established, it provides a basis for analyzing the stoppage of the
CVR and, possibly, the DFDR. The following text should be inserted immediately
before the sentence, "At 09:05:15.6, the cockpit voice recorder (CVR) ceased
recording":
At 0904:55, the PIC said “go back for a while, finish your plate.”
The co-pilot responded “I am.” A series of metallic snaps started
immediately prior to 0905:00, when the PIC said “some water.”
The co-pilot replied “no thanks.”
Further, the CVR transcript in Appendix A is not a complete factual record
of CVR recording that was transcribed by the CVR group. Appendix A does not
include the conversation that took place during preflight activities or the
conversations that transpired between the flight crewmembers or between the flight
crew and the cabin crew while the airplane was on the ground in Jakarta. This
information is critical to the analysis of crewmembers’ overall discipline and of
comments that were made in conversation by the crewmembers. Seen in its entirety,
the CVR transcript indicates that a cordial and professional atmosphere existed on
the flight deck during the period of time the CVR was operating. Moreover, the
publication of the entire CVR transcript is necessary to maintain consistency with
the ATC communications transcripts.
1.5
Personnel Information
1.5.1
Pilot-In-Command (PIC)
Sex
Age
Date of joining SilkAir
Licence country of issue
Licence type
Licence number
Validity period of licence
Ratings
Medical certificate
Aeronautical experience
Experience on type
Male
41 years
1 March 1992
Singapore
ATPL (Airline Transport Pilot Licence)
501923
1 November 1997 to 30 April 1998
Boeing B737; Airbus A310 (not current)
First class – issued 10 October 1997
7173.3 hours
3614.7 hours
N-14
Last 24 hours
Last 7 days
Last 28 days
Last 90 days
Last line check
Last proficiency check
Instrument rating check
1.5.2
1.6 hours
20.1 hours
56.8 hours
216.7 hours
25 January 1997
7 August 1997
7 August 1997
First Officer (F/O)
Sex
Age
Date of joining SilkAir
Licence country of issue
Licence type
Licence number
Validity period of licence
Ratings
Medical certificate
Aeronautical experience
Experience on type
Last 24 hours
Last 7 days
Last 28 days
Last 90 days
Last line check
Last proficiency check
Instrument rating check
Male
23 years
16 September 1996
Singapore
CPL (Commercial Pilot Licence)
503669
1 July 1997 to 30 June 1998
Boeing B737
First class – Issued 4 June 1997
2501.7 hours
2311.8 hours
1.6 hours
21.4 hours
69.8 hours
217.6 hours
10 October 1997
15 September 1997
15 September 1997
Sections 1.5.1 and 1.5.2 do not contain any information about the captain’s
personal background. Although this is not a typical subheading for an accident
report, the high probability of flight crew involvement in this accident makes it
necessary to include this information to complete the factual record and provide the
basis for a thorough analysis.
It is strongly suggested that the NTSC add a section pertinent to the captain’s
personal background information to the draft Final Report. This section should be
identified as “1.5.1.1 Personal Background” and include the following information
about the captain:
•
The captain was born June 17, 1956 in Singapore. His parents were Chinese
immigrants, and he was the second of four children. He was of Buddhist
faith, but he was not reported to be devoutly religious.
N-15
•
He achieved his “O” level education in 1972 and an industrial technician
certificate in electronics in 1974.
•
The captain was married in July 1979, and his wife, also of Chinese descent,
was born in Singapore. The captain and his wife had three sons, born in
1981, 1983, and 1989. In HPG interviews during the investigation, the
captain was described as being a family man, who often spoke about his sons
and spent a lot of time with them.
•
The captain’s family moved to a new house in August 1997 next to where the
captain’s brother and parents lived. He was reported to be interested in
computers and financial markets.
•
According to police representatives, he had no record of criminal activity in
Singapore.
It is also strongly suggested that the NTSC add a section pertinent to the first
officer’s personal background information. This section should be identified as
“1.5.2.1 Personal Background” and include the following information:
•
The first officer was born in New Zealand on March 3, 1974. He was the
second of four children. The first officer was not married but had a close
personal relationship with a stewardess who worked at SilkAir. The first
officer lived with another SilkAir first officer, a close friend who he also lived
with when working for Garuda.
•
The first officer was a Christian, and was described as being devout. He was
described as being close to his family and had a number of close friends in
Singapore. He was described as being very interested in flying and pursuing
a flying career. His other interests were reported to include traveling,
spending time with friends, and playing sport.
•
Police representatives reported that the first officer had no record of
criminal activity in Singapore.
1.16
Tests and Research
1.16.1 CVR Circuit Breaker Actuation Test
Upon the completion of data readout by NTSB, the CVR was taken to AlliedSignal
on 22 January 1998 for further testing. This testing was an attempt to verify if the
termination of the CVR recording was due to loss of power by the pulling of the CVR
circuit breaker or other means. The result was inconclusive. Therefore other tests had to
be performed, see Appendix F.
N-16
There were three tests conducted in a B737-300 aircraft to investigate the CVR
circuit breaker actuation sound signature.
The first test
The first test was carried out on the ground by NTSB and Boeing on 5 February 1998.
The reason for this test was to have quiet ambient condition to provide the best
opportunity for detection of circuit breaker actuation sound signature. The result showed
that the CVR cockpit area microphone did record the CVR circuit breaker actuation.
Actuation of a circuit breaker nearby gave a similar result.
The second test
The test (consisting of on-ground and in-flight tests) was conducted on 14 May 1998 and
15 May 1998 by NTSB.
The purpose of the ground test was to obtain an on-plane, on-ground CVR
recording of the CVR circuit breaker actuation, and the purpose of the flight test was to
obtain an on-plane, in-flight CVR recording of the CVR circuit breaker opening. In both
tests the circuit breaker was actuated manually and through the introduction of faults to
the aircraft’s wiring, i.e. short circuit and overload.
The results of these tests were compared with the accident CVR recording sound
signatures. In the short circuit tests a distinctive 400 Hz tone is recorded on one or more
of the CVR channels. No corresponding signatures could be identified on the accident
recording. The same tests found that the area microphone is able to pick up a distinctive
and identifiable snap sound that the circuit breaker makes when it is violently tripped by
a short circuit. (Note: The CVR continues to run for 250 milliseconds before it runs out of
power from the capacitor. As sounds travel about one foot per millisecond, it would take
only six milliseconds to travel the approximately six feet distance from the circuit breaker
to the area microphone. Hence the CVR is able to record the snap sound of the circuit
breaker.)
The overload tests yielded similar results as the short circuit tests except that
there was a slight time delay for the circuit breaker to trip and the snap sound was
quieter but still identifiable. No corresponding sound signatures could be found in the
accident recording.
The last set of tests was to examine the sound signatures when the CVR circuit
breaker was manually pulled. The snap sound was identifiable on the ground without
engines and air-conditioning operating. However in the flight tests, the addition of the
background cockpit noise present during normal cruise obscures the sounds associated
with the manual in-flight pulling of the cockpit circuit breaker. No corresponding sound
signatures could be found in the accident recording.
N-17
The summaries of the results of the second tests are as follows:
•
During an overload and a short circuit, the sound of the circuit breaker popping is
loud enough to be identified on the CVR‘s area microphone channel, both on the
ground and in-flight.
•
During an overload and a short circuit, the CVR records unique and identifiable
sound signature on one or more of the channels, both on the ground and in-flight.
•
During the manual pull test on the ground, the sound of the circuit breaker is loud
enough to be identified on the CVR recording.
•
In cruise flight, normal cockpit background noise obscured the manual circuit
breaker pull sounds. There are no unique electronic identifying sound signature
recorded on the CVR.
The third test
The test was conducted in-flight using a B-737 SilkAir sister aircraft in Singapore on 16
October 1998 and supervised by the Indonesian NTSC, an FAA avionic inspector
(representing NTSB) and Singapore MCIT representatives.
In the third test, several scenarios were performed where the CVR circuit breaker in the
cockpit was manually pulled. The manual pulls were categorized as “soft”, “hard” and
“string” pull. The soft pull was by pulling the circuit breaker with minimum noise. The
hard pull was by pulling the circuit breaker normally. The string pull was by pulling on a
string that was attached to the circuit breaker. This was to simulate a short circuit
causing the circuit breaker to pop out.
All the tests were conducted with an identical AlliedSignal SSCVR 2-hours recorder as
installed in the accident aircraft.
All four channels of the CVR recordings of the above three tests were analyzed using the
same NTSB signal processing software that was used to analyze the accident CVR
recording.
Several tests were done to document the sound that were recorded on the CVR during a
soft, hard and string pull of the CVR circuit breaker. The test closely matched the data
obtained from the second test (NTSB in-flight test above).
The NTSC draft Final Report’s discussion about the CVR (and DFDR later
in the report) requires a correction of terms so as to avoid confusion when
referencing the electrical power that is being supplied to both recorders. The
“power source” is where the CVR and DFDR receive their respective electrical
power from within the aircraft whereas the “power supply” is a component that is
integral to the CVR and DFDR units. The use of “power supply” when describing a
power interruption from the aircraft power source is incorrect because it implies
that an internal failure occurred or may have occurred within a respective recorder.
N-18
To eliminate any potential confusion regarding the electrical power being supplied
to the CVR and DFDR, the draft Final Report should be reviewed and the term
“electrical power source” inserted where the term “power supply” is currently used.
NTSC’S COMMENTS:
The term “electrical power source” is meant to include
the external as well as the internal power supply to
the individual units.
In addition, examination and testing requested by the NTSC revealed no
evidence to suggest that a mechanical malfunction or failure of either the CVR or
DFDR caused either recorder to stop recording data.
The discussion of the postaccident CVR testing does not address the
distinctive 400 Hz tone (or “hum”) that was recorded on the CVR tape. For
purposes of clarity, the following information (an excerpt summary from the
February 20, 1998, Boeing test report) should be added to this section to discuss the
basis for the 400 Hz tone:
Power line hum components, located at multiples of 400 HZ were evident
in the CVR area microphone signal. The level of some of these hum
components increased when the overload was applied until the circuit
breaker popped. These levels were even more dramatic during the
ground fault conditions….It should be noted that some of these hum
components persisted in the signal, after the breaker has opened, to the
very end of the recorded data. Tones are good candidates for detection
amongst broadband boundary layer excitation. Tracking the amplitude
of a particular power-line hum component may indicate circuit overload
or faults.
NTSC’S COMMENTS:
It has been adequately covered in the report that
failure of the CVR is not due to an overload or a short
circuit condition.
1.18.3 PIC’s Background and Training
1.18.3.1 Professional Background in RSAF
The PIC joined the Republic of Singapore Air Force (RSAF) as a pilot trainee on 14 July
1975. He obtained his ‘wings’ (fully operational) on 25 March 1977. During his RSAF
career, the PIC flew many different types of fighter and training aircraft. He held senior
flying and instructing positions. In 1970s, the PIC was selected to joint the RSAF’s Black
Knights aerobatic team. He reached the rank of Captain in 1980 and was promoted to
N-19
Major in 1989. In 1991, the PIC applied to leave the RSAF under a voluntary early
release scheme. The PIC met the eligibility requirements for the early release scheme as
he was 35 years old and had at least six years in his immediate preceding rank. His
application was accepted.
The PIC’s reason to leave RSAF and join SilkAir was to keep flying and to spend more
time with his family.
The PIC obtained a US Federal Aviation Administration (FAA) Commercial Pilot
Licence on 19 November 1991 and an Air Transport Pilot Licence on 26 November 1991
in Benton Kansas. He left full-time employment in the RSAF on 29 February 1992. He
had approximately 4,100 hours flying experience at that time. The PIC served as a
squadron pilot in the RSAF on a part-time basis from 1 March 1992 to 30 April 1993. He
subsequently served in the military reserve, as a Major, in a non-flying capacity. In
January 1997, the PIC was promoted to Deputy Director Air Liaison Officer in his
reserve unit.
The NTSC draft Final Report’s description of the captain’s “Professional
Background in RSAF” should be expanded to include information developed by the
HPG regarding his military service or significant events that occurred during his
service period. It is strongly recommended that section 1.18.3.1 be revised to
include the following information:
The captain joined the RSAF as a pilot trainee on July 14, 1975. He
obtained his “wings” (fully operational) on March 25, 1977. During
his RSAF career, the captain flew many different types military
fighter and training aircraft. He held senior flying and instructing
positions and reached the rank of captain (military) in 1980 and was
promoted to the rank of major in 1989. The captain became a
member of the RSAF Black Knights in 1990 [not 1970s as stated in the
NTSC draft Final Report].
NTSC’S COMMENTS:
Factually correct however it is felt that there is no
need to include such specific details in the final
report.
In December 1979, when five RSAF pilots (including the captain) and
four T/A-4S aircraft were temporarily stationed in the Philippines for
training, the captain experienced the first of three significant events
during his service in the military. On December 19, the captain was
forced to withdraw from a scheduled training mission because of a
mechanical problem with his aircraft. The other three aircraft
continued with the training mission and collided with terrain after
encountering bad weather in a mountainous area. All of the pilots on
N-20
board the aircraft were killed. Although the HPG investigation
obtained information about this event, it was not possible for the
group to determine the extent to which the captain had been affected
by this event.
On September 6, 1981, the captain took off in an SF 260 training
aircraft as the instructor pilot with a student pilot conducting the
flying duties. During the takeoff roll, the aircraft crashed. The
student was fatally injured. The RSAF investigation found the
circumstances of the accident were not the responsibility of the
captain.
On March 3, 1986, the T/A-4S in which the captain was acting as an
instructor pilot during a training flight experienced loss of control
because of a mechanical malfunction. Both crewmembers ejected
safely from the aircraft. The RSAF investigation found the captain
was not responsible for the event.
During HPG interviews, RSAF personnel described the captain as
being a highly skilled pilot. According to the RSAF, there were no
records of the captain receiving disciplinary action or having any
major setbacks in his career. It was also reported that the captain did
not have a history of disagreements with other personnel while
serving in the RSAF. However, several pilots who worked with the
captain at SilkAir and who were in the RSAF at the same time as the
captain reported that he had disagreements with his commander
while serving with the Black Knights. The RSAF reported that these
disagreements were related to flying maneuvers. The disagreements
were characterized as minor and were reported to have been resolved.
It was reported that the squadron was under a great deal of pressure
and that there were disagreements between many pilots.
In 1991, the captain applied to voluntarily leave the RSAF under an
early release program. The captain met the eligibility requirements
for the early release (he was 35 years old and had at least 6 years in
his immediate preceding rank), and his application was accepted
shortly thereafter.
The captain obtained a United States Federal Aviation Administration
(FAA) Commercial Pilot Certificate on November 19, 1991, and an
Airline Transport Pilot Certificate on November 26, 1991, in Benton
Kansas.
He left the full-time employment of the RSAF on
February 29, 1992, at which time he had accumulated approximately
4,100 hours flying experience.
N-21
The HPG investigation found during an interview with the captain’s
wife that he separated from the RSAF because he wanted to spend
more time with his family. A person considered to be a flying
associate of the captain reported that the captain probably left the
RSAF in order to remain in a flying position because his next job
would have likely been in a nonflying capacity if he remained enlisted.
Under the early release scheme, the captain had the choice of joining
either Singapore Airlines (SIA) or SilkAir. Several SilkAir personnel
who knew the captain reported that he preferred to join SilkAir
rather than SIA because he could be promoted to a command pilot
position at SilkAir within 3 years whereas he would likely have to wait
at least 5 years to achieve the same position at SIA.
Although the captain was flying full-time for SilkAir in March 1992,
he served as a squadron pilot in the RSAF on a part-time basis from
March 1, 1992, to April 30, 1993. He subsequently served in the
military reserve, as a major, in a nonflying capacity and in January
1997 was promoted to Deputy Director Air Liaison Officer in his
reserve unit.
NTSC’S COMMENTS:
The above
accident.
details
are
deemed
not
relevant
to
the
1.18.3.2 Professional Background with SilkAir
The PIC formally joined SilkAir on 1 March 1992. He was initially employed as a Cadet
pilot under a training program for pilots that did not have a Boeing 737 (B737) type
rating and had no previous airline experience.
The PIC was assigned to the Airbus A310 fleet and commenced training on 30 May 1994.
He was appointed First Officer on the aircraft on 15 August 1994. When SilkAir
discontinued A310 operations, the PIC was re-qualified on the B737 in March 1995.
The PIC was selected for B737 command training on 22 October 1995. He was
appointed Captain on 26 January 1996, and confirmed in that position on 27 July 1996.
He was selected as LIP in March 19978 and completed his training on 9 May 1997. He
performed satisfactorily thereafter in this position. On 3 July 1997, the PIC was deappointed from his LIP position following an investigation into an operational incident
which occurred on 24 June (see Appendix I for details).
8 The LIP position was seen as a requirement for further promotion to instructor pilot or into management.
The position also gave a pilot additional allowance of S$ 750 monthly.
N-22
The PIC had no problems with regard to his professional licence medical requirements.
His last licence renewal medical examination was on 2 December 1997.
Section 1.18.3.2 does not provide a complete factual record regarding the
captain’s career at SilkAir nor does the referenced Appendix I (the chronology
presented in Appendix I does not list any professional events that occurred after
August 19, 1997). The addition of a complete career chronology is necessary to
accomplish a thorough analysis of the accident. Thus, it is strongly recommended
that the NTSC include the following information (from version 6.0 of the HPG
report) in section 1.18.3.2 to present a complete chronology of the captain’s airline
history:
The captain was formally employed with SilkAir on March 1, 1992, as
a “cadet” pilot under a special training program for pilots who did
not have a Boeing 737 (B737) type rating and had no previous airline
experience. The captain signed a 7-year training bond agreement
with SilkAir (then known as Tradewinds) that required the captain to
pay the company on a prorated schedule for his training if he should
leave the employ of SilkAir for any reason.
After completing B737 ground school, the simulator checks, and base
checks, the captain was appointed as a second officer on June 26,
1992. At SilkAir, the second officer position is typically identified
with a copilot who requires a mandatory period of supervision before
promotion to first officer. It was a standard appointment for someone
with the captain’s background at the time he joined the airline. He
was appointed as a first officer for a 6-month probation period on
October 14, 1992, commenced line operations in that position, and
completed his probation period on April 14, 1993.
The captain was selected for a conversion to the Airbus 310 (A310)
fleet on April 14, 1994, and commenced training activities on May 30,
1994. He was appointed as a first officer on the aircraft on August 15,
1994. When SilkAir phased out its A310 fleet, the captain was eligible
to leave SilkAir and join SIA. However, he decided to remain with
SilkAir and subsequently completed B737 reactivation training in
March 1995.
After obtaining sufficient flight time to convert his FAA issued Airline
Transport Pilot License (ATPL) to a Singaporean license, the captain
took a written examination (Special Assessment Paper for a Foreign
License Conversion) on July 25, 1995. This examination was
administered by the United Kingdom (UK) CAA under a contract
with the Singaporean CAAS. Candidates are only given one
opportunity to take this examination for which a passing score was 75
N-23
percent. The captain initially scored 68 percent but a subsequent
rescoring elevated his score to 71 percent. On August 1, 1995, the
captain made a special appeal to the UK CAA, citing that there were
“imperfections” in the exam paper. His appeal was accepted and he
was given a passing grade. He applied for a Singapore ATPL on
October 11, 1995, and received a letter 7 days later from the SilkAir
B737 Fleet Manager congratulating him on achieving his Singapore
ATPL and 3 years of service at SilkAir.
The captain was selected for B737 command training on October 22,
1995,9 and commenced training for this position on January 22, 1996.
This training included both simulator and line training, with the
simulator training consisting of five line-oriented flight training
(LOFT) sessions. The captain signed a 3-year bond agreement with
SilkAir for the training. He was appointed to captain on January 26,
1996, and confirmed (after probation) in that position on July 27,
1996.
In March 1997, three management pilots in SilkAir10 met to discuss
the captain’s suitability for a line instructor pilot (LIP) position.11
Although the captain met the minimum requirements for the position,
the managers initially had reservations about the captain’s suitability
as they thought he may have been too reserved, regimented, or strict.
After interviewing the captain, they selected him for the position and
he completed the required training between April 30, and May 9,
1997. He also conducted line operations as a LIP between May 10 and
June 13, 1997. There were no problems reported concerning the
captain’s LIP performance.
During his career at SilkAir, the captain received training in unusual
attitudes, flight control malfunctions, and flight instrument
malfunctions on several occasions. No significant problems were
noted regarding the captain’s abilities to accomplish this training.
During his last training session in these areas (February 17 to 18,
1997), his performance was rated as “sound.” In May 1997, the
captain participated in the Aircrew Resource Management course
conducted within SIA for aircrews.
9
Pilots are not eligible for command selection in SilkAir before they have served 3 years in the
company, flown a total of 4,400 hours, and completed 300 sectors as pilot flying.
10
These management pilots were the Flight Operations Manager, B737 Fleet Manager, and the F-70
Fleet Manager. The B737 fleet manager was the PIC’s immediate supervisor.
11
In SilkAir, a captain is eligible to be considered for an LIP position after serving as a captain for 1
year. In addition to regular duties as commander, LIPs at SilkAir give instruction and guidance to
command candidates or first officers in the final stages of their training. At SilkAir, the LIP position
was considered a requirement for further promotion to instructor pilot or into management. The
position also gave a pilot additional pay of S$750 monthly.
N-24
During 1992 to 1996, the captain was not involved in any known
operational incidents.
However, he was involved in three
nonoperational incidents that SilkAir management described as
minor and as having no effect on the captain’s chances of career
progression. These events involved missing a security briefing,
forgetting his passport for an international flight, and attempting to
cash a cheque at an inappropriate facility.
Between January 1997 and the time of the accident, the captain was
involved in the following four operational events:
•
On March 3, 1997, a go-around was performed on an approach
into Manado, Indonesia;
•
On May 17, 1997, a flight was conducted with a dispatch
authorization for an inoperative parking brake;
•
On June 24, 1997, the captain pulled, and then reset, the CVR
circuit breaker before a flight; and
•
On November 20, 1997, an overweight landing occurred in
Singapore.
As a result of the CVR circuit breaker incident, the captain was
removed from his LIP position on July 3, 1997, after a Divisional
Inquiry.
Although he subsequently appealed this decision, a
Company Inquiry upheld the original decision. During the accident
investigation, the HPG found through interviews with several SilkAir
personnel that the captain was upset by the events that resulted in the
loss of his LIP position. The events surrounding the four incidents
are summarized in Appendix I of the draft Final Report (Appendix
H1 of the HPG version 6.0 report).
During interviews conducted by the HPG, SilkAir instructors
reported that the captain’s transition from military fighter aircraft to
a commercial airliner was average but that his performance soon
improved with experience. He adapted quickly and generally had no
problems with any period of training. Further, SilkAir evaluates each
pilot’s operational performance every 6 months. These base checks
are performed in a simulator, and the pilot’s performance is rated in a
number of key performance areas. The captain’s performance on
almost all base checks from March 1993 to his last check on August 7,
1997, was rated as “above average.” His performance on the annual
line checks was consistent with his performance on base checks.
N-25
Interviews conducted by the HPG with instructors and other pilots at
SilkAir indicated that the captain’s ability as a pilot was above
average and that he was a competent operator. He was commonly
described as being very highly skilled in handling an aircraft and as
having fast reaction times. It was reported that he appeared to
maintain good situation awareness and that he made decisions
quickly, firmly, and confidently. He also appeared to be able to handle
any flight-related pressures quite calmly.
The captain was described by other SilkAir pilots as a good cockpit
manager. They indicated that he provided clear instructions, kept
first officers informed of his decisions, and let first officers make their
own decisions. He was quick to spot copilot mistakes or problems with
the flight. However, he was not known for criticizing copilots, and
liked to teach and show them new ways of doing things.
The HPG interviews of SilkAir pilots also revealed the captain to be
someone who operated “by the book” and who would not exceed
aircraft limitations. However, many pilots also indicated the captain
would vary from normal practices at times. For example, several
copilots reported the captain would often push the aircraft’s speed
beyond the economy speed of .74 mach.12 There were also several
reports of the captain performing higher than normal approaches13
and, unlike other company captains, exploring ways of varying
normal practices to reduce flight durations. Although the captain was
not generally regarded as an unsafe pilot, he was regarded as a pilot
who made his own decisions as to what was and was not safe.
Over a period of 2 years, the captain received several letters of
appreciation from the SilkAir B737 Fleet Manager for being on
standby. Such letters were standard for such events.
In addition, it is strongly suggested that the following events, which were
presented in the HPG report (version 6.0) be added to Appendix I:
24 August: The captain and the first officer who had been involved in
the go-around incident in March and the circuit
breaker incident in June flew together as a result of a
12
Economy speed is a speed that is supposed to optimize the relationship between time enroute and
fuel burn. It is a speed calculated to reduce the cost of operating the flight for the airline considering
several economic factors and not a manufacturer-imposed operating limitation on the aircraft.
13
A high approach means that during the descent from cruise, an airplane is at an altitude higher
than the typical altitude for a given distance from an airport.
N-26
roster change that the captain had requested for
personal reasons.
20 November: The captain was involved in an event that resulted in
an overweight landing. There was an engine power
problem noted during takeoff and climbout. After
discussion with an engineer on board, the flight crew
decided to return to Singapore instead of continuing to
the destination airport. The subsequent overweight
landing was not noted in the voyage report and
technical log as was specifically required by company
procedures. The crew also did not complete other
paperwork associated with the flight including
calculating the landing speeds, flight time, and fuel
flight plan.
The B737 Fleet Manager noted the
discrepancies and sent both pilots letters the next day,
instructing them to "please be more mindful."
10 December: The captain was flying with close friends, one was the
first officer and the other was riding in the jump seat.
During the flight, the captain complained about the
B737 Fleet Manager and the letter he had received
about the overweight landing incident on November 20.
A variety of other topics were discussed (see
recommended additions to section “1.18.3.4 Recent
Behaviour”).
11 December: The captain visited the Flight Operations Manager to
discuss the letter about the overweight landing. The
captain was concerned that he had been sent a negative
letter about a minor issue and no mention had been
made about the significant good work he had done to
return the airplane to Singapore (according to the first
officer on that flight, the captain effectively managed
the increased workload that was associated with this
air-turn back to Singapore including troubleshooting,
making the necessary notifications to ensure that
passengers’ needs were met, and moving the airplane to
the maintenance area). The Flight Operations Manager
reported that he told the captain to send him a letter
outlining his concerns and he would then consider
whether to send him a "thank you" letter. The Flight
Operations Manager did not receive a letter from the
PIC outlining his concerns.
N-27
Finally, information in Appendix I regarding certain events in the captain’s
professional history that occurred before August 19, 1997, is noticeably
condensed when compared to the discussion of the same events in the HPG
report (version 6.0). Some of these events were addressed and acknowledged
publicly by SilkAir following the AAIC interim report that was released in
August 1999. Although this information may have been pared down for the sake
of brevity, the NTSC should provide more information about the March 3, 1997,
Manado event and the June 4, 1997, CVR circuit breaker event and the
associated inquiries and appeals.
NTSC’S COMMENTS:
All group reports have been analysed and integrated
into the final report. Specific details such as those
stated above were not deemed necessary for inclusion in
the final report.
1.18.3.3 Financial Background Information
The financial background data of the PIC was gathered to determine whether financial
factors could have affected the performance of the PIC.
PricewaterhouseCoopers was appointed by the NTSC to conduct an independent review
of the preliminary findings of the NTSC's Human Factors Group concerning the financial
background of the PIC. PricewaterhouseCoopers was not involved in the investigation
itself. Based on the review, PricewaterhouseCoopers made certain recommendations to
the NTSC in order for the NTSC to refine its findings.
At the time of the accident, the PIC operated a securities trading account in Singapore.
This account was operated from June 1990 until the time of the accident. During 1990 –
1997 the PIC traded over 10 million shares, where the value and the volume of the
trading increased significantly every year. The PIC’s accumulated total losses from
share trading increased between 1993 and 1997, with moderate gains during 1997.
There was no period of the PIC negative net worth. The PIC’s trading activities was
stopped on two occasions due to the non-settlement of his debt, i.e. from 15 April to 15
August 1997 and again from 4 December 1997 until the time of the accident. On the
morning of 19 December 1997, the PIC promised the remisier to make a payment when
he returned from his flight.
The PIC had several loans and debts at the time of the accident. The PIC’s (and
immediate family’s) monthly income was calculated to be less (about 6%) than their
monthly expenditure at the time of the accident.
N-28
The probate document indicates that the PIC had a number of insurance policies which
provided benefits on the event of his death. Most of these policies were taken out many
years prior to the accident. In December 1997 he was required by the financial institution
granting the property loan to take a mortgage insurance policy. The PIC underwent
medical tests for the policy on 1 December 1997 and followed this with a formal
application of 5 December 1997. The PIC did not specify the commencement date for the
policy. On 12 December 1997 the insurance company informed the PIC that his
application was accepted pending payment of the insurance premium. A cheque dated 16
December 1997 was sent to the insurance company by the PIC being payment for the
premium. The commencement or the inception date of the policy was set by the insurance
company to be 19 December 1997. This information was not conveyed to the PIC. The
cheque was cleared on 22 December 1997.
The HPG’s examination and evaluation of the captain’s and first officer’s
overall financial status was understood to be sensitive and confidential for the
purpose of publication, and the NTSC’s decision not to present the actual financial
numbers in the report is respected. However, this section’s brevity is of concern
because it does not present the totality of information evaluated and analyzed by the
HPG.
All of the participants in the HPG (representing Indonesia, Singapore, the
United States and Australia) were involved with almost every aspect of the factgathering process regarding the captain’s financial status, which spanned a period
of more than 8 years before the accident. The HPG evaluated this aspect of the
captain’s life and in July 1999, determined through consensus of all members that at
the time of the accident, the captain was experiencing significant financial
difficulties (conclusion 14 in the HPG report version 6.0) The NTSC contracted
Price-Waterhouse-Coopers (an auditing company) to conduct an audit of financial
information that had been gathered during the investigation by the HPG and the
NTSC in the latter stages of the investigation. Although the audit was completed,
HPG members were not provided a copy for its review and evaluation. Given the
significance of this information, the report and its content should be discussed in
greater detail.
The review of section 1.18.3.3, revealed inaccuracies and exclusion of
relevant information. The NTSC draft report briefly summarized the captain’s
stock trading over a period of 7 years. Although the actual number of stock shares
that the captain traded is not relevant, the monetary value of the stocks traded, even
approximated, is significant in that it demonstrates the financial burden that he was
incurring in the later years. In addition, the NTSC draft report should contain
specific information regarding the type of trading performed by the captain (that is,
contra-trading versus buying normal shares) because it establishes that contratrading, which the captain had been conducting for approximately 8 years, is a highrisk activity. Further, a complete discussion of the captain’s trading activities would
provide a basis for explaining his loss of trading privileges on two occasions because
N-29
of nonpayment (the first time requiring a repayment plan that spanned months).
NTSC’S COMMENTS:
The information in 1.18.3.3 is accurate based on the
findings of the HPG report version 6.0 completed in
July 1999 with corrections by PricewaterhouseCoopers in
October 2000.
The NTSC draft Final Report misstates factual information developed by the
AAIC HPG as of July 1999 concerning the mortgage insurance policy that became
effective on the date of the accident. The NTSC draft Final Report states that
mortgage insurance was “required” by the mortgage lender. However, the HPG
found that mortgage insurance, which is purchased voluntarily, is generally
recommended by the lender. Also, it should be noted that the loan secured for the
purchase of this insurance had been established for at least 3 months before the
policy was issued.
It is suggested that the following revisions be made to the draft Final Report
to clarify existing information:
The mortgage insurance application was submitted November 27,
1997. The insurance was recommended and not required as part
of the loan taken by the captain in August 1997.
NTSC’S COMMENTS:
This information was updated
version 6.0 was completed.
after
the
HPG
report
The captain had two training bonds that were in effect at the time
of the accident, which required repayment had the captain
voluntarily separated from SilkAir or was dismissed.
The statement that “The PIC’s accumulated total losses from
share trading increased between 1993 and 1997, with moderate
gains during 1997,” should be clarified to indicate that the captain
had experienced net losses during this period.
NTSC’S COMMENTS:
The information is based on the findings of the HPG
report version 6.0 completed in July 1999 with
corrections by PricewaterhouseCoopers in October 2000.
In addition, the NTSC draft Final Report does not adequately present
sufficient information regarding the captain’s liquidity of assets at the time of the
accident.
N-30
Finally, the statement, “The PIC had several loans and debts at the time of
the accident” does not provide the specificity that is necessary to demonstrate the
significant debt. Also, this statement does not address the significant amount of
money that was due to be paid to the stock remiser at the time of the accident or the
fact that there were no known liquid assets from which to pay this loan. Further,
the draft Final Report does not address the credit cards debt that had been incurred
by the captain at the time of the accident. A more thorough description of this
information, even in general terms, is necessary to demonstrate that the captain’s
debts exceeded his assets. In addition, for clarity that maintains confidentiality of
specific financial amounts, the report should be modified to characterize the
magnitude of the captain’s stock losses and debts about the time of the accident in
terms of his average annual income.
NTSC’S COMMENTS:
Additional evidence not previously available to the HPG
provided a more accurate estimate of the PIC’s monthly
expenditures.
As stated in 1.18.3.3 of the final
report, the PIC had no period of negative net worth.
1.18.3.4
Recent Behavior
The PIC’s family reported that events and activities were normal in the days before the
accident. The PIC was reported to have slept and eaten normally. There were no
reported changes in his recent behavior. He was organizing his father’s birthday party
that was planned for 21 December 1997. No medical problems were reported or noted by
airline’s appointed medical clinics.
Work associates who observed the PIC on the day of the accident and on his most recent
flights, reported nothing odd or unusual in his behavior.
The NTSC draft Final Report provides very brief information about the
captain’s activities in the days preceding the accident; however, excerpted
information from the HPG report (final report and version 6.0) cites other pertinent
information that documents the captain’s behavior during the 45-day period prior
to the accident. It is suggested that NTSC draft Final Report be revised to include
the following information:
The PIC’s family reported that events and activities were normal in
the days before the accident. The PIC was reported to have slept and
eaten normally. No recent changes in his behavior were reported by
his family. It was also reported that the PIC was assisting in
preparations for his father’s birthday party, which was to occur on
December 21.
N-31
Several work associates who observed the PIC on the day of the
accident reported nothing odd or unusual in his behavior. One
associate noted that the PIC was quite reserved in the briefing room
but that he had behaved that way on some occasions.
In the month prior to the accident, the PIC conducted flights on
November 21, 22, 23, 28, 29 and 30, and December 10, 11, 12, 13 and
16. Some company personnel on these flights reported nothing
unusual or noteworthy in the PIC’s behavior. Some personnel
reported that the PIC was quieter than normal, and other personnel
reported that he complained about company management and its
maintenance of aircraft. The following specific events were recounted:
On December 10, the PIC was flying with close friends, one was the
first officer and the other was riding in the jump seat. During the
flight, the PIC complained about the B737 Fleet Manager and the
letter he received about the overweight landing incident. A variety of
other topics were also discussed. The PIC asked one of the other pilots
about a crash involving a Malaysian Airlines B737 at Gelang Patah in
1984.14 The PIC was also reported to have discussed the TWA800
B747 accident and mentioned the helplessness of a pilot in such an
accident.
The captain took leave from December 1 to 7, 1997. He applied for
this leave on November 26, which coincided with school holidays.
In addition, the NTSC draft Final Report does not include factual information
related to the behavioral characteristics of both pilots that was developed by the
HPG. The information is necessary for a complete understanding of the human
factors aspects of the investigation and critical to a complete and proper analysis
being accomplished.
Therefore, it is strongly recommended that the NTSC add a section that
describes the behavioral characteristics of the captain and first officer. The new
section for the captain should precede “section 1.18.3.4 Recent Behaviour” in the
draft Final Report (thereby making “Recent Behaviour” section 1.18.3.5) and be
identified as “1.18.3.4 Behavioral Characteristics.” This section should contain the
following information:
The HPG found during its investigation regarding the captain that the
RSAF did not require any psychological assessment during his career nor
14
The crash was determined to be the result of a hijacking. Information recorded on the CVR
assisted authorities in making this determination. This information was told to the PIC during this
flight.
N-32
was such an assessment required or conducted for his employment with
SilkAir.
HPG interviews with a variety of people who worked with the captain
revealed that he was generally a quiet and reserved person. He would not
initiate conversation with casual acquaintances but would talk if asked
questions or shared a common interest with the other person. Some
people described him as distant and difficult to get to know while other
people described him as friendly and easy-going.
The captain’s wife described him as a perfectionist and others who knew
him in the RSAF described him as a typical fighter pilot. He was also
described by people in the RSAF and SilkAir as a very assertive person;
he would state his opinion if he disagreed with some professional issue,
often in an undiplomatic manner. The interviews revealed that he was
very confident and proud of his flying skills and was very proud of his
RSAF experience. The captain was described as motivated to obtain the
best for himself and it was reported that his career was important to him.
The HPG did find through interviews that there were no reports of the
captain having any unusual or abnormal personal habits in the cockpit
during flights and that it was routine for him to leave the cockpit to use
the toilet, get a drink, or chat with the flight attendants.
The relationships between the captain and the three management pilots in
SilkAir were reported to be quite cordial. It was reported that when he
joined the company, management regarded him as a pilot who would
have no trouble reaching a command position and who had the potential
to become a management pilot. The managers were impressed with the
captain’s performance during his time as a first officer and as a captain.
HPG interviews revealed that during this period, the captain often visited
the Flight Operations Manager or the B737 Fleet Manager to offer
suggestions to improve the company’s operations. However, following
the incidents and subsequent investigations in 1997, the captain’s
relationship with the B737 Fleet Manager became less amicable.
Although there were no reported arguments, there was little interaction
between them. The captain continued to visit and communicate with the
Flight Operations Manager but this relationship became somewhat
strained after the captain lost his LIP status.
The captain was well respected by other Singaporean pilots in the
company, was one of the first two Singaporeans to join SilkAir, and was
one of the first two Singaporeans to be selected for command training. As
such, he was regarded as a natural leader to the Singaporeans who
subsequently joined the airline. The captain was one of several pilots
N-33
who were involved in efforts to improve the employment conditions of
Singaporean pilots. He was also known to defend other Singaporean
pilots or encourage them to question any unfair treatment.
During the accident investigation, the HPG received reports that the
captain did not have a good relationship with some of the expatriate
pilots in the airline. Some of the pilots stated that the captain did not
accept advice or criticism well from other captains. Further, some pilots
also stated that that the captain had been promoted to LIP over more
experienced captains who had airline instructional experience before
coming to SilkAir. The Flight Operations Manager said that the captain
was promoted because he had previous training experience in the RSAF,
had a good record, and had all the markings of a good instructor. It was
also stated that because he was Singaporean, he was preferred over
expatriates.
The captain was generally popular among flight attendants and it was
reported that he was typically easy-going and often joked with them. He
never made any special demands, and he often completed sectors faster
than other captains. Flight dispatchers also reported that the captain was
easy-going and sociable. Line engineers reported that the captain was
friendly, helped when there were engineering problems, and was quite
reasonable about accepting any defects.
The HPG also received reports during its investigation that during the latter
part of 1997, the captain criticized or complained about SilkAir management
and the B737 Fleet Manager during flights. It was reported that he appeared to
be upset about the inquiries that had taken place and his loss of LIP status and
that he believed he had been unfairly treated. However, there were also reports
that the captain had accepted his demotion.
It is also suggested that the appropriate section of NTSC draft Final Report be
revised to include the information regarding the captain’s use of medical leave to
demonstrate a change in his behavior in the 3 months before the accident. For
example, during his 6-year career at SilkAir, the captain visited the airline’s
preferred medical group on 31 occasions. No major, excessive, or extra-ordinary
medical problems were ever reported or noted. The PIC submitted six medical
certificates, each requiring a day off flying duties during his SilkAir career.
These occurred on August 1, 1995, July 23, 1996, August 10, 1997, October 1,
1997, and November 12 and 24, 1997. All certificates were associated with
temporary conditions such as upper-respiratory tract infection, flu, or
gastroenteritis.
As previously stated, a section describing the first officer’s behavioral
characteristics should be added to the draft Final Report. This section should
N-34
precede “section 1.18.4.3 Recent Behaviour” in the draft Final Report (thereby
making “Recent Behaviour” section 1.18.4.4) and be identified as “1.18.4.4
Behavioral Characteristics.” The following information should be included:
During HPG interviews, the first officer was described as a quiet and
reserved person. However, it was reported that if something needed to be
said, he would say it. He was described as being mature for his age,
confident, and likeable.
The HPG received no reports of any
interpersonal problems between the first officer and other pilots or other
employees at SilkAir. There were no indications that he was experiencing
any personal problems.
While at Massey University, the first officer completed a California
Personality Inventory. The results indicated that the he was welladjusted, conservative, stable, and confident. There were no reports that
the first officer had any unusual or abnormal habits in the cockpit during
flights.
The NTSC draft Final Report should also be revised to include information
about the first officer’s medical history while at SilkAir. Specifically, the first
officer visited the airline’s preferred medical group on three occasions, and no
medical certificates were submitted requiring time off of work due to illness.
NTSC’S COMMENTS:
All group reports have been analysed and integrated
into the final report. Specific details such as those
stated above were not deemed necessary for inclusion in
the final report.
N-35
Additional Comments on NTSC Factual Section
Page 15, 1.12.3.2, Flight control surface diagrams should accompany the discussion
of the spoiler actuators, leading edge flaps, etc., to facilitate the reader’s
understanding of these mechanisms.
Page 16, 1.12.3.3f, The statement, “As the actuator is of the piston type, the position
may be indicative of the last position at impact”, should be changed to read “As the
actuator is of the piston type, the position may not be indicative of the last position
at impact.” This change would make the statement consistent with a similar
statement made in section 1.12.3.3.a.
Pages 16 and 17, 1.12.3.3.g and 1.12.3.3.h, The heading for each paragraph should
be changed to read “Aileron Power Control Units and Autopilot Servos” and
“Elevator Power Control Units and Autopilot Servos” respectively.
Page 21, 1.16.4, Paragraph 3, The statement that the rudder balance weight
separation did not occur “…while the aircraft was cruising at FL350 but at a lower
altitude” is correct. However, it should be emphasized that the balance weight
separated after the airplane departed cruise flight. Thus, it is suggested that the
sentence be revised to read “The rudder balance weight did not separate in while
the aircraft was in cruise flight at FL350 but at a significantly lower altitude, after
the airplane departed cruise flight.”
Page 21, 1.16.4, This section only discusses the BASI trajectory study. The NTSB
performed a “desktop” simulation that is referenced as Appendix G. The results of
this simulation should be included in this section to provide a complete picture of
the aircraft trajectory and break-up after departure from cruise flight.
Page 24, 1.18.2.3, There is no “Figure 15” as referenced in this paragraph.
NTSC’S COMMENTS:
Figure 15 is included in the final report. It was not
ready at the draft report stage.
Page 25, 1.18.3.2, Paragraph 4, It is suggested that instead of using the word “deappointed” to describe the captain’s loss of his LIP status, the words “demoted” or
“removed” be used.
NTSC’S COMMENTS:
The word “de-appointed” is appropriate. The LIP status
is not a promotion position but an additional lateral
appointment.
N-36
2.
2.1
ANALYSIS
Introduction
The following statement in section 2.1 should be clarified:
In accordance with Annex 13, a report was made to the relevant
aviation security authorities in late 1999. While the technical
investigation continued, aviation security authorities conducted a
separate investigation, which is not covered in this report.
Currently there is no support text in the factual part of the NTSC draft Final
Report that explains why the contents of the investigative report were made
available to the relevant aviation security authorities. In accordance with ICAO
Annex 13 requirements regarding suspicion or evidence that an accident was the
result of a criminal act, paragraph 5.11 states "If, in the course of an investigation it
becomes known, or it is suspected, that an act of unlawful interference was involved,
the investigator-in-charge shall immediately initiate action to ensure that the
aviation security authorities of the State(s) concerned are so informed.” In the case
of MI 185, this notification was necessary because the technical examination of the
aircraft wreckage revealed there was no evidence of a mechanical malfunction of the
aircraft structure, systems or powerplants that would have caused the aircraft to
depart cruise flight. Further, the HPG developed sufficient personal background
data pertaining to the captain to warrant the “relevant aviation security
authorities” to conduct a further investigation of the captain.
NTSC’S COMMENTS:
The notification to the relevant aviation security
authorities was not made because there was no evidence
of a mechanical malfunction of the aircraft structure,
systems or powerplants. The findings of the HPG at that
time (July 99) was taken into consideration in the
notification.
2.3.3
Explanation to the Break Up of the Empennage
Close examination of the wreckage (Section 1.22) supports the results of the flutter
analysis (Section 2.3.2) and the trajectory analysis (Section 2.3.1).
The above results suggest that the separation of the empennage parts could have had
occurred at an altitude near or below 12,000 ft, due to an unstable flutter as the aircraft
exceeded 1.2 Vd.
N-37
These two sentences may be misleading. To provide clarification, the NTSC
draft Final Report should be revised to state that evidence indicates that the
separation of the empennage components/parts was not the cause of the
departure from cruise flight or the accident but was the result of an
overspeed condition that occurred after the airplane departed cruise flight.
NTSC’S COMMENTS:
The two sentences are satisfactory.
2.4
Power Control Units and Actuators
2.4.1
Main Rudder PCU
In the controlled laboratory test condition [Reference 16], it was found that problems
due to thermal shock can arise. This can happen if the warm hydraulic fluid (at +77°C)
rushes into a cold-soaked servo valve (at -40°C) causing the slides to expand against the
valve housing. In such a temperature difference, a valve jamming could occur causing the
rudder to move uncommanded or in a direction opposite to the rudder pedal command
(rudder reversal). In real flight, the hydraulic temperature would not reach that high
(+77°C) a level.
An introductory paragraph should be included in section 2.4.1 that explains
why the rudder PCU was examined and described in greater detail than the other
actuators/PCUs. This introduction would provide the reader with a brief
background about the known rudder PCU anomalies identified in previous
accidents and the reason for the additional examination.
Further, for clarification, the last two sentences in this section should be
revised to read, “In such a temperature difference, if the valve jams and the pilot
commands additional rudder input, the result could be an unintended rudder
movement in a direction opposite to commanded input (rudder reversal). However,
the temperature of the hydraulic fluid rising to +77°°C is not likely in normal inflight operations.
Finally, this section should contain a conclusion statement that indicates that
the investigation determined that the rudder PCU was not a cause or contributing
factor in the accident.
NTSC’S COMMENTS:
In view of the FAA 737 Flight Controls Engineering Test
and Evaluation Board Report and the FAA AD --- which
recommends that -- , the NTSC is of the view that this
section should not contain a conclusion statement on
whether the rudder PCU was or was not a cause or
contributing factor in the accident.
N-38
2.4.5
Horizontal Stabilizer Jackscrew
A malfunction affecting both trim switches on a control wheel could also cause a runaway. It was not possible to ascertain if such an occurrence took place. However, had a
run-away occurred due to a malfunction of the main electrical trim system, it would take
about 10 seconds to change from 4.5 to 2.5 units (at a rate of trim change of 0.2 unit/sec
at flaps retracted position). The trim wheel would turn continuously. The movement of the
trim wheels and the sound produced would have been noticed by the pilots. Both pilots
were trained to recognize such a condition and to take appropriate corrective actions.
This paragraph should be modified to include a conclusion that based on the
evidence derived from the last recorded FDR position, the NTSB simulation and the
physical evidence found during the wreckage examination, the stabilizer trim was
moved to the full nose-down limit through pilot input to the main electric trim
system and not due to an “uncommanded” or “runaway” trim condition.
NTSC’S COMMENTS:
There is no evidence to conclude that pilot input was
responsible for the final position of the stabilizer.
The effect of a system run-away of the horizontal stabilizer trim was simulated in the
Garuda Indonesia Training Simulator as well as Boeing M-Cab Simulator, see Appendix
G. A trim change from 4.5 to 2.5 units changed the aircraft attitude from a nose-up to a
nose-down attitude. The simulator results showed that, with such a trim change, it took 1
minute and 23 seconds to descend from 35,000 feet to 19,500 feet. However, the last five
ATC radar points showed a much faster descent of the accident aircraft, i.e. 32 seconds
from 35,000 feet to 19,500 feet. Therefore, if the simulation was correct, the change of
horizontal stabilizer trim position alone would not have resulted in the fast descent after
leaving FL350.
The phrase in the preceding paragraph, “if the simulation was correct”
should be removed unless there are specific doubts regarding the accuracy of the
simulation. If there is evidence to support the accuracy or inaccuracy of the
simulation, this information should be discussed in detail in the factual and the
analysis. Further, Boeing does not have any additional qualifiers about the
simulation accuracy other than the verification by flight test to .89 Mach and
extrapolation to .99 Mach.
NTSC’S COMMENTS:
This qualifying phrase recognises that simulation tests
cannot
fully
replicate
actual
flight
conditions
especially flight conditions beyond the normal flight
envelope.
N-39
2.4.6
Other Actuators
During the tear down examination, the following components were found to be in the
stowed or retracted position:
• Flight spoiler actuators
• Outboard ground spoiler actuators
• Inboard ground spoiler actuators
• Trailing edge flap ballscrews
• Leading edge flap actuators
• Leading edge slat actuators
• Mach trim actuator
• Thrust reverser actuators
The fact that these actuators were found in the stowed or retracted position does not
necessarily suggest that their respective systems were not activated during the descent. If
the respective systems remained in the stowed or retracted positions, they would not have
been factors contributing to the accident.
There was sufficient evidence to indicate that the actuators had performed as
intended. Thus, this section should be modified to include a conclusion that there
was no evidence of a mechanical malfunction or failure of any flight control PCU or
actuator that either caused or contributed to MI 185’s departure from cruise flight
or the resulting accident.
NTSC’S COMMENTS:
As only 73% of the highly fragmented wreckage was
recovered, airplane-related mechanical malfunctions or
failures cannot be totally ruled out.
2.6 Stoppage of the CVR and FDR
2.6.1 CVR Stoppage
The CVR recording ended while the aircraft was still cruising at an altitude of 35,000
feet, about seven minutes before the last radar return. Up to the CVR stoppage, the
conversations in the cockpit was consistent with normal flight operation.
The CVR stoppage could have occurred due to a malfunction of the unit itself or a loss of
power to the unit. The loss of power to the unit could be due to power interruption to the
Electronics Bus 1 that supplies power to the CVR, short circuit or overload, CVR circuit
breaker pulling or break in the wiring.
The entire two-hour recording was found normal. There were no observed anomalies
when power was transferred on the ground in Jakarta. It appeared that the recorder’s
internal energy storage capacitor was operating normally by providing continuous
N-40
recorder operation in spite of momentary aircraft electrical power interruptions,
[Reference 4].
The examination of the CVR unit performed by the manufacturer (Appendix F) confirmed
that the CVR was functioning properly. The recording had characteristics that would be
expected of a normal electrical power shutdown of the CVR. Therefore, the stoppage of
the CVR could be a result of the loss of power to the unit.
According to the aircraft wiring diagram 24-58-11 (Figure 16) the power to the CVR was
from the Electronics Bus 1 (Elex Bus 1). The Elex Bus 1 also supplies power to other
systems, such as the FDR, DME-1, TCAS, ATC-1 etc. Parameters of DME-1 and TCAS
were recorded in the FDR. Analysis of the FDR recording showed that six minutes after
the CVR stopped, the FDR was still recording TCAS and DME-1 parameters. This
indicates that the CVR stoppage was not due to power loss at Elex Bus 1.
The CVR is equipped with an energy storage capacitor. The function of this capacitor is
to provide power for 250 milliseconds after electrical power is removed from the unit
such as when the aircraft power is switched from ground power to APU generators or the
engine generators. Another function of this capacitor is to enable continued recording for
another 250 milliseconds after power loss to the unit.
Had there been an overload or short circuit, the resultant popping of the CVR circuit
breaker in the cockpit would have been recorded as a unique and identifiable sound
signature by the CVR (see Section 1.11.1). Based on the examination of the results of the
circuit breaker pull tests, there was no such sound signature in the MI 185 CVR
recording found. This indicates that there were no short circuit or overload to cause the
CVR circuit breaker to pop out.
The results of the CB pull tests showed that the sound signature associated with manual
pulling of the circuit breaker is obscured by the cockpit ambient noise. Hence, no
conclusion can be drawn whether the circuit breaker had been pulled manually.
A break in the wire supplying power to the CVR could also lead to CVR stoppage without
any sound being recorded on the CVR. However, from the limited quantity of wiring
recovered it could also not be determined if a break in the wiring had caused the CVR to
stop.
Thus, the cause of the CVR stoppage could not be concluded.
The conclusions presented by the NTSC regarding the stoppage of the CVR
are not in full agreement with the evidence. As previously stated in the comments to
the factual portion of the draft Final Report, postaccident examination and testing
proved that there were no mechanical malfunctions or failures of either the CVR or
DFDR that would have caused the recorders to stop recording data.
N-41
Further, it is highly unlikely that the CVR lost power because of a broken
wire, as the NTSC’s analysis suggests, without a related “short circuit” or power
loss to other systems in a related wiring bundle or electrical bus, which likely would
have been reflected on the DFDR. If the short circuit had occurred, the circuit
breaker would have popped, which would have been recorded on the CVR. NTSB
tests established that if an “overloaded and a short circuit” condition had occurred,
the sound of the circuit breaker popping is unique and loud enough to be identified
on the CVR area microphone channel on the ground and in flight. No such sound
was recorded on the CVR from MI 185.
Sufficient evidence has been documented, based on postaccident testing and
examination, to conclude that the failure was not the result of a “fault” or the CVR
“internal power supply and “hold up” capacitor, which appeared to be operating
normally. Additionally, postaccident examination and testing revealed that the
CVR recording exhibited characteristics that would be expected of a normal
electrical power shutdown of the CVR.
NTSC’S COMMENTS:
The evidence indicates that the stoppage of the CVR was
not due to a short circuit or an overload condition.
As only a limited amount of electrical wires was found,
the failure of the CVR as a result of a broken wire
cannot be completely ruled out.
2.6.2
FDR Stoppage
The FDR stopped recording at 09:11:33.7, or 6 minutes and 18.1 seconds after the CVR
stoppage, and approximately 35.5 seconds before the aircraft started its descent, see
Section 1.11.1 and Figure 2. Data recorded by the FDR indicates that the flight was
normal until the FDR stoppage time. It was concluded that until the stoppage of the FDR,
there were no indications of unusual disturbance (e.g. atmospheric turbulence, clear air
turbulence, or jet stream upsets, etc.) or other events affecting the flight.
The FDR stoppage could have occurred due to a loss of power supply to the FDR, or the
malfunction of the unit itself.
The recording of the ATC radar plots during the descent of the aircraft until 19,500 ft
indicated that the aircraft ATC transponder continued operating after the FDR had
stopped recording. SilkAir stated that generally flight crews use ATC-1 flying outbound
from Singapore, and ATC-2 inbound. ATC-1 is on the same bus as the FDR, while ATC-2
is powered from Elex Bus 2, i.e. a different power source. No conclusion could be drawn
as to the reasons for the CVR and FDR stoppage at different times.
The FDR was determined to be functioning normally until it stopped. The stoppage of the
FDR could not be determined from the available data.
N-42
There were no evidence found that could explain the six-minute time difference between
stoppage of the CVR and FDR.
The NTSC draft Final Report’s discussion regarding the stoppage of the FDR
needs to be revised to indicate that in addition to the possibilities mentioned, the
DFDR’s stoppage can also be explained by someone manually pulling the circuit
breaker. This discussion should also be revised to reflect that the DFDR is powered
through the same electrical bus (Electronics Bus 1) as ATC-1 (one of the airplane’s
two radar transponders) and the Mach trim actuator. The radar transponder
(which was likely ATC-1 during the accident flight) continued to operate and return
data for a short time after the DFDR stopped. In addition, the Mach trim actuator
was found at its high speed (not cruise speed) setting, indicating that it was powered
and operational during the airplane’s high-speed dive. It can be concluded that the
absence of a malfunction of the DFDR up to the point at which it stopped, combined
with the fact that the transponder continued to transmit and the Mach trim actuator
continued to operate after the DFDR had stopped, indicates that the stoppage was
not due to a loss of power to Electronics Bus 1. However, the stoppage could be
explained by someone manually pulling the circuit breaker.
The NTSC draft Final Report’s discussion of the SilkAir practice of flight
crews using ATC-1 when flying outbound from Singapore and ATC-2 when
returning to Singapore is not documented in either the AAIC Operations or HPG
reports. Further, this statement is contrary to information provided to the HPG
group that the transponder in use during a flight typically corresponds to the pilot
flying. Therefore, in the case of the accident flight, ATC-1 would have been
selected. It is suggested that this statement be corrected.
Finally, the NTSC draft Final Report should include a discussion of human
actions as a possible cause of the CVR and DFDR stoppage.
NTSC’S COMMENTS:
As the data pertaining to the use of the ATC-1 and ATC2 is not captured by the FDR, it cannot be confirmed
for the accident flight, whether ATC-1 or ATC-2 had
been selected. The NTSC draft Final Report contains a
discussion of the human aspects of stoppage of the CVR
recording. Discussion of the human aspects of stoppage
of the FDR recordings was not meaningful as there was
no available CVR recording to assist in the discussion.
2.10
Simulated Descent Profile
The last five ATC radar points recorded represent the flight trajectory of the aircraft
from the cruise altitude 35,000 feet to approximately 19,500 feet. Each point consisted of
data relating to time, altitude and geographical coordinates.
N-43
Simulator tests and computer simulation fly-out studies were done to determine failures
or combination of failures of the flight control and autopilot systems that could result in
the extreme descent trajectory. Aircraft flight data were not available for the time period
after the stoppage of the FDR. The initial condition for these tests and studies was cruise
configuration at 35,000 feet based on the last known FDR data. The altitude range for
the simulations was from 35,000 feet to approximately 19,500 feet.
The results of these simulation studies (Appendix G) are summarized as follows:
•
Any single failure of the primary flight controls such as hard-over or jamming of
aileron, rudder or elevator did not result in a descent time history similar to that of
the last ATC radar points. In simulations of these flight control failure conditions the
aircraft could be recovered to normal flight manually.
•
Any single failure of the secondary flight controls such as hard over or jamming of
yaw damper, or runaway of the stabilizer trim would not result in a descent time
history similar to that of the last ATC radar points. In simulations of these flight
control failure conditions the aircraft could be recovered to normal flight manually.
•
Manipulation of the primary flight controls without horizontal stabilizer trim would
result in a descent time history similar to that of the last ATC radar points. But this
required large control column input forces and the aircraft was subjected to a
loading exceeding 2 G. However, if the control column input forces were relaxed, in
the simulations the aircraft would recover from the steep descent due to its inherent
stability.
•
Among other possibilities, a combination of changing the stabilizer trim from about
4.5 to 2.5 units and an aileron input could result in a descent time history similar to
that of the last ATC radar points. This simulated descent trajectory would result in
the aircraft entering an accelerating spiral and being subjected to a loading of less
than 2 G. Furthermore, the aircraft would continue in the spiral even when the
control forces were relaxed. This would result in a descent at a speed exceeding 1.2
Vd, in agreement with the analysis on the break up of the empennage as discussed in
Section 2.3.
Bullet 3 should be modified as follows for correctness and clarity:
Although manipulation of the primary flight controls without
horizontal stabilizer trim would result in a descent time history
similar to that of the last ATC radar points, this would and large
control require control column forces greater than 50 pounds column
inputs. However, the simulations indicated that if the control column
input forces had been relaxed, the aircraft would have initiated a
return to a nose-up attitude due to its inherent stability.
N-44
Bullet 4 should be modified as follows for correctness and clarity:
Among other possibilities, a combination of either control column
inputs and/or changing the stabilizer trim from about 4.5 to 2.5 units,
in combination with aileron inputs could result in a descent time
history similar to that of the last ATC radar points. This simulated
descent trajectory would result in the aircraft entering an accelerating
spiral and being subjected to a loading of less than 2 G. Furthermore,
the aircraft would continue in the spiral even when the control forces
were relaxed. This would result in a descent at a speed exceeding 1.2
Vd, which is in agreement with the analysis of the breakup of the
empennage as discussed in section 2.3.
Based on the data derived from the simulations, the following conclusion can
be made regarding the maneuvers necessary for the airplane to fly a profile similar
to that of MI 185:
No single mechanical failure of the airplane structure or flight
control systems was found that would have resulted in movement of
the airplane that matched the recorded radar data points. Further,
there was no evidence of any combination of systems failures. Thus,
no known or postulated mechanical failure was found that resulted in
a flight profile that matched the radar data. However, changing the
flight control input manually in multiple axes did provide a flight
profile that matched the last recorded ATC radar data points.
Therefore, it is probable that the airplane was likely responding to
sustained flight control inputs from the cockpit.
NTSC’S COMMENTS:
As there was no FDR data available from before the
commencement of descent up to time of impact, it is not
possible to conclude that the airplane was responding
to sustained flight control inputs from the cockpit.
2.11
High Speed Descent Issues
2.11.1 Mach Trim System and its Function
The aircraft was equipped with a Mach Trim system to provide stability at the higher
operating speeds, i.e. higher Mach numbers. Mach trim is automatically accomplished
above Mach 0.615. When the Mach Trim system is operative it will normally compensate
for trim changes by adjusting the elevator with respect to the stabilizer, as the speed
increases. With the Mach Trim inoperative, the aircraft could exhibit a nose down
N-45
tendency ("Mach Tuck") as speed increases. However, the expected control forces to
overcome the “Mach Tuck” are light. Additionally when the speed exceeds the maximum
limit, audible overspeed warnings are activated.
Since the aircraft was cruising at subsonic speed (Mach 0.74) and trimmed for level
flight, the aircraft will eventually return to the trimmed condition after a minor speed
disturbance.
For the aircraft to dive, a significant disturbance resulting in an increasing speed must
have taken place. Such a disturbance could be initiated by changing aircraft elevator or
stabilizer trim. Should the airspeed increase to the point where it becomes transonic, and
as the lift resultant moves aft and local supersonic flow develops, the nose-down pitching
moment could be sufficiently large that the aircraft becomes speed unstable, i.e.
continuing speed increase of the aircraft. Once the aircraft is in a transonic dive,
recovery from the dive becomes more difficult because of an increase in control column
forces due to the aircraft’s increasing nose down pitching moment as well as a large
reduction of elevator effectiveness due to the formation of shock induced air flow
separation in front of the elevator.
It is possible to recover from a transonic dive by timely action of the pilot, by reducing
thrust and deploying the speed brakes. Should the pilot not initiate a prompt recovery
action, the recovery becomes more difficult.
During the tear down examination, the mach trim was found in the fully retracted
position. The fact that this actuator was found in the retracted position may not
necessarily indicate that the mach trim system is a factor contributing to the accident.
Because the Mach Trim system was not implicated as a cause or contributing
factor in the accident, the discussion regarding this system is irrelevant. Therefore,
it is strongly suggested that the discussion in this section be substantially reduced
and that a definitive conclusion be included indicating that there was no evidence of
a Mach Trim system failure that would have been causal or contributing to the
accident.
2.11.2 Emergency Descent due to Fire, Smoke or Depressurization
An emergency descent is necessary when there is a rapid cabin depressurization or when
a fire or smoke occurs in flight. The procedure is to simultaneously retard the thrust
levers, deploy the speed brakes and bank the aircraft to initiate the descent. (Appendix
K). Some forward stabilizer trim is applied to attain a dive which will accelerate the
aircraft towards the maximum speed limit. Once the maximum speed is reached aircraft
is re-trimmed to maintain the speed. This facilitates a limit on maximum rate of descent
to the minimum safe altitude.
N-46
The last pilot radio transmission about two and a half minutes before the descent
sounded normal and there was no mention of any in-flight fire or smoke. Furthermore,
examination of the wreckage showed no evidence of in-flight fire or explosion.
Examination of the recovered oxygen generators showed that they were not activated.
This indicated that there was no rapid depressurization at high altitude.
Based on the above findings, there was no indication of an emergency descent due to fire,
smoke or rapid depressurization.
The first paragraph in this section is not a statement of analysis but of fact.
The information presented refers to procedures (included in the NTSC draft Final
Report as appendix K) to be employed by the flight crew in the event that an
emergency descent is necessary. Moreover, the statement that the emergency
descent procedures call for the pilot to “bank the aircraft to initiate the descent” is
incorrect. The procedures do not specify banking the aircraft as the method to be
used to initiate the descent. Because this information has no relationship to the
accident and implies that a true emergency descent profile is similar to the derived
profiles used in the simulator to match the accident descent profile, this statement
should be removed. However, if it is to remain in the analysis, it must be corrected
by removing the statement “banking the aircraft to initiate descent.”
2.13
Human Factors Aspects of the CVR and ATC Recordings
2.13.1 CVR
(a) The conversations and sounds recorded by the CVR before it stopped were examined.
The CVR transcript (Appendix A) showed that at 09:04:55 the PIC indicated his
intention to go to the passenger cabin " .… go back for a while …. finish your
plate….". At 09:05:00 the PIC offered water to the F/O, and at about the same time,
several metallic snapping sounds were recorded. Thirteen seconds later, at
09:05:13.6 the CVR ceased recording. Analysis of the recording indicated that the
metallic snapping sounds were made by a seatbelt buckle striking the floor. (See
Section 1.16.2)
(b) During the period recorded by the CVR, all door openings or closings were
related to pre-departure activities, in-flight meal service and normal pilot-cabin
crew interaction. In the four minutes following the last meal service, there were no
sounds associated with cockpit door opening or closing. After takeoff from Jakarta,
conversations within the flight deck were between pilot-to-pilot, pilot-to-flight
attendants, and normal pilot-to-ATC radio communications. During the flight, except
for cabin attendants serving meals and drinks to the pilots, there were no indications
of any other person(s) in the cockpit. It is concluded that after the last meal service
and until the stoppage of the CVR, the recording did not reveal any indications that
N-47
person(s) other than the flight crew and cabin attendants attending to their duties
were in the cockpit.
(c) Analysis of the CVR stoppage indicated that the failure of the CVR could not
have been caused by a short circuit or overload. This is because either occurrence
would have resulted in the CVR recording a “pop” sound which was heard on the
test recording but not on the accident recording.
The CVR in-flight tests could not identify the sound of the CVR circuit breaker
being manually pulled as the ambient noise obscured the sound made. The accident
tape did not contain any identifiable sound attributable to manual pulling of the
CVR circuit breaker. It was not possible to determine from the CVR tests if there
was a pulling out of the CVR circuit breaker.
The information presented in paragraph (c) regarding the CVR is
redundant. Because no conclusions are drawn, it is not necessary to discuss this
information again.
In addition, as noted in the summary, this section should address the fact
that the captain was in the process of leaving the cockpit at the time the CVR
stopped recording. It should provide a description of the position of the CVR and
DFDR circuit breakers in relation to the captain’s seat, the door, etc. Finally, the
information about the captain’s previous CVR event (in June 1997) should be
emphasized.
NTSC’S COMMENTS:
The PIC’s previous CVR event has no bearing on this
accident. The location of the CVR circuit breaker is
provided in in-flight documentation.
2.13.2 ATC Recordings
The data transcribed from the ATC communications recording of the air-to-ground
conversation indicates that at 09:10:26, or 5 minutes and 10.4 seconds after the CVR
stoppage, the F/O acknowledged the “abeam Palembang” call from the ATC. The F/O
was positively identified by voice analysis examination. This confirms that the F/O was in
the cockpit when the aircraft was abeam Palembang. However, it is not possible to
conclude whether the PIC was in the cockpit at the time. It was also not possible to
determine events or persons present in the cockpit from the time of the last transmission
to ATC.
The absence of a distress call could suggest that the pilots were preoccupied with the
handling of an urgent situation. However, it is not possible to conclude on the reason for
the absence of a distress call.
N-48
The NTSC’s conclusion that the absence of a distress call likely indicates that
the pilots were attempting to “handle an urgent situation” is misleading because it
implies that the pilot(s) perceived the situation as an emergency. The discussion in
this section should be modified to make clear the possibility that the absence of a
distress call could suggest that the pilot(s) did not consider the situation a condition
of distress, that is, the airplane was doing what a pilot commanded it to do.
NTSC’S COMMENTS:
Just as it is not possible to conclude that the pilots
perceived the situation as an emergency, it is also not
possible to conclude that the airplane was doing what a
pilot commanded it to do.
2.14
Specific Human Factors Issues
In this section, the specific, personal, financial backgrounds and recent behavior of the
PIC and the F/O are examined.
2.14.1 Personal Relationships
Evidence obtained from family and friends of both the PIC and F/O reported no recent
changes or difficulties in personal relationships.
It was concluded there was no evidence that either pilot was experiencing difficulties in
any personal relationships.
2.14.2 First Officer (F/O)
The investigation into the F/O's personal and professional history revealed no unusual
issues. No records of incidents or unusual events were found, and no career setbacks or
difficulties were experienced. Financial records showed no evidence of financial
problems. Interviews with family, close friends and relations seem to indicate that the
F/O was a well-balanced and well-adjusted person, and keen on his job, and planning to
advance his a flying career. There were no reports on recent changes in his behavior.
2.14.3 Pilot-in-Command (PIC)
The investigation into the personal and professional career revealed that the PIC was
considered to have been a good pilot, making his transition from a military pilot to
commercial pilot smoothly. His career at SilkAir showed that he was well accepted and
given higher responsibilities. He was considered to be a leader among the Singaporean
pilot community in SilkAir.
N-49
During his professional career at SilkAir, he was involved in a few work-related events,
which were in general considered minor operational incidents by the management.
However in one particular event, for non-technical reasons the PIC infringed a standard
operating procedure, i.e. with the intention to preserve a conversation between the PIC
and his copilot, the PIC pulled out the CVR circuit breaker, but the PIC reset the circuit
breaker in its original position before the flight. This was considered a serious incident
by the management, and the PIC was relieved of his LIP appointment. The PIC was
known to have tried through existing company procedures to reverse the management
decision. Although there were some indications of the PIC being upset by the outcome of
the events, the magnitude of the psychological impact on the PIC could not be
determined.
The PIC’s financial history was investigated for the period from 1990-1997. Based on the
data available to the NTSC it was noted that the PIC’s accumulated losses in share
trading increased between 1993 and 1997 and his trading activity was stopped on two
occasions due to non-settlement of his share trading debt. The data available also
showed that his loans and debts were greater than his realizable assets and his monthly
income (including his immediate family’s income) was less (about 6%) than his estimated
monthly expenses at the time of the accident.
2.14.4 Recent Behaviour
The PIC’s recent behaviour was analysed from statements made by family members,
friends and peers during interviews. The PIC’s family reported no recent changes in his
behaviour. Work associates who observed the PIC on the day of the accident and on his
most recent flights, reported nothing odd or unusual in his behaviour.
2.14.5 Insurance
Based on the data available, it was found that at the time of the accident, the PIC had a
number of life insurance policies. The majority of these were taken up earlier in his life.
The most recent policy was a mortgage policy which was required by the financial
institution from which he took the loan for his house in line with normal practice for
property purchases in Singapore. The PIC applied for the mortgage policy on 27
November 1997. The insurance company approved the policy on 12 December 1997
pending payment of the first premium. The PIC submitted a cheque dated 16 December
1997 for the first premium payment. The commencement or the inception date of the
policy was set by the insurance company to be 19 December 1997. This information was
not conveyed to the PIC. The cheque was cleared on 22 December 1997. From the data
available to the NTSC there was no evidence to indicate if this mortgage policy has any
relevance to the accident.
NTSC concluded that the combination of financial situation and his work related events
could be stressors on the PIC. However, NTSC could not determine the magnitude of
these stressors and its impact on the PIC’s behavior.
N-50
The deficiencies, inaccuracies, and omissions of relevant information
pertaining to the captain’s and first officer’s personal, career, and financial
backgrounds have been previously discussed. It is imperative that complete and
accurate factual information be presented for analysis so that a proper and
thorough analysis can be accomplished. The factual information suggested for
inclusion will serve as the basis for revising the analytical discussion and
conclusions of the human factors issues in this section.
Finally, it should be noted in this section that the captain had been told that
the insurance policy would go into effect upon receipt of the first premium payment.
NTSC’S COMMENTS:
NTSC notes that the above details about the crew’s
personal background, the PIC’s professional background
in
RSAF
and
SilkAir
and
financial
background
information and the crew’s recent behaviour which are
proposed to be included in the final report are taken
from the HPG Report version 6.0. As explained earlier,
the HPG Report version 6.0 was overtaken by information
from the
PricewaterhouseCoopers Audit. Relevant
information from the HPG Report version 6.0 which had
or might have a bearing on the accident were integrated
into the final report. Whatever had no bearing was not
included.
In accordance with 5.12 of Annex 13 to the Convention
on International Civil Aviation, private information
regarding persons involved in the accident shall be
included in the final report or its appendices only
where pertinent to the analysis of the accident. Parts
of the private information not relevant to the analysis
shall not be included.
N-51
3.
3.1
CONCLUSIONS
Findings
Engineering and Systems
•
There was no evidence found of in-flight fire or explosion.
•
From flutter analysis and wreckage distribution study, the empennage break-up
could have occurred in the range between 5,000 and 12,000 feet altitude.
•
Examination of engine wreckage indicated that the conditions of the engines at
impact were not inconsistent with high engine rotation speed. No indications were
found of in-flight high energy uncontained engine failures. Therefore, the engines
were considered to be not a factor contributing to the accident.
•
Examination of the actuators of flight and ground spoilers, trailing and leading edge
flaps, as well as engine thrust reversers indicate retracted or stowed positions of the
respective systems.
•
Examination of the main rudder power control unit (including the servo-valve), the
yaw damper modulating piston, the rudder trim actuator, the rudder trim and feel
centering unit, the standby rudder PCU, the aileron PCUs, the elevator PCUs, and
the horizontal stabilizer jack-screw components, revealed no indications or evidence
of pre-impact malfunctions.
Based on the evidence and postaccident testing, a definitive conclusion can be
made regarding the flight control systems. Therefore, it is suggested that the
NTSC’s draft Final Report be modified to include the following:
There was no evidence of a mechanical failure of any of the flight
control systems or related components that would have been causal or
contributing to the accident.
Also, it is suggested that the following conclusion be added for completeness:
Separation of the empennage components/parts were not the cause of
the departure from cruise flight or the resulting impact with terrain
but, rather, were the result of an overspeed condition that occurred
after the airplane departed cruise flight.
N-52
•
Examination of the 370 kg of recovered electrical wires, connectors and circuit
boards showed no indication or evidence of corrosion, shorting, burning or arcing in
these wires or parts.
•
The CVR stopped recording at 09:05:15.6 and the FDR stopped recording at
09:11:33.7. The examination of the CVR and FDR showed no malfunction of the
units. The stoppages could be attributed to a loss of power supply to the units.
However, there were no indications or evidence found to conclude on the reason for
the stoppages due to the loss of power. The cause of the CVR and FDR stoppages
and the reason for the time difference between the stoppages could not be concluded.
The NTSC draft Final Report suggests that the cessation of the CVR and
DFDR could in each case be explained by a broken wire. Although this is
technically correct, the probability of two such unrelated wire breaks occurring
several minutes apart and affecting only the CVR and DFDR is so highly
improbable that it cannot be considered a realistic possibility.
•
The inspection of the aircraft maintenance records did not reveal any defects or
anomalies that could have affected the airworthiness of the aircraft or that may have
been a factor contributing to the accident.
•
The horizontal stabilizer trim was found to be in the 2.5 units position which
matched the forward nose-down limit of the manual electrical trim.
This conclusion should be expanded to include a definitive statement that the
2.5 units of nose-down trim was the result of a sustained manual input and not
attributed to a malfunctioning system resulting in a “runaway.”
Flight Operations
•
Weather and Air Traffic Control were not factors contributing to the accident.
•
Audio spectral analyses on Air Traffic Control communications and the accident
CVR indicate that the last communication from the MI 185 at 09:10:26, occurring at
a position approximately abeam Palembang was performed by the F/O.
•
The examination of the flight deck noise and sounds concludes that the metallic snap
recorded on the CVR was made by a seatbelt buckle hitting against a metal surface.
•
Based on flight simulations, it was observed that the simulated descent trajectory
resulting from any single failure of flight control or autopilot system would not
match the radar data.
•
Based on the same flight simulations, it was also observed that the trajectory shown
by the radar data could have been, among other possibilities, the result of the
N-53
combination of lateral and longitudinal inputs together with the horizontal stabilizer
trim input to its forward manual electrical trim limit of 2.5 units.
To clarify the conclusion at bullet 5, it is suggested that the following
sentence be added to the end:
Despite the stabilizer trim being at the 2.5 unit nose-down setting (its
forward limit), the aircraft would have remained controllable and
appropriate flight control input would return the airplane a normal
flight attitude.
Additionally, the following should be added to the draft Final Report:
No single mechanical failure of the airplane structure or flight control
systems was found that would have resulted in movement of the
airplane that matched the recorded radar data points. Further, there
was no evidence of any combination of systems failures. Thus, no
known or postulated mechanical failure was found that resulted in a
flight profile that matched the accident radar data. However,
changing the flight control input manually in multiple axes did
provide a flight profile that matched the last recorded ATC radar
data points. Therefore, it is probable that the airplane was likely
responding to flight control inputs from the cockpit.”
Human Factors
•
Both pilots were properly trained, licensed, and qualified to conduct the flight.
•
There was no evidence found to indicate that the performance of either pilot was
adversely affected by any medical or physiological condition.
•
Interviews with respective superiors, colleagues, friends and family revealed no
evidence that both the flight crew members had changed their normal behaviour
prior to the accident.
This conclusion is not representative of the findings of the HPG investigation.
Although consistent with the HPG’s conclusion (in HPG report Version 6.0, July 30,
1999) regarding the first officer, the conclusion in the draft report is inconsistent
with the HPG’s conclusion regarding the captain. The HPG report states, "There
were some indications that the captain’s behavior or lifestyle changed prior to the
accident." It is suggested that this conclusion be separated to accurately describe
the captain’s and first officer’s behavior.
N-54
•
There was no evidence found to indicate that there were any difficulties in the
relationship between the two pilots either during or before the accident flight; or had
been experiencing noteworthy difficulties in any personal relationships (family and
friends).
•
Until the stoppage of the CVR, the pilots conducted the flight in a normal manner
and conformed to all requirements and standard operating procedures.
•
Although a flight attendant had been in the cockpit previously, after the last meal
service and until the stoppage of the CVR there was no indication that anyone else
was in the cockpit other than the two pilots.
•
In the final seconds of the CVR recording the PIC voiced his intention to leave the
flight deck, however there were no indications or evidence that he had left.
•
Interviews and records showed that in 1997 the PIC had experienced a number of
flight operations related events, one of which resulted in his being relieved of his LIP
position.
In its evaluation of the data collected, the HPG made a more definitive
conclusion regarding the captain’s career in the 6 months prior to the accident. The
HPG conclusion, "During 1997 the PIC experienced multiple work-related
difficulties, particularly during the last 6 months" should be used to modify the
existing conclusion.
•
The PIC was involved in stock-trading activities, but no conclusions could be made
indicating that these activities had influenced his personal behavior.
The first part of this conclusion, “The PIC was involved in stock-trading
activities” is a statement of fact and does not provide the basis for a conclusion.
Further, the factual report substantiates that at the time of the accident, the PIC
had been requested to pay a significant amount of money for outstanding debts and
did not have liquid assets from which to pay these debts. This latter information
forms the basis for a conclusion regarding the captain’s financial stressors. In the
HPG report, the conclusion was made that "At the time of the accident the PIC was
experiencing significant financial difficulties." Also, this information was presented
in the NTSC’s interim report issued August 1999. Therefore, it is suggested that the
NTSC revise this conclusion to be consistent with information cited in the AAIC
HPG report and that was disseminated to the public in 1999.
•
From the data available to the NTSC there was no evidence found to indicate if the
mortgage policy taken out by the PIC in connection with his housing loan has any
relevance to the accident.
N-55
Finally, the NTSC’s conclusions do not address the crash of the three
training aircraft from the captain’s squadron while he was serving in the military.
As discussed in the comments to section 1.18.3.1, the HPG examined the effect this
event may have had on the captain but could not determine the extent to which he
may have been affected. It is strongly suggested that the NTSC’s draft Final Report
include the conclusion from the AAIC HPG report that states, “The accident [in
Palembang] occurred on the same date as the 1979 RSAF crash in the Philippines;
the extent to which the PIC was affected by this event could not be determined.”
3.2 Final Remarks
•
The NTSC investigation into the MI 185 accident was a very extensive, exhaustive
and complex investigation to find out what happened, how it happened, and why it
happened. It was an extremely difficult investigation due to the degree of destruction
of the aircraft resulting in highly fragmented wreckage, the difficulties presented by
the accident site and the lack of information from the flight recorders during the final
moments of the accident sequence.
•
The NTSC accident investigation team members and participating organizations
have done the investigation in a thorough manner and to the best of their conscience,
knowledge and professional expertise, taking into consideration all available data
and information recovered and gathered during the investigation.
•
Given the limited data and information from the wreckage and flight recorders, the
NTSC is unable to find the reasons for the departure of the aircraft from its cruising
level of FL350 and the reasons for the stoppage of the flight recorders.
•
The NTSC has to conclude that the technical investigation has yielded no evidence to
explain the cause of the accident.
The technical investigation has, in fact, yielded sufficient information,
which was derived through the on-scene and postaccident investigation activities, to
definitively conclude that there were no mechanical anomalies with the aircraft,
there were no environmental anomalies, nor were there any other significant
technical factors that would have caused or contributed to the accident.
Additionally, the statement regarding the participating organizations in
bullet 2 is made with a level of certainty that may not truly reflect the opinions of
the “participating” organizations. Further, the remaining concluding statements do
not necessarily reflect an analysis of all of the facts, conditions, and circumstances
revealed during the course of this accident investigation.
NTSC’S COMMENTS:
With respect to the proposals relating to the
conclusions of the final report, please refer to
comments made at the relevant sections above.
N-56
3
RECOMMENDATIONS TO MANUFACTURERS
1. It is recommended that a comprehensive review and analysis of flight data recorders
and cockpit voice recorders systems design philosophy be undertaken by aircraft and
equipment manufacturers. The purpose of the review and analysis would be to
identify and rectify latent factors associated with stoppage of the recorders in flight,
and if needed, to propose improvements to ensure recording until time of occurrence.
2. It is recommended that a review of the flight recorders design philosophy be
undertaken by aircraft and equipment manufacturers to include recording of actual
displays as observed by pilots in particular for CRT type of display panels.
3. It is recommended that a review of the flight crew training syllabi be undertaken by
aircraft manufacturers to include recovery from high speed flight upsets beyond the
normal flight envelope. The purpose of developing the additional training is to
enhance pilot awareness on the possibility of unexpected hazardous flight situations.
4
GENERAL RECOMMENDATION
4. It is recommended that regional investigation framework for co-operation in aircraft
accident investigations be established to enable fast mobilization of resources and
coordination of activities to support those states that do not have the resources and
facilities to do investigations on their own.
The factual evidence does not support recommendations 1 and 3 because the
postaccident tests and examination suggest that the CVR stopped recording as a
result of the unit’s circuit breaker being pulled. This scenario also likely explains
why the DFDR stopped recording.
The investigation did not reveal any evidence to suggest that a mechanical
malfunction or failure of a particular system caused an unexpected upset. If such a
scenario had occurred, the flight crew should have been able to take immediate
corrective action because they had received training at SilkAir in the recovery from
unusual attitudes. Based on the evidence, the departure from cruise flight was likely
an intentional maneuver; therefore, recommendation 3 is without merit.
NTSC’S COMMENTS:
There was no evidence to positively conclude that the
departure from cruise flight was an intentional
maneuver. Present line crew training does not include
recovery from high speed flight upset beyond the normal
flight envelope from whatever cause. If such crew
training is developed, it would enhance crew awareness
on the possibility of unexpected hazardous flight
N-57
situations and provide them with the necessary recovery
techniques to handle any high speed flight upsets.
N-58
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
50
51
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
52
Figure 3 Boeing B737-300 – Three view drawing
53
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Figure 4.a Flight Data Recorder with an opening corner
Figure 4.b The tape inside FDR
Figure 4.c Sketch of FDR tape layout
Figure 5.a Solid State CVR Module immersed in the river water
Figure 5.b Memory module of the Solid State CVR
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Figure 6.a Early stage of debris recovery (by divers)
Figure 6.b Debris recovery using dredge
Legend:
A
B
C
D
E
F
H
I
Engine parts
Tail sections
Wing sections
Fuselage parts
Human remains
Documents
Cockpit Voice Recorder (CVR)
Avionics equipment/parts
Figure 6.c Sonar photo of the debris distribution buried in the river bottom
Location of components on ground - B 737, Palembang
Sections of stabilisers, elevators and rudder
Rudder composite panels
Elevator composite panels
RH elevator LE
RH elevator trim tab
56
56.5
57
57.5
58
east
58.5
59
59.5
60
25.00
25.50
26.00
8
east
shore
line
9
26.50
F
south
27.50
28.00
16
21 E
27.00
B
D
A
C
4
5
Crash
site
14
28.50
15
29.00
29.50
30.00
Figure 7 Sketch of the wreckage pieces found on land
17
6
10
7
1
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
(a)
(b)
Figure 8 Picture of the wreckage recovered from the river
(c)
(d)
Figure 8 Picture of the wreckage recovered from the river (continued)
(e)
(f)
Figure 8 Picture of the wreckage recovered from the river (continued)
(g)
(h)
Figure 8 Picture of the wreckage recovered from the river (continued)
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
(a)
(b)
Figure 9 Picture of wreckage of the fuselage skin patch repair
64
Figure 10.a Section 48 and Empennage
65
Figure 10.b Horizontal Stabilizer L.H. Side
66
Figure 10.c L.H. Side Elevator
67
Figure 10.d Horizontal Stabilizer R.H. Side
68
Figure 10.e R.H. Side Elevator
69
Figure 10.f Horizontal Stabilizer Center Section
70
Figure 10.g Front Spar Vertical Stabilizer
71
Figure 10.h Rear Spar Vertical Stabilizer
72
Figure 10.i Trailing Edge Beam Vertical Stabilizer
73
Figure 10.j Rudder Honeycomb Panels
74
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
Figure 11 Picture of the reconstructed empennage
75
(a)
(b)
Figure 12 Picture of the impact marks at the cam feel centering unit
76
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
(a) Position of the Stabilizer Control System
(b) Schematic Diagram of the Horizontal Stabilizer Trim
Figure 13 Horizontal Stabilizer Trim
77
(c) Stabilizer Jackscrew
Figure 13 Horizontal Stabilizer Trim (Continued)
78
(d) Captain Control Wheel
(e) Stabilizer Trim Deflection Scale
Figure 13 Horizontal Stabilizer Trim (Continued)
79
National
Transportation
Safety
Committee
Aircraft Accident Report
SILKAIR FLIGHT MI 185
BOEING B737-300
9V-TRF
MUSI RIVER, PALEMBANG, INDONESIA
19 DECEMBER 1997
Jakarta, 14 December 2000
Department of Communications
Republic of Indonesia
(a) NTSB Results based on separation at 35 000 ft
Figure 14 Diagram of debris distribution analysis
80
(b) NTSB Results based on separation at 12 000 ft
Figure 14 Diagram of debris distribution analysis (Continued)
81
Location of components on ground - B 737, Palembang
Sections of stabilisers, elevators and rudder
Rudder composite panels
Elevator composite panels
RH elevator LE
RH elevator trim tab
56
56.5
57
57.5
58
east
58.5
59
59.5
60
25.00
25.50
26.00
8
east
shore
line
9
26.50
F
16
21 E
south
27.00
27.50
28.00
B
D
C
point 27
4
A
5
Crash
site
6
10
7
1
point 26
14
28.50
29.00
Flight path
based on radar
data
15
17
Segm e nt of alternative
flight paths
point 25
29.50
point 24
30.00
(c) BASI Results
Figure 14 Diagram of debris distribution analysis (Continued)
82
0.74M
CRUISE
0.78M
0.88M
STABLE
UNSTABLE
ALTITUDE (FEET)
0.95M
VC
VD
DESCENT
0.97M
1.2VD
FLIGHT TEST
CLEARANCE
ANALYTICAL
CLEARANCE
PREDICTED FLUTTER POINTS
l 22 Hz ANTISYMMETRIC FLUTTER MODE
n 12 Hz ANTISYMMETRIC FLUTTER MODE
TRUE AIRSPEED (KNOTS)
Figure 15 Diagram of Boeing B737 Flutter Flight Envelope
83
(a)
(b)
Figure 16 Corrected Radar Data
84
Figure 17 Wiring Diagram 24-58-11
85
Figure 18 Flight Control Surfaces Location
86
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement