Fly Dubai Accident

This document is an English translation of the Final Report on the accident involving the
Boeing 737-8KN aircraft registered A6-FDN that occurred on March 19, 2016 (00:42 UTC) at
Rostov-on-Don aerodrome, the Rostov Region, Russian Federation.
The translation was done as accurate as a translation may be to facilitate the understanding of
the Final Report for non-Russian speaking people. The use of this translation for any purpose
other than for the prevention of future accidents could lead to erroneous interpretations.
In case of any inconsistence or misunderstanding, the original text in Russian shall be used as
the work of reference.
INTERSTATE AVIATION COMMITTEE

AIR ACCIDENT INVESTIGATION COMMISSION
FINAL REPORT
Type of occurrence Fatal accident

Type of aircraft Boeing 737-8KN, airplane
Nationality and registration marks A6-FDN
Owner Celestial Aviation Trading 38 Limited,
Shannon, County Clare, Ireland
Operator Dubai Aviation Corporation, P.O. Box: 353,
United Arab Emirates
Aviation authority of the place of occurrence Southern FATA Interregional Territorial
Department
Place of occurrence Russian Federation, the Rostov region, the
Rostov-on-Don aerodrome, reference position:
47º15′54.7″ N, 039°49′43.8″ E
Date and time 19.03.2016, 03:42 local time (00:42 UTC),
nighttime
In accordance with the ICAO Standards and Recommended Practices this report is issued with the sole objective to
prevent the air accidents.
It is not the purpose of this report to apportion blame or liability.
Criminal aspects of the accident are tackled within separate criminal case.
Boeing 737-8KN A6-FDN Fatal Accident - Final Report 2
ABBREVIATIONS ..................................................................................................................................................... 3
GENERAL INFORMATION ..................................................................................................................................... 9
1. FACTUAL INFORMATION ............................................................................................................................. 11
1.1. HISTORY OF FLIGHT ....................................................................................................................................... 11
1.2. INJURIES TO PERSONS .................................................................................................................................... 11
1.3. DAMAGE TO AIRCRAFT .................................................................................................................................. 12
1.4. OTHER DAMAGE ............................................................................................................................................. 30
1.5. PERSONNEL INFORMATION ............................................................................................................................. 31
1.5.1. Flight crew information ..................................................................................................................... 31
1.5.2. Ground service personnel information .............................................................................................. 39
1.6. AIRCRAFT INFORMATION ............................................................................................................................... 47
1.7. METEOROLOGICAL INFORMATION .................................................................................................................. 48
1.8. AIDS TO NAVIGATION ..................................................................................................................................... 57
1.9. COMMUNICATIONS ......................................................................................................................................... 66
1.10. AERODROME INFORMATION ........................................................................................................................... 66
1.11. FLIGHT RECORDERS ....................................................................................................................................... 71
1.12. WRECKAGE AND IMPACT INFORMATION......................................................................................................... 72
1.13. MEDICAL AND PATHOLOGICAL INFORMATION................................................................................................ 77
1.14. SURVIVAL ASPECTS ........................................................................................................................................ 78
1.15. SAR AND FIREFIGHTING OPERATIONS ............................................................................................................ 79
1.16. TESTS AND RESEARCH .................................................................................................................................... 81
1.16.1. Wreckage layout ................................................................................................................................. 81
1.16.2. The examination of the stabilizer jackscrew ...................................................................................... 82
1.16.3. Stabilizer trim control switch of the F/O control wheel ..................................................................... 83
1.16.4. The examinations of the elevator control system PCUs ..................................................................... 84
1.16.5. The evaluation of the condition and the serviceability of the electric stab trim motor ...................... 88
1.16.6. The assessment of the language proficiency level of the approach control unit officer ..................... 89
1.16.7. Engineering simulation (Mathematical modelling)............................................................................ 89
1.16.8. On the control column forces ............................................................................................................. 95
1.16.9. The reconstruction of the Head-Up Display/HUD readings .............................................................. 97
1.17. ORGANIZATIONAL AND MANAGEMENT INFORMATION. .................................................................................. 98
1.18. ADDITIONAL INFORMATION ........................................................................................................................... 98
1.18.1. On the fatal accident to the Ilyushin Il-86 RA-86060 aircraft at the Sheremetyevo airport .............. 98
1.18.2. On the stabilizer control at the Boeing 737-800 aircraft ................................................................. 100
1.18.3. On the PFD monitoring at the forward deflection of the control column ........................................ 102
1.19. USEFUL OR EFFECTIVE INVESTIGATION TECHNIQUES ................................................................................... 103
2. ANALYSIS ......................................................................................................................................................... 104
2.1. DESCRIPTION OF THE FLIGHT........................................................................................................................ 104
2.2. ON THE PECULIARITIES OF THE TRIM (RELIEF) OF FORCES ............................................................................ 159
2.3. ON THE USE OF HUD IN PROGRESS OF GO-AROUND ..................................................................................... 162
2.4. ON THE POSSIBLE IMPACT OF SOMATOGRAVIC ILLUSIONS ............................................................................ 166
3. CONCLUSION .................................................................................................................................................. 170
4. THE SHORTCOMINGS, REVEALED IN THE INVESTIGATION .......................................................... 172
5. SAFETY RECOMMENDATIONS .................................................................................................................. 173

ABBREVIATIONS
A/P  autopilot
A/T  autothrottle
AAIC  Air Accident Investigation Commission
AAIS  Air Accident Investigation Sector
AAISTSC  Air Accident Investigation Scientific-Technical Support
Commission
ABOT  aircraft blocked-off time
ACCREP  Accredited Representative
ADF  automatic direction finder
ADS-B  automatic dependent surveillance – broadcast
AFDS  autopilot flight director system
AFM  aircraft flight manual
AIII  HGS Approach III mode (CAT III approach mode or status)
AIP  Aeronautical Information Publication
AMS  Aerospace Material Standards
AOA  angle of attack
APU  auxiliary power unit
AR  Aviation Regulations
ASTM  American Society of Testing Materials
ATC  air traffic control
ATIS  automatic terminal information service
ATM  air traffic management
ATP  Acceptance Test Procedure
ATPL  airline transport pilot’s license
ATS  air traffic services
AV  automatic voice
BEA  Bureau d'Enquêtes et d'Analyses pour la Sécurité de l'Aviation
Civile / Bureau of Investigation and Analysis for Civil Aviation
Safety (France)
C/A  cabin attendant
CA  civil aviation
CAT  category

CDS  Common Display System

CIAIAC  Comisión de Investigación de Accidentes e Incidentes de Aviación
Civil / Civil Aviation Accident and Incident Investigation
Commission (Spain)
CIS  Commonwealth of Independent States
CMM  component maintenance manual
COSPAS-SARSAT  satellite-based search and rescue distress alert detection and
information distribution system
CPL  commercial pilot licence
CPT  captain
CT  computer tomography
CVR  cockpit voice recorder
DC  direct current
DH  decision height
DME  distance measurement equipment
DNA  deoxyribonucleic acid
DVOR  Doppler very high frequency omnidirectional range
E  eastern longitude
EASA  European Aviation Safety Agency
EGPWS  enhanced ground proximity warning system
ELT  emergency locator transmitter
EMERCOM  Ministry of the Russian Federation for Affairs for Civil Defense,
Emergencies and Elimination of Consequences of Natural Disasters
F/O  first officer
FAA  Federal Aviation Administration of United States of America
FAR  Federal Aviation Regulations
FAR-362  Federal Aviation Regulations “Procedure for Radio Communication
in the airspace of Russian Federation”
FATA  Federal Air Transport Agency
FCC  flight control computer
FCOM  Flight Crew Operating Manual
FCT 737 NG (TM)  Boeing 737 aircraft Flight Crew Training Manual
FCTM  Flight Crew Training Manual

FD  flight dispatcher of the airline

FDP  flight duration period
FDR  flight data recorder
FFS  full flight simulator
Fig.  figure
FIR  flight information region
FL  flight level
FMC  flight management computer
FMS  flight management system
FPM  flight plan manager
FR  frame
FSBI  federal state budgetary institution
FSI  federal state institution
FSUAE  Federal State Unitary Aviation Enterprise
FSUE  federal state unitary enterprise
ft  feet
FTOA  Federal Transport Oversight Agency
FWD  forward
G, g  gravity
G/A  go-around
G/S  glideslope
GCAA  General Civil Aviation Authority
GL  ground level
GPS  Global Positioning System
H24  twenty-four hour operation
HF  high-frequency
HGS  Head-Up Guidance System
hPa  hectopascal
HQC  High Qualification Commission
hrs  hours
HUD  head-up display
IA  initial approach
IAC  Interstate Aviation Committee

IAS  indicated airspeed

ICAO  International Civil Aviation Organization
IDG  integrated drive generator
IFR  instrument flight rules
ILS  instrument landing system
IMB  inner marker beacon
IMC  instrument meteorological conditions approach – HGS mode of
operation
kg  kilogram
KRAMS-4  Integrated Aerodrome Radiotechnical Weather Station
kt  knots
kVA  kilovolt-ampere
lb  pound
LH  left-hand
LIM  locator at the inner marker
LLC  limited liability company
LOC  localizer
LOM  locator at the outer marker
LPC  line proficiency check
LVOPS  low visibility operations
MAC  mean aerodynamic chord
MAP  missed approach
MCP  Mode Control Panel
MEL  Minimum Equipment List
METAR  METeorological Aerodrome Report
MHz  megahertz
min  minutes
MRO  maintenance, repair and overhaul
MSN  manufacturer serial number
MTOW  maximum takeoff weight
N  northern latitude
ND  navigation display
NDB  non-directional beacon

nm  nautical mile
NTSB  National Transportation Safety Board
NVM  non-volatile memory
OEI  one engine inoperative
OM  Operations Manual
OMB  outer marker beacon
OMDB  Dubai International Airport ICAO code
OPC  operator proficiency check
P.O.  post office
p/n  part number
PAPI  precision approach path indicator
PCU  power control unit
PF  pilot flying
PFD  primary flight display
PIC  pilot-in-command
PJSC  public joint stock company
PM  pilot monitoring
PRI  HGS primary mode of operation
PWA  printed wiring assembly
PWS  Predictive Windshear
QFE  atmospheric pressure at runway threshold
QNH  atmospheric pressure adjusted to mean sea level
QRH  Quick Reference Handbook
R/T  radiotelephony
RAMC  Rostov-on-Don air meteorological center
RF  Russian Federation
RH  right-hand
RNP  Required Navigation Performance
Roshydromet  Federal Service of Russia for Hydrometeorology and Monitoring of
the Environment
RQC  regional qualification commission
RVR  runway visual range
RWY  runway

s/n  serial number

SAR  search and rescue
SatCom  Satellite Communications
SB  service bulletin
SIGMET  Significant Meteorological Information
SOP  standard operating procedures
SSFDR  solid state flight data recorder
STAR  standard instrument arrival
STC  Supplementary Type Certificate
STD  scheduled time of departure or synthetic training device (by context)
TAF  terminal aerodrome forecast
TAS  true airspeed
TBO  time between overhauls
TCAS  traffic collision avoidance system
TO/GA  takeoff/go-around
TOW  takeoff weight
TQC  territorial qualification commission
TSN  time since new
TWY  taxiway
UAE  United Arab Emirates
UATMS  unified air traffic management system
URRR  Rostov-on-Don Airport ICAO code
USA  United States of America
UTC  coordinated universal time
Vfe  maximum flap extended speed
VMC  visual meteorological conditions
VNAV  vertical navigation
Vref  landing reference speed
vs  versus
WS WRNG  windshear warning

General Information
On March 19, 2016 at 00:42 UTC1, at nighttime, in the progress of the go-around maneuver
after the missed approach to the RWY 22 of the Rostov-on-Don international airport the fatal
accident occurred to the Boeing 737-8KN A6-FDN aircraft (further on referred to as B737-8KN),
operated by the Dubai Aviation Corporation airline (further on referred to under the Flydubai
trading name (UAE) , having performed the international scheduled passenger flight FDB 981 on
route Dubai (OMDB/DXB) – Rostov-on-Don (URRR/RVI).
There were 62 people on board (2 flight crewmembers, 5 cabin crewmembers and 55
passengers), having been the citizens of India, Spain, Cyprus, Colombia, Kyrgyzstan, Russia,
Seychelles, Ukraine and Uzbekistan.
The IAC AAIC was notified of the air accident at 01:22 on 19.03.2016.
To investigate the occurrence the investigation team was assigned by the IAC AAIC
Chairman Order No 9/765-р of 19.03.2016.
In compliance with ICAO Annex 13, the notifications on the air accident were forwarded
to NTSB (USA) representing the aircraft State of Design and State of Manufacture, to BEA
(France) representing the State of Design of the engines, to AAIS (UAE) representing the State of
Registry and the State of the Operator, as well as to the air accident investigation authorities of the
states of the killed occupants’ citizenship. USA, France and UAE have appointed the ACCREPs
to the investigation.
1
Hereinafter, unless otherwise stated, UTC is indicated, the local time is UTC + 3 hrs
The representatives-experts of NTSB, FAA, the Boeing Company, being the aircraft
manufacturer, AAIS, the Flydubai airline, FATA, the Federal Transport Oversight Agency, the
Federal Service for Hydrometeorology and Environmental Monitoring of Russia, the State Air
Traffic Management Corporation, the Rostov-on-Don International Airport, LLC, the Rodina
scientific plant, the Gromov Flight Research Institute, as well as the command and junior flight
crew and instructor personnel of several Russian airlines assisted to the investigation.
To transcript and translate the segment of communications, recorded by CVR, that had
been conducted in Spanish, the investigation team addressed to CIAIAC. CIAIAC has appointed
the ACCREP and granted the necessary assistance.
The investigation was opened on March 19, 2016.
The investigation was closed on November 25, 2019.
In order to render assistance to the affected persons and to the bereaved families, as well
as for the assistance to the post-accident damage control and recovery the Government
Commission, chaired by Minister of Transport of the Russian Federation, has been assigned by the
Chairman of the Government of the Russian Federation Order No 459-р of 19.03.2016.
The pretrial criminal investigation had been conducted by the Russian Federation
Investigative Committee General Directorate for Major Investigations.

1. Factual information
1.1. History of flight
At the overnight into 19.03.2016 the Flydubai airline flight crew, consisting of the PIC and
F/O, was performing the round-trip international scheduled passenger flight FDB 981/982 on route
Dubai (OMDB) – Rostov-on-Don (URRR) – Dubai (OMDB) on the B737-8KN A6-FDN aircraft.
At 18:37 on 18.03.2016 the aircraft took off from the Dubai airport. The flight had been
performed in IFR.
At 18:59:30 FL360 was reached. The further flight has been performed on this very FL.
The descent from FL has been initiated at 22:17. Before starting the descent, the crew
contacted the ATC on the Rostov-on-Don airport actual weather and the active RWY data.
In progress of the glide path descent to perform landing with magnetic heading 218°
(RWY22) the crew relayed the presence of “windshear” on final to the ATC (as per the aboard
windshear warning system activation). At 22:42:05 from the altitude of 1080 ft (330 m) above
runway level performed go-around.
Further on the flight was proceeded at the holding area, first on FL080, then on FL150.
At 00:23 on 19.03.2016, the crew requested descent for another approach.
It was an ILS approach. The A/P was disengaged by the crew at the altitude of 2165 ft
QNH (575 m QFE), and the A/T at the altitude of 1960 ft QNH (510 m QFE). .
In the progress of another approach the crew made the decision to initiate go-around and
at 00:40:50, from the altitude of 830 ft (253 m) above the runway level, started the maneuver.
After the reach of the altitude of 3350 ft (1020 m) above the runway level the aircraft
transitioned to a steep descent and at 00:41:49 impacted the ground (it collided the surface of the
artificial runway at the distance of about 120 m off the RWY22 threshold) with the nose-down
pitch of about 50⁰ and IAS about 340 kt (630 km/h).
1.2. Injuries to Persons
Injuries to persons Crew Passengers Others
Fatal 7 55 0
Serious 0 0 0
Minor/None 0/0 0/0 0/0

1.3. Damage to Aircraft

The condition of the aircraft systems and engines
Fuselage
The fuselage of the airplane is disintegrated in many fragmentary elements, of which the
largest ones are:
 the fragment of the upper section of the fuselage (from FR 460 to FR 500D) with the
GPS antennas;
 the fragment of the RH fuselage section (from FR 927 to FR 986.5) (the area adjacent
to the aft RH service door) and the aft RH service door itself;
 FWD cabin door;
 the fragment of the LH fuselage section (from FR 578 to FR 610) with the fragment of
the overwing emergency exit;
 the fragment of the skin (from FR 500F to FR 540) with the LH window openings;
 the upper fuselage section (from FR 694 to FR 727A) with the ADF antennas;
 the fragment of the pressure bulkhead.
The windshield of the cockpit and passenger cabin is totally destroyed.
The aircraft flight controls and the equipment, located inside the cockpit are almost
completely destroyed (Fig. 1).
Fig. 1. The elements of the cockpit and the nose landing gear leg

The partly survived F/O control column part with the elements of the control wheel was
discovered having been considerably deformed (Fig. 2). The control wheel elements integrated the
partly destroyed trim switch.
Fig. 2. The element of the F/O control column
The elements of the elevator, ailerons and rudder control systems, mounted in the cockpit
underfloor compartment, are presented on Fig. 3.
Fig. 3. The elements of the elevator, ailerons and rudder control systems, mounted in the cockpit underfloor
compartment

Wing
The wing is disintegrated (Fig. 4, Fig. 5).
The LH and RH wingtips (winglets) with the fragments of the lower wing panels, the single
unit tracks, along which the flaps travel, the fragments of the LH outboard and inboard flaps, the
fragments of the skin of the RH wing panel lower section survived.
The wing to fuselage attachment fittings are destroyed.
Fig. 4. The fragments of the RH wing panel
Fig. 5. The fragments of the LH wing panel

Wing high-lift devices (flight control surfaces)

As the result of the aircraft impact with ground, the slats had been totally destroyed, in the
progress of the layout only the smaller fragments of the flaps were retrieved. The transmission of
the flaps, laid along the aft wing spar, is destroyed. All the eight jackscrews and eight single unit
tracks of the flaps retraction/extension mechanism had been discovered (Fig. 6). The elements in
question are mechanically damaged, the screw parts are destroyed and deformed.
Fig. 6. Jackscrews and single unit tracks
It has not been possible to identify the position of the wing control surfaces as of the
moment of the aircraft impact with ground.2
Horizontal and vertical stabilizer (Fig. 7)
The vertical stabilizer, disintegrated in several parts, separated from the fuselage at the area
of attachment. The composite rudder is completely destroyed as the result of the impact with
ground. The skin and the frame of the vertical stabilizer are considerably damaged. There are no
signs of fire and soot on the vertical stabilizer. The horizontal stabilizer had been destroyed and
separated from the vertical stabilizer structure.
2
As per the FDR data, at the moment of aircraft impact with ground the flap handle was at 15⁰ position, flaps - at
10⁰ position due to activation of automatic load relief function, the slats were in the intermediate position.
2
1
Fig. 7. The vertical and horizontal stabilizer elements (1 – LH half of the stabilizer, 2 - RH half of the
stabilizer, 3- the vertical stabilizer elements)
Landing gear
The nose landing gear is destroyed (Fig. 8). The retraction/extension actuator with the
elements of the hinge fittings is separated from the rod. The torsion link is totally destroyed, the

nose landing gear steering actuator cylinders are separated from the standard attachment points.
The nose landing gear wheels attachment pivot and the wheels’ disks are totally destroyed.
Fig. 8. Nose landing gear
The right and left main landing gear legs (Fig. 9 and Fig. 10) are considerably damaged. The
landing gear legs are separated from the fuselage attachment beams. The torsion links of the
shock struts are destroyed. The retraction/extension actuators are separated off the legs main
structure. The disks of the brakes are broken off. The brakes hubs are considerably damaged.
The hubs of the wheels are separated off the tires.

Fig. 9. The elements of the right main landing gear leg
Fig. 10. The elements of the left main landing gear leg

Control System
Pitch control system (elevator and stabilizer)
In the progress of the post-accident inspection of the aircraft wreckage there were retrieved
the PIC and the F/O control columns fragments, the right and the left elevator PCUs, the fractured
stabilizer jackscrew with the cable drum and the elevator PCUs output shaft.
The cable control linkage is totally defragmented subsequent to the aircraft impact with
ground and is impossible to identify.
The fragments of the elevator control system (the input rods, the PCUs and the elevator
autopilot electrohydraulic actuator) (Fig. 11) are significantly damaged as the structure had
disintegrated at the aircraft impact with ground.
Fig. 11. The elevator control system output shaft (pointed with arrow)
The left elevator PCU is separated off the output shaft and partially broken, the right PCU
rests at its standard attachment point.
The elevator control kinematics can be seen deformed as the result of the structure
disintegration after the aircraft impact with ground.
In the progress of the accident site inspection, the majority of the stabilizer trim actuator
gearbox had been retrieved. The aft cable drum had been separated from its connection point on
the stab trim actuator. The drive mechanism had been damaged and deformed.
The fragment of the crossbeam with the stab trim mechanism ball nut had been discovered
as well. The crossbeam was disintegrated at the points of its attachment to fuselage as the result of

the off-design loads application. The jackscrew with the attachment fitting to stabilizer (Fig. 12)
had been discovered apart of the stabilizer being disintegrated due to the off-design loads.
Fig. 12. The recovered elements of the stab trim actuator
Roll control system

The cable linkage, passing in the fuselage, was disintegrated into separate smaller
fragments at the aircraft impact with ground and is impossible to identify. The cable linkage, laid
along the LH and RH half-wing aft spar, survived with the signs of the off-design mechanical loads
application and the fire.
The LH and RH ailerons are almost completely destroyed.
The spoilers control system is totally destroyed. As for the system elements, there were
retrieved several actuators and the control surfaces, whose position as of the moment of the
accident is impossible to identify.
The directional channel of the control system
The pilots’ rudder pedals are completely disintegrated. The cable control linkage, passing
in the fuselage is broken off and is impossible to identify. The main control actuator is torn off its
standard mounting because of the vertical stabilizer disintegration.
Hydraulic system
The hydraulic pumps of the main hydraulic systems A and B as well as the pumps of the
auxiliary hydraulic system are totally disintegrated, so are the hydraulic reservoirs and
accumulators. In the progress of the aircraft structure layout, only the smaller fragments of the
hydraulic lines had been retrieved.

The aircraft fuel system

The integral fuel tanks, the elements of the aircraft fuel system, the assemblies and pipes
are totally disintegrated and burnt out in a ground fire, the seat of which had been located at the
area of the fuel tanks.
The air conditioning and pressurization system
The assemblies of the air conditioning system are completely destroyed. At the inspection
of the accident site, only the recirculation system deformed fan had been retrieved.
The pressurization control assemblies are totally disintegrated as well. In the progress of
the accident site inspection only the outflow valve out of this system had been discovered (Fig. 13).
Fig. 13. The air conditioning system outflow valve
Water supply system and catering equipment

The water supply system assemblies are completely destroyed.
The cabin dividers and the overhead bins had been lost. The passenger seats are
disintegrated into the smaller fragments amounting to the passenger seats units’ bases and frames.
The seatbelts are impossible to identify. The galleys equipment is completely destroyed as well.
APU
The APU is disintegrated (Fig. 14). The inspection of the APU gas generator plumbing and
wiring assemblies did not reveal the ducting and control elements uncoupling.
Fig. 14. The APU fragment with the exhaust duct and the fuselage tailcone
Engines
The left and the right engines are totally destroyed (Fig. 15, Fig. 16).

Fig. 15. The left engine fragments
Fig. 16. The right engine fragments

In the progress of the accident site inspection there were retrieved the fragments of the left
and the right engines: the compressor cases with the elements of the compressor control system
and the engines rotor front support, the disks of the of the engines fan stages.
Avionics
At the on-site examination of the fragments of the airframe, cockpit, passenger cabins and
service bays it has been stated that subsequent to the aircraft impact with ground and the further
ground fire the avionics is significantly damaged or totally destroyed.
Autopilot
The avionics units, integrated to the autopilot, are completely destroyed and impossible to
identify.
The two out of four autopilot actuators had been discovered.
Radio communication equipment
The radio communication units are completely destroyed and are impossible to identify.
There were retrieved the fragments of the HF radio set switching unit (p/n 822 0990 004,
s/n 1871X9) (Fig. 17).
Fig. 17. The HF radio set switching unit fragments
Electrical power supply system

The electrical power supply system integrates two alternators, located on the aircraft
engines - the IDG1 and IDG2 integrated drive generators, having a capacity of up to 90 kVA, the
APU starter-generator with the capacity of up to 90 kVA. The direct current secondary electrical

supply system consists of three transformer-rectifier units. The electrical battery is operated as the
direct current emergency power supplies. The generators and electrical battery are completely
destroyed subsequent to the impact. There was discovered the rotor of one of the generators
(Fig. 18).
Fig. 18. The generator rotor
Several integral units to the electrical system were recovered:

 the standby power control unit (SPCU) (Fig. 19);
Fig. 19. The standby power control unit
 the transformer-rectifier unit (TRU) (Fig. 20);

Fig. 20. The transformer-rectifier unit
 the bus power control units (BPCU) (Fig. 21, Fig. 22).
Fig. 21. The bus power control unit No 1

Fig. 22. The bus power control unit No 2
All the units sustained considerable damage.

At the site of the aircraft impact with ground the debris of the aircraft electrical supply were
retrieved along the impact heading (Fig. 23). The aircraft electrical supply is totally destroyed,
there were no signs revealed of the short circuit, might have been evident in the electrical wiring
isolation melt.
Fig. 23. The debris of the aircraft electrical supply
Fire protection system

The fire protection system comprises two cylinders of the engines bays, one APU bay fire
extinguishing cylinder and three control handles in the cockpit.
The fire protection system components had not been discovered due to their total
destruction in the accident sequence.
Flight management system
The part of the FCC case was retrieved. The circuit boards were not discovered.
The stab trim motor, p/n 6355C0001-01, s/n 2062had been retrieved (Fig. 24).
Fig. 24. The stab trim motor elements
The oxygen supply equipment

The flight crew main stationary oxygen cylinder had been retrieved (its mounting location
is in the electronics and avionics bay). Based on the nature of its destruction it can be concluded
that the cylinder had been destructed in the impact sequence (Fig. 25).
Fig. 25. Main stationary oxygen cylinder

Navigation equipment
The navigation assemblies and units, revealed at the accident scene, are totally
disintegrated.
The fragments of the fuselage skin with the GPS and TCAS antennas had been discovered
(Fig. 26).
Fig. 26. The GPS and TCAS antennas
On the accident site there were retrieved as well:

 the EGPWS unit (Fig. 27);
Fig. 27. The EGPWS unit
 the TCAS computer, p/n 940 0300 001, s/n TPA03476 (Fig. 28);

Fig. 28. The TCAS computer
 the standby altimeter (Fig. 29).
Fig. 29. The standby altimeter
All the units have the signs of the physical damage on them.
1.4. Other damage

As the result of the accident 332.5 m² of the RWY concrete surface at the area of its left
edge, at about 150 – 200 m off the RWY 22 threshold, sustained damage. Besides the aerodrome
lighting between TWYs D and C was damaged, five LH edge lights and two RH edge lights,
namely, as well as two axial low-voltage cables together with four edge lights axial low-voltage
cables.
1.5. Personnel information

1.5.1. Flight crew information
Position Pilot-in-command
Sex Male
Date of birth August 05, 1978
Pilot’s license (ATPL) No 51549
License date of issue November 7, 2012
Issued by General Civil Aviation Authority (GCAA) of the
License expiry date Valid till January 25, 2023
Medical report I class medical certificate No 51549, issued on
September 30, 2015, valid till October 14, 2016
Low visibility qualification ICAO CAT IIIA
Total flying time as a pilot 5965 hrs
Transition training for Boeing 737-800 October 27, 2012
Flying time on the Boeing 737-300-900 4682 hrs/1056 hrs
aircraft/of which as a PIC
Flying time within a year (2015) 843 hrs
Flying time within 28 days 78 hrs 36 min
Flying time within 2 days 8 hrs
Flying time within last 24 hours (on the day 06 hrs 05 min
of the accident)
Duty time as of the moment of the accident 07 hrs 57 min
The intervals in flights within the last year The PIC’s leave periods within 2015:
March 19-25/19.03-25.03 – 7 days;
April 9-15/09.04-15.04 – 7 days;
June 25 – July 1/25.06-01.07 – 7 days;
July 30 – August 12/30.07-12.08 – 14 days;
October 8-14/08.10-14.10 – 7 days.
42 days in total in compliance with OM
The date of the last line check December 18, 2015

Simulator training and check January 7, 2016

Preflight preparation Self-preparation before departure at the Dubai
airport
The crew rest 15 hrs at home
English language proficiency ICAO level 6
Air accidents and incidents in the past None
Based on the data, submitted by the Flydubai airline, before April 2004 the PIC had been
performing flights on the aircraft under 5700 kg. There are no data available on these aircraft types.
Further on he underwent the Boeing 737-300/400/500 aircraft transition training and was
employed by the Helios Airways airline (Cyprus) where he was working as the first officer from
April 23, 2004 until November 29, 2006.
From December 23, 2006 until September 12, 2008, he was working as the Boeing 737-
300/400/500 first officer in the XL Airways (Great Britain).
From February 12, 2011 until September 12, 2012 the PIC was working as the Boeing 737-
300/400/500 in the Malaysian Airlines System airline (Kuala Lumpur, Malaysia).
On September 30, 2012, he was employed by the Flydubai airline as the first officer.
Through October 14-27, 2012, he has completed the Boeing 737-800 aircraft transition
training in the Emirates - CAE Flight Training air training center (Dubai, United Arab Emirates)
and was issued a certificate No EK-CAE/DXB/148.
In January 2015, he underwent the Boeing 737-800 command upgrade training.
Position First officer

Sex Male
Date of birth March 23, 1979
Pilot’s license (ATPL) No 66543
License date of issue October 21, 2014
Issued by General Civil Aviation Authority (GCAA) of the
License date of expiry October 20, 2022
Medical report I class medical certificate No 66543, issued on
August 31, 2015, valid till September 14, 2016
Low visibility qualification ICAO CAT IIIA
Total flying time as a pilot 5767 hrs
Flying time/ of which as a PIC Cessna- 421, ATR-42/72, A-320 –
4667 hrs/258 hrs

Boeing 737-800 aircraft transition training October 20, 2014

Flying time on the Boeing-737-800 1100 hrs
aircraft type
Flying time within a year (2015) 784 hrs
Flying time within the last 24 hours (at the 6 hrs 05 min
day of the accident)
Duty time as of the moment of the 7 hrs 57 min
accident
The intervals in flights within the last year The F/O’s leave periods within 2015:
February 26 – March 4/26.02-04.03 – 7 days;
May 21-27/21.05-27.05 – 7 days;
September 17 - October 07 – 21 days;
November 20-26/20.11-26.11 – 7 days.
42 days in total in compliance with OM
January 01-07, 2016/01.01.-07.01.2016 – 7 days
The date of last line check October 20, 2015
The simulator training and check September 7, 2015
Preflight preparation As per briefing system under the authority of the
PIC
The crew rest 20 hrs at home
English language proficiency ICAO level 5
Air accidents and incidents in the past None
Based on the data submitted by the Flydubai airline and by Spanish Aviation Safety and
Security Agency (Agencia Estatal de Seguridad Aérea – AESA) the first officer on May 29, 2000,
after graduating from SENASA, Services and Studies for Air Navigation and Aeronautical Safety
air training center, was issued a CPL with the authorization to operate single- and multi-engine
land airplanes in IFR.
From June 1, 2005 till September 30, 2005 he was working as a Cessna-421 aircraft PIC at
the Regional Geodata regional carrier (Spain).
From May 31, 2006 till March 10, 2008 he was employed as the ATR-42/72 aircraft first
officer at the Islas Airways airline (Spain).

From February 25, 2009 till April 17, 2014 he was working as the ATR-42/72 aircraft first
officer, PIC, A3203aircraft first officer.
On March 25, 2009, he was issued an ATPL.
On May 8, 2012, he was authorized to operate ATR-42/72 aircraft as a PIC.
On August 24, 2014, he was employed by the Flydubai airline as a Boeing 737-800 first
officer.
From September 7, 2014 till October 20, 2014 he completed the Boeing 737-800 aircraft
transition training at the Emirates - CAE Flight Training air training center (Dubai, United Arab
Emirates) and was issued a certificate No EK-CAE/DXB/B-077.
On February 26, 2015, he was authorized to operate the Boeing 737-300/900 aircraft.
The analysis of the submitted documents
According to the airline OM the duty cycle shall not exceed:
over 28 consecutive days – 100 hrs of flight time and 190 hrs of duty time.
The minimum rest between flying duty periods at the base aerodrome shall be not less than
twelve hours.
The actual data on the flight crew flight time and duty time as well as the rest time are
given in the below table.
PIC F/O
Flying time over last 28 days: 78 hrs Flying time over last 28 days: 80 hrs
Duty time over last 28 days: 136 hrs Duty time over last 28 days: 139 hrs
The preflight rest: 15 hrs The preflight rest: 20 hrs
As per their health status, the crew were approved for flight operations, had a sufficient
rest period, held the valid pilot’s licenses and underwent all the necessary procedures to operate
solo (without the mandatory presence of the flight instructor) flights.
As per the submitted documents, the Flydubai airline the flight crews training is carried out
at the certified simulators in compliance with the Program, drawn up with the FCOM as a core
document that determines the standard operating procedures.
The go-around performance SOP are specified in FCOM, volume 1 (NP.21.54) with the
additional instructions in the Flydubai Procedures and Policies manual (Section 12 Go-Around
and Missed Approach and Appendix F (Standard Callouts)). These two documents are in use
together with the FCTM guidelines. Furthermore, the Upset Recovery Training Aid, second
edition of November 2008, having been relevant as of the date of the accident, integrates the
additional recommendations on the go-around performance.
3
On May 10, 2013, he completed the A320 transition training at the Baltic Aviation Academy air training center.
The go-around procedure in the initial training

Once the pilots are employed by the airline, they undergo the initial training and a
Conversion course (for type-rated pilots) or a full Type Rating course (for non-type-rated pilots).
The syllabuses for both courses include go-arounds flown both with two engines operative
and with one engine inoperative.
Competency demonstration for all pilots occurs at the License Skill Test (the UAE pilot’s
license).
Go-around in the progress of the HUD training
All the aircraft in the airline fleet are equipped with the HGS (Head-Up Display Guidance
System) Model 4000 system that integrates HUD4 as a display unit. According to the airline SOP
(Annex D, Section D.1.2) the use of HUD, if it is operative, is mandatory throughout the entire
flight.
HUD is a supplementary cockpit onboard avionics unit, designed by Rockwell-Collins
(currently Collins). Rockwell-Collins had certified the equipment in question and was issued the
Supplementary Type Certificate/STC ST00845SE, which, for the Boeing 737-800 aircraft,
involves the installation of the only HUD at the PIC’s duty station. This Certificate does not
provide for the dual installation (including the F/O duty station) 5. The HGS was installed on the
aircraft by the Boeing Company on request of Flydubai prior to its delivery.
Before starting line operations at Flydubai all pilots undergo a Head Up Display (HUD)
training dedicated simulator session that includes:
 two go-arounds with two engines operative;
 two go-arounds with one engine inoperative.
Pilots will fly the maneuver twice, one time as Pilot Flying (PF) and repeat as Pilot
Monitoring (PM), the go-arounds start at different distances from the runway (from the different
altitude) as recorded in relevant training form.
Irrespective of the fact that all the Flydubai aircraft have the HUD fitted only on the left
side (the PIC side) all pilots, including first officers, go through the HUD training simulator
session.
The PIC underwent the training and simulator session on October 27, 2012.
The First Officer underwent the training and simulator session on October 18, 2014.
There were no comments stated by the instructor staff on the pilots’ simulator session go-
through.
4
Hereinafter both terms are referred to as synonyms.
5
Currently, as for the HUD Model 6000 (STC ST02522SE) and for the Boeing 737NG and MAX, the option is
certified of its installation on both flight crew’s duty stations.
Go-around during low visibility operations initial training and check

During the Initial LVOPS training simulator session the flight crew members perform go-
around maneuvers both with two engines operative and one engine inoperative.
The altitude at which the maneuver starts varies along the training, ranging from around
1000 feet/300 meters to DH.
Current Low Visibility Initial Simulator Training program includes go-around training as
follows.
Go-arounds with two engines operative (12 go-arounds in total):
 7 go-arounds due different systems failure occurring between 1000 feet and the
appropriate minima/ DH;
 1 go-around due insufficient RVR;
 1 go-around due windshear;
 1 go-around due excessive deviation from the G/S;
 1 go-around due HGS/HUD failure with reversion for conventional instruments (PFD);
 1 go-around due pilot incapacitation.
Go-arounds with one engine operative (2 go-arounds in total):
 1 go-around due insufficient RVR;
 1 go-around due system failures.
Go-around during recurrent training and operator and license proficiency checks
(OPCs/LPCs)
Within the recurrent training and OPC/LPC (carried out twice a year in compliance with
the OM D1), the pilots underwent the go-around training that included:
 1 go-around with one engine inoperative from minimum altitude;
 1 go-around on RNP approach with failure requiring go-around (two or one engine
operative);
 1 go-around from low visibility approach due engine failure, incapacitation or system
failure (two engines or one engine operative).
As of the moment of the accident, in the Flydubai airline there were no simulator sessions
carried out on go-around from different altitudes with low aircraft weight.
Windshear during initial training
Similar to the go-around training, windshear is included in both Conversion courses and
Type Rating courses and all pilots joining Flydubai will undergo one or the other course.
Before starting line operations at Flydubai, all pilots undergo ground training course which
includes theoretical training on:
 windshear avoidance and recovery;
 windshear avoidance and recovery from predictive and actual windshear using the
HUD.
On completion of the ground training, the pilots undergo the windshear HUD training
session that implies the training both for PF and PM.
Windshear during low visibility operations initial training and check
Low Visibility Initial Simulator Training program includes windshear. Captains are trained
using the HUD and first officers use conventional head-down instruments (PFD).
Windshear during recurrent training
Windshear training is present in the Recurrent Airplane/STD Training Schedule B737‐800
(OM D1 Appendix 9.A). This accounts for the inclusion of windshear training in the Recurrent
Training program at least twice in a 3-year cycle during recurrent training, done every six months.
The training is based on FCOM, QRH and FCTM guidance for windshear avoidance,
precautions and recovery.
In the low visibility recurrent checking program (every six months), at least one go-around
(two engines) is conducted due to either insufficient RVR or windshear.
The PIC underwent the windshear training on July 6, 2015.
The first officer underwent the windshear training on September 7, 2015.
Manual stabilizer trim operation
Malfunctions requiring manual operation of the stabilizer trim are trained in both
conversion course and full type rating course pilots go through before starting line operations in
the Flydubai airline.
The recurrent training on the flight controls and trim system failures is specified in the
respective section in the Flydubai’s Recurrent Airplane/STD Training Schedule B737‐800
(OM D1 Appendix 9.A).
The upset recovery training
Similar to the training on the operations at the windshear within the Conversion and the
Type Rating training, all the pilots employed by the Flydubai airline, before they start performing
flights, undergo the four-day ground training that includes the theoretical course in upset avoidance
and recovery.
On completion of the ground training the pilots undergo the HUD/HGS simulator session
that involves the flight training both for pilot flying and for pilot monitoring.
The upset recovery training is integrated in the Boeing 737-8KN flight crew members
recurrent training (OM D1 Appendix 9.A) and is conducted triply within a three-year cycle,
divided into 6 six-month sessions.
The training is based on the FCOM , QRH and FCTM guidance on the upset recovery.

The aircraft’s PIC underwent the upset recovery training on January 7, 2016.
The first officer underwent the upset recovery training on September 6, 2015.
The PIC and the first officer underwent the FFS recurrent training at regular intervals in
compliance with the airline OM.
PIC’s recurrent training and checks
Date Training type Comments6
July 5-6, 2015 Recurrent training and 1. First attempt flown within limits but
proficiency check became un-stabilized around 500FT. G/A
carried out followed by debriefing and
repetition to an acceptable standard.
2. Speed was below limits for the existing
flaps, the PIC failed to recognize the dual
GPS failure. Approach was discontinued
later as per ATC instruction. Repetition to
satisfactory standard.
3. Remember the proper call out to set the
go around thrust as per FCOM I and QRH.
4. Windshear at takeoff.
TOGA not pressed at first alert, and not
pressed again after 400ft while VNAV used
during takeoff. Briefed and repeated.
January 6-7, 2016 Recurrent training and Simulator instructor comments:
proficiency check 1. For both Normal and Non Normal
checklists brief the approach before calling
for the Descent checklist.
2. Arrival procedure – If cleared direct to a
beacon as part of an arrival (IA) be sure to
have it identified.
3. On the first attempt /name/ did not
sufficiently arrest his descent rate when
visual, thereby generating an APCH WARN
(approach warning) and hard landing. The
second maneuver was to training standards.
6
Only the comments, related to the circumstances of the accident, are stated
F/O recurrent training and checks

Date Training type Comments
March 17-18, 2015 Recurrent training and 1. Leadership and teamwork
proficiency check Needs to be quite a bit more assertive in
what is needed from the Captain. Tell
him/her what you want done and do not
wait for the Captain to enquire with you or
direct you in this regard. Need to be more
decisive in taking actions when needed.
2. Go-around maneuver
Use caution on directional control during
OEI MAP. Be aware that the TOGA mode
shows wings level information and is not
linked to runway heading or track.
3. Always check the approach RNP when
setting up/briefing
September 6-7, 2015 Recurrent training and 1. High altitude stall
proficiency check Must maintain pitch altitude better during
recovery, secondary stick shaker occurring.
2. Approach
Did not descend until late turning base and
stopped turn causing aircraft to fly through
centerline. Did not recover in time but flew
approach to landing. Debriefed and flown
to a high standard.
3. Go-around
No call for G/A thrust. Debriefed
1.5.2. Ground service personnel information

Position The Rostov-on-Don aerodrome air operations
supervisor
Sex Male
Date of birth April 15, 1976
Education Higher, Ulyanovsk Higher School for Civil
Aviation in 1998

Class I class. The RF Ministry of Transport FTOA CA

HQC work team protocol No 8 of 24.06.2005
Duty authorization Authorized for duty as an air operations
supervisor – the FATA Southern department RQC
protocol No 3 of 11.04.2011
Duty assignment The State ATM Corporation South Air Navigation
branch general manager order No 19/ок of
04.02.2014.
ICAO language proficiency ICAO level 4 – certificate РНД No 1219.7619,
valid till 28.01.2017
Qualification and English language At the Saint-Petersburg State University for CA
advanced training air training center (the city of Saint-Petersburg) as
per the program «The qualification and English
language advanced training for the ATC air
operations supervisors», 27.12.2014
Medical certificate validity Till 16.10.2016
License, date of issue СД No 019275, issued on 21.12.1998 by Federal
Aviation Service of Russia
License validity Till 05.11.2016
Competence check Air operations supervisor – 16.02.2016
Simulator training 04.03.2016 – radar control unit
05.03.2016 – approach control unit
Total in-service time 16 years
Total in-service time as an air operations 2 years
supervisor
Position Control tower chief officer (at the day of the

accident functioned as the radar control unit
officer)
Sex Male
Date of birth November 5, 1966
Education Secondary professional, Riga flight-technical
college for CA in 1989
Class I class. The RQC protocol No 5 of 12.03.1997

Duty authorization Authorized for duty as a chief controller – The

RQC protocol No 104 of 06.12.2009
Duty assignment The State ATM Corporation South Air Navigation
branch general manager order No 427/ок of
19.04.2010
Qualification and English language At the Institute for air navigation south branch
advanced training (the city of Rostov-on-Don):
– as per the program «The ATC officers training
(for chief controllers, instructor controllers,
simulator instructor-controllers)», on 27.03.2015;
– the aviation English language (R/T phraseology
for the ATC personnel, authorized for the
international air operations support in the English
language), on 14.04.2015
Competence check Chief controller – on 21.02.2016
Simulator training Approach control unit – on 28.02.2016,
Radar control unit – on 05.03.2016
Total ATC in-service time 26 years
Total in-service time as a chief controller 5 years 11 months
Position Aerodrome control center officer (at the day of

the accident functioned as an approach control
unit officer, further on referred as APPR-1)
Sex Female
Date of birth July 19, 1975
Education Higher, Rostov-on-Don State University in 2006,
Saint-Petersburg air transport college in 2012

Class III class. FATA HQC ATC personnel qualification

subcommission Decision of 30.10.2013
Duty authorization Approach control unit – on 05.11.2013
Radar control unit – on 28.02.2016
ICAO language proficiency ICAO Level 4 – certificate СПБ No 1324.18026,
– as per the program «ATC personnel training (for
the ATM controllers)», on 29.05.2015;
Competence check Approach control unit – on 13.10.2015
Simulator training Approach control unit – on 27.02.2016
Total ATC in-service time 2 years 4 months
The assessment of the language proficiency level of the approach control unit officer is
given in Section 1.16.6.

the accident functioned as the radar control
unit officer, further on referred as RDR-1)
Sex Male
Date of birth September 1, 1985
Education Higher, Saint-Petersburg Academy for CA in
2008
Class II class. Southern FATA interregional territorial
department RQC protocol No 17 of 20.02.2013
Duty authorization Radar control unit – on 12.11.2015

ICAO language proficiency ICAO Level 4 – certificate РНД No 217.2017,

Competence check Radar control unit – on 06.11.2015
Simulator training Radar control unit – on 16.02.2016

the accident functioned as the approach
control unit, the radar control unit officer)
Sex Male
Education Higher, Saint-Petersburg State University for CA
in 2010
department TQC protocol No 114 of 05.12.2013
Duty authorization Approach control unit – on 15.03.2011; radar
control unit – on 24.01.2013
tower control unit – on 12.10.2015; ground control
unit – on 14.01.2016
ICAO language proficiency ICAO level 4 – certificate issued by MLS
International College (the city of Bournemouth,
Great Britain), valid till 30.10.2018

Qualification and English language At the Institute for air navigation south branch (the
advanced training city of Rostov-on-Don):
– as per the program «The air operations support
in the English language (for the ATM
controllers)», on 08.10.2013;
for the State ATM Corporation, FSUE, ATC
personnel that underwent training at MLS
International College (the city of Bournemouth,
Great Britain), on 06.11.2015
Competence check Approach control unit – on 04.06.2015, radar
control unit – no data available
Simulator training Radar control unit – on 10.03.2016; approach
control unit – on 11.03.2016
Position The aerodrome control center officer (at the

day of the accident functioned as the tower
control unit officer, further on referred as
TWR-1)
Sex Male
Education Higher, Yeysk Air Force Higher Flight School in
2008
Saint-Petersburg State University for CA in 2012
Class II class. The Southern FATA Interregional
Territorial Department Decision No 66 of
22.12.2015
Duty authorization Tower control unit on 26.06.2014
Ground control unit – on 15.12.2015


– as per the program «Air traffic management (for
the ATC controllers, authorized for the
Competence check Tower control unit – 19.08.2015
Simulator training No data available
Total ATC in-service time 1 year 8 months
Position The aerodrome control center officer (at the

day of the accident functioned as tower control
unit officer)
Sex Male
Date of birth April 12, 1961
Education Higher, Yeysk Air Force Higher Flight School in
1984
department TQC subcommission protocol No 30
of 10.04.2014
Duty authorization Tower control unit – on 22.09.2010
Ground control unit – on 10.02.2012


Competence check Tower control unit – on 19.08.2015
Simulator training No data available
On 18.03.2016, at nighttime, the air traffic management within the Rostov-on-Don

aerodrome air traffic service area of responsibility was carried out by shift No 5 under the authority
of the Rostov-on-Don aerodrome air operations supervisor.
The shift personnel in full underwent the pre-shift medical inspection prior to the duty
changeover. The shift underwent a briefing in full.
The following control units exercised the FDB981 flight support:
 approach control unit;
 radar control unit;
 tower control unit.
The air traffic service/management, in compliance with the regulations in effect, is carried
out by the one ATC controller at the main or standby duty station.
In the case the necessity arises to engage the relief personnel, the controller who effects the
engagement in question, shall take the duty over according to the ATC officer SOP. The
monitoring and the responsibility for the relief personnel engagement is held by the air operations
supervisor (the chief controller).
At the point of the duty handover/takeover the radio communication with the aircraft flight
crews is conducted by the controller who hands the duty over until the personal password is entered
to the Alfa ATC automated system and the air operations supervisor reports the following (to the
chief controller): «Controller (surname) has taken the duty over» and the respective entry is
introduced by the controller who takes the duty over to the duty takeover book.
Within a period when the A6-FDN aircraft was approaching for landing, the following
officers functioned as the main controllers at the control units: APPR-1, RDR-1, TWR-1, who
were relieved by other controllers when required under the clearance of the air operations
supervisor.
1.6. Aircraft information
Aircraft type Boeing 737-8KN

Manufacturer The Boeing Company (Seattle, USA)
MSN 40241
Date of manufacture 19.01.2011
Registration A6-FDN
Registration certificate No 06/11 of 19.01.2011, issued by GCAA of UAE
Airworthiness certificate ARC-FZ-FDN-4 of 10.01.2016, issued by GCAA
of UAE, valid till 18.01.2017
TSN 21257 hrs, 9421 landings
Service life and life limit Not assigned by the manufacturer, the aircraft was
operated on condition
TBO and service life Not assigned by the manufacturer, the aircraft was
operated on condition
Last scheduled maintenance Base check, carried out on 21.01.2016 at TSN
20656 hrs, 9161 cycle at the facilities of the
certified MRO company
Last line maintenance At the amount of daily check – it was carried out
on 18.03.2016 at the Dubai airport by the Flydubai
airline maintenance personnel at the
TSN 21247 hrs and 9419 cycles (logbook page
No FDN02415).
Prior to departure from the Dubai airport to the
Rostov-on-Don airport the flight en route Dubai –
Kiev – Dubai had been performed. The works
were carried out at the amount of Pre-Departure
Inspection/PDI (logbook page No FDN02417 and
No FDN02418).
On 18.03.2016 prior to departure from the Dubai
airport to the Rostov-on-Don airport the Flydubai
airline maintenance personnel had carried out the
last line maintenance at the amount of PDI.
The CFM International engines manufactured at Snecma Corporation, Centre de

Villaroche, were installed on the accident aircraft.

Engines Left (No 1) Right (No 2)

Type CFM56-7B27/3B1F CFM56-7В27/3B1F
MSN 804660 804538
Date of manufacture January 2011 November 2010
Date of installation on the 12.07.2014 03.04.2014
aircraft
Service life 20000 cycles 20000 cycles
TSN, hours/cycles 20284/8977 20994/9375
Remaining service life 11023 cycles 10625 cycles
TBO On condition On condition
The APU data: model 131-9В, manufactured by Honeywell International Inc. Engines &
Systems on 07.12.2010, MSN 3800702-1, s/n Р-8816, TSN – 15011 hrs, 10998 cycles, it was
repaired once on 24.09.2013.
1.7. Meteorological information

On 18.03.2016, at 17:32, at the briefing at the Dubai airport the flight crew had been
provided by the meteorological documentation as follows: the TAF and METAR reports as for the
Dubai departure aerodrome, the Rostov-on-Don destination aerodrome and the alternate
aerodromes of Trabzon, Volgograd, Krasnodar, Mineralnye Vody; the FL300, FL340, FL390
weather and the wind forecast charts together with the London World Area Forecast Center
significant weather charts through FL100-450, FL250-630 of specific times of 18:00 on
18.03.2016 and 00:00 on 19.03.2016, two vertical profile charts, drawn up at 12:00 on 18.03.2016
for a term of 06 and 12 hrs.
The weather package did not contain the Rostov-on-Don FIR SIGMET No 6 information,
valid through 17:30 – 21:00 on 18.03.2016, drawn up on 18.03.2016 at 17:04 and transmitted at
17:06 to the Vienna and Brussels Regional weather data banks.
URRV SIGMET 6 VALID 181730/182100 URRR-
URRV ROSTOV FIR SEV TURB FCST N OF N48 W OF E048 SFC/FL150 STNR NC
SIGMET 6 forecast valid through 17:30 on 18.03.2016 – 21:00 18.03.2016: stationary
severe turbulence to the south of 48° N and to the west of 48° E from GL up to FL150, no change.
The Rostov-on-Don destination aerodrome METAR-coded actual weather as of 17:00 on
18.03.2016.
METAR URRR 181700Z 24011G16MPS 8000 -RA BKN020 OVC100 06/03 Q1003
R22/290057 TEMPO 25013G18MPS 1000 SHRA BR SCT003 BKN020CB RMK QFE744/0993.

Surface wind 240° 11 m/s, the gusts of 16 m/s, the visibility 8000 m, light rain, broken
clouds (5-7 octants), the cloud base7 600 m, overcast (8 octants), cloud base 3000 m, air
temperature 6°, dew point 3°, QNH 1003 hPa; RWY22: friction coefficient 0.57; temporary
surface wind 250° 13 m/s, gusts 18 m/s, visibility 1000 m, rain shower, mist, cloud: scattered
(3-4 octants), cloud base 90 m, broken (5-7 octants) cumulonimbus, cloud base 600 m,
QFE 744 mm of mercury/993 hPa.
The TAF report of the Rostov-on-Don destination aerodrome of 18.03.2016:
TAF URRR 181401Z 1815/1915 25007G13MPS 9999 SCTC010 SCT020CB TEMPO
1815/1821 25012G18MPS 3000 -SHRA BR SCT005 BKN020CB FM182100 26010G17MPS 3000
–SHRA BR SCT005 BKN020CB TEMPO 1821/1906 28017G25MPS 1000 SHRA BR SCT003
BKN020CB TEMPO 1906/1915 30011G17MPS.
The forecast is drawn up on 18.03.2016 at 14:01, valid through 15:00 on 18.03.2016 –
15:00 on 19.03.2016: surface wind 250° 7 m/s, gusts 13 m/s, visibility 10 km or more, cloud:
scattered (3-4 octants), cloud base 300 m, scattered (3-4 octants) cumulonimbus cloud base 600 m;
temporary through 15:00 of 18.03.2016 – 21:00 of 18.03.2016: surface wind 250° 12 m/s, gusts
18 m/s, visibility 3000 m, light shower rain, mist, cloud: scattered (3-4 octants), cloud base 150 m,
broken (5-7 octants) cumulonimbus, cloud base 600 m; from 21:00 on 18.03.2016: surface wind
260° 10 m/s, gusts 17 m/s, visibility 3000 m, light shower rain, mist, cloud: scattered (3-4 octants),
cloud base 150 m, broken (5-7 octants) cumulonimbus, cloud base 600 m; temporary through
21:00 on 18.03.2016 – 06:00 on 19.03.2016: surface wind 280° 17 m/s, gusts 25 m/s, visibility
1000 m, rain shower, mist, cloud; scattered (3-4 octants), cloud base 90 m, broken (5-7 octants)
cumulonimbus, cloud base 600 m; temporary through 06:00 – 15:00 on 19.03.2016: surface wind
300° 11 m/s, gusts 17 m/s.
At 18:37 on 18.03.2016, the FDB981 flight departed from the Dubai aerodrome to Rostov-
on-Don. According to the forecast the wind of 270° about 200 km/h had been expected on FL360
at the first half of the route; at the second half of the route the 300° 80 km/h wind has been
forecasted. The en route forecast on the FL100-450 significant weather chart indicated the jet
stream of 270° and 280 km/h at the Shiraz – Isphahan area, at the Mineralnye Vody – Rostov route
leg the forecast indicated the moderate icing below FL150, moderate turbulence at the layer below
FL150.
The air operations meteorological support within the area of responsibility of the Rostov-
on-Don UATMS area center (according to the UAMTS Rostov area center Instruction on the
meteorological service, approved on 19.05.2015 by the Aviamettelecom of Roshydromet, FSBI,
7
Hereinafter at the reference to the altitude of cloud base, the value is computed from the GL. The elevation of the
Rostov-on-Don aerodrome is 85.6 m (280 ft).
North Caucasian branch General Manager and by the State ATM Corporation, FSUE, South Air
Navigation branch General Manager) and at the ATM Rostov division area (according to the
Instruction on the meteorological service to air operations at the Rostov-on-Don international
airport, approved on 29.05.2015 by the Rostov-on-Don airport, LLC Executive director) had been
accomplished by the duty shift of the Aviamettelecom of Roshydromet, FSBI, North Caucasian
branch Rostov-on-Don air meteorological center (further on referred to as RAMC), licensed by
Roshydromet on 26.03.2014, registration number Р/2014/2527/100/Л.
On 18.03.2016 at 20:10, the RAMC issued the next SIGMET No 7 information on the
anticipated significant weather – the severe turbulence within the area of responsibility of the
UAMTS Rostov-on-Don area center that was transmitted to the Meteo hardware-software complex
of the Rostov-on-Don UAMTS area center and to the weather data banks.
URRV SIGMET 7 VALID 182100/190100 URRR-
URRV ROSTOV FIR SEV TURB FCST NOF N44 W OF E048 SFC/FL150 STNR NC
SIGMET 7 forecast valid from 21:00 18.03.2016 till 01:00 19.03.2016: severe turbulence
to the north of 44° N and to the west of 48° E from GL up to FL150, stationary, no change.
At 21:40 on 18.03.2016, the flight FDB981 entered the area of responsibility of the Rostov-
on-Don UAMTS area center (Rostov Control).
At 21:50, the Rostov UAMTS area center ATC officer transmitted the factual weather on
the Rostov-on-Don aerodrome as of 21:30 to the flight crew: wind 250° 9 m/s, gusts 15 m/s,
visibility 5 km, cloud base 390 m, QNH 1000 hPa, as well as the SIGMET No 7 information.
At 22:24, the FDB981 flight entered the area of responsibility of the ATM Rostov division.
At the Rostov-on-Don aerodrome the measurement and the processing of the weather
parameters on the artificial RWY is carried out by the KRAMS-4 complex radio technical
aerodrome weather station (the main and standby assemblies), its main and standby sensors are
located at the areas of the inner markers, in the landing area and at the middle of the artificial
RWY.
The values of the surface wind parameters, the visibility, the cloud base or the vertical
visibility (at the inner marker), the temperature, the air humidity, the atmospheric pressure are
measured automatically-remotely.
The RAMC provides with the meteorological information that is necessary to perform their
functions the aircraft flight crew members, the Rostov ATM air traffic authorities (the air
operations supervisor, the radar control unit, the approach control unit, the tower control unit, the
ground control unit), the operations and dispatcher service, the head of the comprehensive duty,
the aerodrome service, the search and rescue authorities of the EMERCOM Southern regional
center air rescue service – under the agreement of 14.01.2016.

The observations are carried out at the main observation station that is located at the area
of the artificial RWY threshold with the magnetic heading 038. All the sensors of the weather
instruments are remotely connected to KRAMS-4. The weather equipment meets the ICAO CATII
requirements for takeoff and landing operations.
On 18.03.2016 the weather conditions at the Rostov-on-Don area were determined by the
rear section of the high-level trough, the axis of which passed along Samara – Makhachkala. At
the ground surface they noted the interaction of the high pressure area over Turkey and the eastern
sector of the Mediterranean sea and the extensive cyclone with the center over Perm that had been
shifted to the south with the speed of 10 km/h. The polar and arctic front systems were allied to
this cyclone.
The polar front with the waves was passing over from Aktobe to Aktau, Lagan, Kamensk-
Shakhtinsky, Donetsk, Kremenchuk and was shifted along the steering current (270°-290°) to the
south – south-east with the speed of about 50 km/h. The passing over Rostov-on-Don was expected
at about 00 hrs on 19.03.2016.
The arctic cold front was passing over from Uralsk to Penza, Tambov, Minsk, shifted along
the steering current with the speed of 60 km/h. The passing over Rostov-on-Don was expected at
about 06 hrs on 19.03.2016.
This very weather pattern contributed to the increase of the atmospheric pressure gradient,
the increase of the surface wind, the development of the atmospheric turbulence and windshear.
Note: The windshear is a change in wind speed and/or direction in space, including
updrafts and downdrafts (Doc 9817 Manual on Low-Level Windshear).
Turbulence is an air circulation, at which the air particles move in an unsteady,
chaotic way along the complex trajectories. The movement is characterized by
the presence of vortices of different dimensions that are shifted with different
speeds in a general (average) air current. The vortex nature of the air circulation
specifies the presence of the fluctuating ripples of the wind speed within the
turbulent area, including the ripples of the wind vertical component, which
considerably affects the aircraft flight.
The windshear and turbulence are included in the local weather reports
according to the data from air reports.
At 23:30 on 18.03.2016, the radiosonde was launched by the Rostov-on-Don aerological
station. Based on the radiosonde data the aerological graph was constructed, of which it is evident
that the labile stratification had been observed within a ground – 1300 m layer that contributed to
the cumulonimbus cloud formation (with the cloud ceiling of up to 4.5 km) at the aerodrome area
that caused the shower rain precipitation. The freezing level was observed at 700 m. The high

humidity ratio contributed to the icing formation above the level of 700 m in the stratocumulus
and cumulonimbus cloud that was confirmed by the air reports transmitted by the crews of the
aircraft having approached and departed the Rostov-on-Don aerodrome. As per the air reports the
moderate icing within the layer of 900-1500 m in the stratiformis cloud was reported, above
1500 m the cumulonimbus cloud was noted present that is confirmed by the NOAA-19 and
Meteosat-10 weather earth satellites cloud shots of 18-19.03.2016.
Note: The cumulonimbus cloud forecast implies the moderate/severe turbulence and
moderate/severe icing.
The wind pattern parameters based on the results of the atmospheric radiosonde by the
Rostov-on-Don aerological station (as of 00:00 on 19.03.2016).
Altitude, m Wind direction Wind speed, m/s
100 233 8
200 235 14
300 235 21
400 239 25
500 245 28
600 248 29
900 254 29
1000 257 29
1500 261 29
2000 265 30
3000 258 33
The North Caucasian Hydrometeorological Service administration issued the storm

warning into the Rostov region: «At the overnight into 19.03.2016 and before noon on 19.03.2016
into the territory of the Rostov region the strong and high south-western, western wind with the
gusts of 25-30 m/s is anticipated; at the city of Rostov-on-Don with the gusts of 23-28 m/s».
Note: The storm warnings are communicated to all the concerned weather authorities,
including the Aviamettelecom of Roshydromet, FSBI, North Caucasian branch
Rostov-on-Don air meteorological center. This information may be used by the
forecaster in drawing up forecasts. This information is not communicated to the
aircraft flight crews directly.
According to the data of the North Caucasian Hydrometeorological Service administration
weather stations into the Rostov region the southwestern wind with the gusts of up to 20 m/s had
been observed within a period of 12:00 on 18.03.2016 – 02:00 on 19.03.2016.

The RAMC duty forecaster at 19:59 issued the regular forecast into the Rostov-on-Don
aerodrome:
TAF URRR 181959Z 1821/1921 25007G13MPS 3000 -SHRA BR SCTC010 SCT020CB
TEMPO 1821/1906 25013G20MPS 1000 -SHRA BR SCT005 BKN020CB FM1906
29010G17MPS 3000 –SHRA BR SCT005 BKN020CB TEMPO 1906/1918 30017G25MPS 1000
SHSNRA BR SCT003 BKN020CB.
The forecast drawn up at 19:59 on 18.03.2016, valid within a period of 21:00 on 18.03.2016
– 21:00 on 19.03.2016: surface wind 250° 7 m/s, gusts 13 m/s, visibility 3000 m, cloud: scattered
(3-4 octants), cloud base 300 m, broken (5-7 octants), cumulonimbus, cloud base 600 m;
temporarily within a period of 21:00 on 18.03.2016 – 06:00 on 19.03.2016: surface wind 250°
13 m/s, gusts 20 m/s, visibility 1000 m, light shower rain, mist, cloud: scattered (3-4 octants),
cloud base 150 m, broken (5-07 octants) cumulonimbus, cloud base 600 m; from 06:00 on
19.03.2016: surface wind 290° 10 m/s, gusts 17 m/s, visibility 3000 m, light shower rain, mist,
cloud base: scattered (3-4 octants), cloud base 150 m, broken (5-7 octants) cumulonimbus, cloud
base 600 m; temporarily within a period of 06:00 – 18:00 on 19.03.2016: surface wind 300°
17 m/s, gusts 25 m/s, visibility 1000 m, moderate snow shower with rain, cloud: scattered (3-4
octants), cloud base 90 m, broken (5-7 octants) cumulonimbus, cloud base 600 m.
At 20:00 there was issued a regular warning No 3 into the Rostov-on-Don aerodrome on
the wind increase. The aerodrome warning No 3 was drawn up at 20:00 on 18.03.2016, valid
within a period of 21:00 on 18.03.2016 – 06:00 on 19.03.2016: the wind speed of 13 m/s, max
20 m/s is forecasted.
This very warning as per Supplement E of the Instruction on the meteorological service to
air operations at the Rostov-on-Don international airport had been transmitted to the approach
control unit officer (at 20:00), ground control unit officer (at 20:01) with the delivery confirmation,
to the radiotechnical communications and navigational equipment service, the operations and
dispatcher service, the operations coordination and planning department, to the duty forecasters of
the international air operations and air force.
As per the data of the air report, transmitted by the UTA497 flight in the progress of
approach to the Rostov-on-Don aerodrome, the moderate windshear had been observed on final
that was introduced in the local weather report – displayed at the MeteoDisplay automated
information system – and incorporated into the Tango ATIS information.
Note: According to the Instruction on the meteorological service to air operations at
the Rostov-on-Don international airport items 3.4, 3.5 the information on the
presence of the specific conditions and occurrences is transmitted by the aircraft
crew to the meteorological office via the ATC officer.

At 21:55 on 18.03.2016, the forecasted moderate windshear warning was issued into the
Rostov-on-Don aerodrome:
URRR WS WRNG 1 182155 VALID 182200/1190600 MOD WS FCST.
The Rostov-on-Don aerodrome windshear warning No 1 drawn up at 21:55 on 18.03.2016,
valid from 22:00 on 18.03.2016 until 06:00 on 19.03.2016: the moderate windshear is forecasted.
This very warning had been transmitted to the approach control unit officer (at 21:55), the
ground control officer (at 21:57) with the delivery confirmation, the international air operations
duty forecaster, and the air operations supervisor.
Note: The windshear aerodrome warning is drawn up according to item 5.4.4 of the
Instruction on the meteorological service to air operations at the Rostov-on-Don
international airport. According to item 7.1.4 of the Instruction the warning
shall be transmitted by the ATC officer to the aircraft throughout the period of
its validity or until the receipt of the cancellation messages.
As per item 5.3.16 of Doc 9817 Manual on Low-Level Wind Shear:
ATS units should continue to transmit information on wind shear conditions until
it is confirmed, either by subsequent aircraft reports or by advice from the
associated MET office, that conditions are no longer significant for operations
at the aerodrome…The ATS unit should continue to relay air-reports of wind
shear to other aircraft concerned until such time as the reports have been
incorporated into a wind shear warning by the associated MET office.
Thereafter, the wind shear warning will be transmitted to all aircraft concerned
until cancelled by the MET office.
As per item 4.3.4 of Doc 9817 Manual on Low-Level Wind Shear:
The wind shear warnings, issued at an airport, serve to alert the pilot to the
possibility of wind shear and permit appropriate action to be taken.
At 22:24 at the radio contact with the Rostov-on-Don ATC officer the Boeing 737-8KN
A6-FDN aircraft flight crew reported that they had monitored the Uniform ATIS information.
The ATIS information Uniform of 22:00: «…wind 240⁰ 10 m/s gusts 15 m/s, visibility
4200 m, light shower rain, cloud: scattered at 390 m, broken cumulonimbus at 990 m, temperature
6⁰, dew point 4⁰, QFE 742 mm/990 hPa, QNH 1000 hPa, moderate windshear, moderate
turbulence from GL to 1000 m, moderate icing in cloud within 900-1500 m; temporarily: surface
wind 250⁰ 13 m/s gusts 18 m/s, visibility 1000 m, shower rain, mist, cloud: scattered at 90 m,
broken cumulonimbus at 600 m…».
At 22:25 the ATC officer recommended that the crew monitored the ATIS information
Whisky: «…wind 230⁰ 10 m/s, gusts 17 m/s, visibility 2900 m, light shower rain, cloud: scattered

at 480 m, broken cumulonimbus at 990 m, temperature 6⁰, dew point 3⁰, QFE 742 mm/990 hPa,
QNH 1000 hPa, moderate turbulence from GL up to 1000 m, moderate icing in cloud from 900 m
up 1500 m; tempo: surface wind 250⁰ 13 m/s gusts 18 m/s, visibility 1000 m, shower rain, mist,
cloud: scattered at 90 m, broken cumulonimbus at 600 m…».
As per the air report transmitted at 22:30 by the flight crew of the Ural Airlines flight SVR
2758 having been in progress of the approach the light-to-moderate windshear and the moderate
turbulence were observed on final. The air report data were incorporated into the local weather
report and the ATIS information Zulu.
The ATIS information Zulu: «… wind 230⁰ 10 m/s gusts 17 m/s, visibility 4100 m, light
shower rain, cloud: scattered at 480 m, broken cumulonimbus at 990 m, temperature 6⁰, dew point
3⁰, QFE 742 mm/990 hPa, QNH 1000 hPa, moderate windshear, moderate turbulence from GL
up to 1000 m, moderate icing in cloud from 900 m up to 1500 m; temporarily: surface wind 250⁰
15 m/s gusts 20 m/s, visibility 1000 m, shower rain, mist, cloud: scattered at 90 m, broken
cumulonimbus at 600 m…».
At 22:42, the FDB981 flight crew made the decision to initiate the go-around having
advised to the ATC officer the windshear as the reason to perform the maneuver.
In the progress of the climb at 22:47, the flight crew relayed the air report to the ATC and
requested FL80 due to the moderate icing.
At 22:49 the ATC officer transmitted the moderate icing be present at FL50 to the RAMC
duty forecaster.
At 22:55 the flight crew of the AFL1166 flight having approached the Rostov-on-Don
aerodrome reported the moderate windshear on final and go-around.
At 23:07 the ATC officer transmitted to the meteo office that in the progress of the missed
approach the AFL1166 crew had reported the moderate windshear at the altitude of 400-300 m.
At 23:27 the air operations supervisor reconfirmed, having transmitted the information via
loudspeaker communication system, that the AFL1166 aircraft in the progress of the third go-
around had reported the moderate-to-strong windshear.
Note: Doc 9817 Manual on Low-Level Wind Shear item 5.2.7:
…pilots, when reporting wind shear, may use the qualifying terms ‘moderate’,
‘strong’ or ‘severe’, based to a large extent on their subjective assessment of the
intensity of the wind shear encountered, and such qualifications have to be
included unchanged in the report.
The information on the windshear presence at the aerodrome had been transmitting to ATIS
up to 00:02 on 19.03.2016.

At 23:47 the FDB981 flight crew asked the tower control unit officer whether the
windshear warning had been issued. The controller replied in the negative: «Negative», but in no
more than 15 s the tower control unit officer relayed the presence of the moderate windshear to
the crew.
At 00:04 the ATC officer recommended that the crew monitored the ATIS information
Bravo.
Information Bravo of 00:02: «…wind 230⁰ 11 m/s gusts 14 m/s, visibility 7 km, light shower
rain, cloud: scattered at 570 m, broken cumulonimbus at 1200 m, temperature 6⁰, dew point 3⁰,
QFE 741 mm/988 hPa, QNH 998 hPa, moderate turbulence from GL up to 1000 m, moderate
icing in cloud from 900 up to 1500 m; temporarily: surface wind 250⁰ 15 m/s gusts 20 m/s,
visibility 1000 m, shower rain, mist, cloud scattered at 90 m, broken cumulonimbus at 600 m…».
At 00:20 the tower control unit officer relayed the weather information to the aircraft:
«Weather as of 00:20: visibility 5 km, cloud base 630 m, wind 230⁰ 13 m/s gusts 18 m/s, light
shower rain, mist, on final the severe turbulence and moderate windshear».
At 00:22 the approach control unit officer relayed to the aircraft: «…as of 22…230-14 gusts
18, visibility 6 km, scattered 480 m, correction 630 m, meteorological office is not reported about
windshear on the RWY».
At that moment, the aerodrome windshear warning remained in effect and should have
been displayed in red color in a form of WS WRNG conventional English language abbreviation
at the controllers’ MeteoDisplay automated information system display window.
At 00:23 the controller relayed to the crew: wind 230-15 gusts 19 m/s.
At 00:25 the ground control unit officer asked the forecaster via loudspeaker
communication system: «…you are not transmitting the moderate windshear in the ATIS now, are
you? ». The response of the meteo office via loudspeaker communication system: «Actually we
are not, but the warning is active». The ground control unit officer via loudspeaker communication
system: «Yes, the warning is active, at actually at 2 hrs you are not transmitting it temporarily for
landing, are you? » The meteo office response via loudspeaker communication system: «The
windshear is not transmitted for landing».
At 00:32, in the progress of departure from the Rostov-on-Don aerodrome the SVR2757
flight crew reported:
«In climb the wind is gusting, now… now 260 degrees, 53 knots, cloud base 2500 ft, in
climb light icing in cloud».
At 00:34 the ATC controller relayed to the FDB981 crew that the departing aircraft
reported the wind at 600 meters 260 degrees 53 knots and light icing.
At 00:40 the FDB981 flight crew reported go-around.

At that moment the ATIS information Echo of 00:30 had been transmitted on air:
«… surface wind 230 degrees 12 m/s gusts 19 m/s, visibility 3800 m, light shower rain, cloud:
scattered at 540 m, broken cumulonimbus at 1080 m, moderate turbulence from GL up to 1000 m,
moderate icing in cloud from 900 up to 1500 m; temporarily: surface wind 250 degrees 17 m/s
gusts 25 m/s, visibility 1000 m, mist cloud; scattered at 90 m, broken cumulonimbus at 600 m».
The air accident occurred at 00:42.
The actual weather parameters, withdrawn from the KRAMS-4 sensors archives as of the
moment of the accident at 00:42: surface wind magnetic 230°-13 m/s gust 18 m/s, visibility
7000/7000/3700 m (touchdown/midpoint/rollout), light shower rain, cloud: scattered (4 octants),
cloud base 420 m, broken (5-7 octants) cumulonimbus, cloud base 1080 m, overcast (8 octants),
cloud base 3000 m, temperature + 6.3°С, dew point + 3.6°С, relative humidity 84%, QNH
998.0 hPa, QFE 742 mm of mercury/988 hPa; moderate icing in cloud within 900-1500 m,
moderate turbulence from GL up to 1000 m.
At 00:55, the tower control unit officer via loudspeaker communication system requested
the technician-forecaster that he measured the weather parameters as per the alarm signal (in
13 min after the alarm signal had been transmitted via GORN alerting system).
The actual weather as documented at the Rostov-on-Don aerodrome after the alarm signal
declaration was transmitted as of 00:55: Surface wind magnetic 230°-16 m/s gust 22 m/s, visibility
3500/3900/2400 m (active heading/middle/inactive), light shower rain, cloud: broken (5-7 octants)
cloud base 450 m, cumulonimbus cloud, cloud base 1100 m, middle overcast (8 octants),
temperature + 6.2°С, dew point +3.8°С, relative humidity 84%, QNH 997.8 hPa, QFE
740.8 mm of mercury/987.8 hPa, moderate icing in cloud within 900-1500 m, moderate
turbulence from GL up to 1000 m.
1.8. Aids to navigation

The Rostov-on-Don airport is equipped with the navigation aids in compliance with the
List of Equipment. The detailed information is given in the AIP of Russian Federation and CIS.
There were no failures or anomalies occurred at the time of the accident flight as far as the navaids
operation is concerned, all the systems were serviceable, there was no switching to the standby
power supply.
The condition of the radio navigation to air operations and the aeronautical
telecommunication at the moment of the accident complied with the operation and maintenance
documentation requirements, FAR, Airworthiness Requirements to Aerodrome Operation. The
aids were functioning as designed with the magnetic heading for landing of 218°.
The State ATM Corporation South Air Navigation Rostov division branch Radiotechnical
Communications and Navigational Equipment Department carries out the operation and

maintenance of the radio navaids on the Rostov-on-Don aerodrome. The Radiotechnical

Communications and Navigational Equipment Department is certified within the Rostov division.
The following radio aids to navigation ensured the operation of ATC approach, radar and
tower control units:
The aids to radio navigation and landing
Air navigation data name The air navigation data element value
Type and category ILS-22 (SP-80M) I category
Radio aid designation LOC 22
Antenna magnetic variation 7.8º East
Callsign ИРВ/IRW
Frequency 110.3 MHz
Station magnetic variation 7.8º East
Reference position of the antenna installation
location (latitude and longitude in degrees, 47º14′44.72″ N, 039º47′57.46″ E
minutes, seconds and centiseconds)
Hours of operation (UTC) period of flight operations
Radio aid designation G/S 22
Callsign ИРВ/IRW
Frequency 335.0 MHz
Hours of operation period of flight operations
Radio aid designation LOM 22
Callsign РВ/RV
Frequency 320 KHz

Radio aid designation OMB 22

Callsign 2 dashes/sec
Frequency 75 MHz
Radio aid designation LIM 22
Callsign В/V
Frequency 659 KHz
Radio aid designation IMB 22
Callsign 6 dots/sec
Frequency 75 MHz
Radio aid designation the Platan ADF
Frequency 118.0-136.975 MHz (reception)
Hours of operation (UTC) H24
Radio aid designation the Lira aerodrome surveillance radar
Antenna magnetic variation 7º8 East
1030, 2753.5-2766.5 MHz (transmission);
Frequency 740, 1090, 2753.5, 2763.5, 2756.5,
2766.5 MHz (reception)
Station magnetic variation 7º8 East
Hours of operation (UTC) H24

 the SP-80M localizer, MSN 8525, year of manufacture 1988, introduced to service on
07.09.1988, modified in compliance with the manufacturer SBs in 2003, 2006 and 2010, TSN as
of February 2016 (inclusively) 95900 hrs, the last maintenance check (TO-С) carried out on
16.03.2016. Installation landing magnetic heading-218;
 the SP-80M glideslope, MSN 8525, year of manufacture 1988, introduced to service on
07.09.1988, modified in compliance with the manufacturer SBs in 2003, 2006 and 2010, TSN as
of February 2016 (inclusively) 96655 hrs, the last maintenance check (TO-С) carried out on
 the PAR-10C locator at the outer marker, MSN 1000933, year of manufacture 1985,
introduced to service on 30.12.1985, TSN as of February 2016 (inclusively) 99138 hrs, the
maintenance-3 carried out on 16.02. 2016. Installation landing magnetic heading-218; the Е-615.1
outer marker, MSN 19009, year of manufacture 1989, introduced to service on 26.09.1990, TSN
as of February 2016 (inclusively) 89298 hrs, the last maintenance check (ТО-3) carried out on
 the PAR-10C locator at the inner marker, MSN 1000895, year of manufacture 1985,
introduced to service on 30.12.1985, TSN as of February 2016 (inclusively) 97512 hrs, the ТО-3
maintenance carried out on 29.02.2016. Installation landing magnetic heading-218; the Е-615.1
inner marker, MSN 19006, year of manufacture 1989, introduced to service on 26.09.1990, TSN
as of February 2016 (inclusively) 87512 hrs, the last maintenance check (ТО-3) carried out on
 the Platan (DF-2000) ADF, MSN ПЛТ-015.12, year of manufacture 2012, introduced
to service on 26.12.2012, TSN as of February 2016 (inclusively) 28266 hrs. The last maintenance
check (ТО-3) carried out on 10.03.2016. Installed at the Rostov-on-Don aerodrome;
 the Lira-A10 aerodrome surveillance radar, MSN 210001, year of manufacture 2010,
introduced to service on 21.12.2011, TSN as of February 2016 (inclusively) 36228 hrs. The last
maintenance check (ТО-С) carried out on 12.09.2015 (as far as the Avrora secondary radar is
concerned, the last maintenance check (ТО-4) was carried out on 28.12.2015). Installed at the
Rostov-on-Don aerodrome;
 the 1090 ES НС-1А ADS-B ground station, MSN 063/14, year of manufacture 2015,
introduced to service on 29.09.2015, TSN as of February 2016 (inclusively) 3696 hrs. The last
maintenance check (ТО-5) carried out on 18.11.2015. Installed at the Rostov-on-Don aerodrome.
There were no comments received on the operation of the radio navigation to air operations
and the aeronautical telecommunication equipment on 18.03.2016 and 19.03.2016 from the ATC
officers, other airport services and the aircraft flight crews.

There was no off-performance or the off-design operation detected of the radio navaids to
air operations or the aeronautical telecommunication equipment.
The ER 22 STAR (Fig. 30) and the RWY22 ILS approach chart (Fig. 32) are given here
below. The data are taken from the AIP of Russia.
The variants of charts, available to the crew are presented on Fig. 31, Fig. 33.
Fig. 30. The RWY22 STAR at the Rostov-on-Don airport (from the AIP of Russia)

Fig. 31. The variants of the RWY22 STAR available to the crew

Fig. 32. The RWY22 ILS approach chart at the Rostov-on-Don airport (from the AIP of Russia)

Fig. 33. The variant of the RWY22 ILS approach chart available to the crew

The Transcon lighting was installed with the landing magnetic heading 38º and landing
magnetic heading 218º in 2006. It was subject to renewal in 2015. The equipment configuration
complies with FAR-262.
The Rostov-on-Don aerodrome as per the landing magnetic heading 218° is equipped with
a high-intensity lights lighting equipment.
No in Air navigation data name The air navigation data element
order value
RWY designation 22
1 Approach lighting system type OVI-1 high intensity approach
lighting system with centerline bar
lights
2 Approach lighting system length 900 m
3 Approach lighting system luminous intensity 24000 candelas
4 RWY threshold lights (color) green
5 Touchdown zone wing bar lights none
6 Visual glide path indication system PAPI
7 Approach slope 2°40'
8 Location of visual glide path indication system to the left of the RWY at the distance
of 525 m off the runway threshold
9 Length of the RWY touchdown zone lights none
10 Length of the RWY centerline lights 2500 m
11 RWY centerline lights luminous intensity 2500 candelas
12 RWY centerline lights installation intervals 15 m
13 RWY centerline lights color At the area up to 1600 m – white;
beyond 600 m – red/white; last
300 m – red
14 RWY landing (edge) lights length 2500 m
15 RWY landing (edge) lights installation 60 m
intervals
16 RWY landing (edge) lights luminous intensity 13000 candelas
17 RWY landing (edge) lights color at the area up to 1900 m – white; the
last 600 m – yellow
18 RWY end lights color red
19 Stopway zone wing bar lights none
20 Length and color of stopway lights none

There was neither switching nor overvoltage registered at the lighting equipment. The
equipment in question had not switched to the standby power supply.
1.9. Communications
The detailed information on the communications are not stated throughout the report, since
the air accident is not anyhow related to their serviceability. Prior to the accident, the radio contact
with the flight crew involved had been stable and intelligible. The crew-ATC radio
communication, as well as the radio exchange at the other communication channels, subject to
recording, had been stored by the ground recorders, decoded and used by the investigation team.
The last radio contact between the aircraft flight crew and the tower controller was
established at 00:40:57. In reply to the crew report on the initiation of the go-around the tower
controller instructed the crew that they contacted the radar controller with the indication of the
frequency. The crew confirmed the receipt of information. Further on the crew did not establish
any radio contact with the radar controller, did not respond to the ATC officers’ calls.
At 00:42:21 the ground controller, having seen the bright flash and the subsequent fire at
the area of TWY Delta, declared an alarm.
1.10. Aerodrome information

The Rostov-on-Don aerodrome (ICAO code - URRR) is a civil aerodrome of a B class (an
ICAO 4D class).
The ground area is state-owned.
The RWY, TWYs, stands and apron are owned by Rostovaeroinvest, PJSC.
Hours of operation – H24.
The Rostov-on-Don aerodrome is registered in the State register of the civil aerodromes of
Russian Federation. The RF Ministry of Transport Air Transport Department issued the aerodrome
Certificate of registration and operability No 57 of 16.06.1993.
On 21.05.2012 FATA renewed the Certificate with an expiry on May 31, 2017 based upon
the Aerodrome Compliance Audit Report to the requirements on the aerodromes operation of
15.12.2011.
The Certificate No 010 А-М of 22.10.2015, issued by the IAC aerodromes and equipment
certification commission, reads the following:
«…the Rostov-on-Don aerodrome complies with the certification requirements of the
AP-139.
RWY 04/22:

with magnetic landing heading 38° is equipped with OVI-II high-intensity lights, ILS-II,
designed for ICAO CAT I precision approach up to DH of 60 m at the RVR not less than 550 m,
ICAO CAT II up to DH of 30 at the RVR not less than 350 m;
with magnetic landing heading 218° is equipped with OVI-II high-intensity lights, ILS-II,
designed for ICAO CAT I precision approach up to DH of 60 m at the RVR not less than 550 m.
The aerodrome is qualified for international air operations…».
The aerodrome is qualified for different types of aircraft to be operated, Boeing 737-800,
757, 767-200, 767-300 and their modifications among them.
The Rostov-on-Don aerodrome (Fig. 34, Fig. 34a) is located at the distance of 10.5
kilometers to the northeast of the railroad station of the same name.
The aerodrome reference point geographic position: 47°15′30″ N, 039°49′05″ E.
The aerodrome reference position elevation (the true altitude) is +85.6 m, the magnetic
variation is +8º. The time zone number – 3 (UTC+3 hrs).

Fig. 34. The Rostov-on-Don aerodrome chart as published in the Aerodrome Air Navigation Passport

Fig. 34а. The Rostov-on-Don aerodrome chart available to the crew
The airfield is a polygon with the dimensions of 3300*260*1900*1000*1800 m. The

airfield surface is flat, the ground is soft and the soil is loam. The airfield surface is unfit for the
takeoff and landing operations.
The aerodrome integrates the artificial runway.
The artificial runway (RWY 04/22) is of В class, of 2501 m length with the active width
of 45 m, permanent throughout the length. The surface is fibercrete, strengthened with asphalt
concrete.
The RWY 04/22 thresholds are coincident with the RWY approach end.
The RWY 04/22 thresholds takeoff and landing magnetic track angles, numbers, reference
positions and true altitudes are given in the table below.

Threshold Takeoff and landing Threshold reference position Threshold true altitude
No magnetic track
angle
04 038° 47º15′01.31″ N, 039º48′22.28″ E 72.58 m
22 218° 47º15′57.96″ N, 039º49′47.27″ E 85.64 m
The runway strip runs in a lateral direction on either side of the runway axis, along the
entire length of 150 m to the either side. The runway strip cleared and graded area runs 72.5 m off
the runway axis.
The runway strip runs beyond the runway stop end:
 with landing magnetic heading 38°: 150 m;
 with landing magnetic heading 218°: 150 m.
The runway strip hard section dimensions amount to:
 at magnetic heading 38°: the inner width – 60 m, length – 50 m;
 at magnetic heading 218°: the inner width – 49 m, length – 50 m.
There are no objects, located within the runway strip cleared and graded area, except for
those, disposed as per their functional purpose.
Within a distance of 120 m to either side off the artificial runway axis there are no
obstacles.
The aerodrome operation engineer – the head of aerodrome shift at 21:15 carried out the
inspection of the airfield with the aim to evaluate its condition. The friction coefficient
measurement was conducted with the use of ATT-2 No 2 MSN 1505 braking cart with the
subsequent entry into the Rostov-on-Don aerodrome airfield condition log. The norm friction
coefficient amounted to 0.57/0.57/0.57. At 0:11 another friction coefficient measurement was
done. The norm friction coefficient amounted to 0.46/0.46/0.5.
The inspection of the artificial surfaces was carried out in compliance with the provisions
of the Manual on the civil aerodromes operation in Russian Federation (REGA RF-94).
After the accident at 00:52 the additional entry was introduced that the artificial runway
with the magnetic heading 218° is unserviceable due to the foreign objects presence (the aircraft
wreckage) and the artificial runway surface disintegration.
At 00:52 the aerodrome had been closed.

1.11. Flight recorders

Flight data recorder
The Honeywell SSFDR p/n 980-4700-042 with the secure memory unit s/n 35907 had been
installed on the accident aircraft. On the accident scene, the secure memory unit was retrieved
being out of the recorder case (Fig. 35).
Fig. 35. The FDR secure memory unit
The recovery and readout of the stored information was performed at the facilities of the
MAK-IAC AAISTSC by the team, consisted of the following experts:
 of the Interstate Aviation Committee;
 Bureau d'Enquêtes et d'Analyses pour la Sécurité de l'Aviation Civile (BEA), France;
 General Civil Aviation Authority of United Arab Emirates.
The analysis of the recovered information revealed that the FDR had been serviceable and
recorded all the analog parameters and the discrete signals in compliance with the Digital Flight
Data Acquisition Unit 737-600/-700/-700c/-800/-900, Data Frame Interface Control And
Requirements Document D226a101-2 within 17.03.2016 – 19.03.2016 (the total duration of the
record amounts to 26 hrs 34 min), including the Boeing 737-800 A6-FDN flight of 19.03.2016,
ended up with the air accident at the Rostov-on-Don aerodrome. The quality of the record is good.
Cockpit voice recorder
The L3 FA2100 2100-1020-00 cockpit voice recorder (CVR) had been installed on the
accident aircraft. The recorder case had been removed from the accident site. At the inspection, it
was noted that the case sustained considerable mechanical damage (Fig. 36).

Fig. 36. The external view of the CVR
At the MAK lab the recovery works had been performed on the removal of the memory
unit from the damaged case and its re-installation in the technological unit. With the use of the
standard L3 Communications software-hardware complex the readout and conversion of all the
data volume of the recorder sound information was performed. The total volume of the data
amounted to 02 hrs 4 min 14.5 s. The record is consistent with the Boeing 737-800 A6-FDN
aircraft flight of 19.03.2016, ended up with the air accident at the Rostov-on-Don aerodrome. The
quality of the record is good.
The data, withdrawn from the flight data and cockpit voice recorders, had been used in the
investigation team activities.
1.12. Wreckage and impact information

The accident scene is located at the Rostov-on-Don airport airfield. The first point of
impact with ground (the RWY concrete surface) occurred at the ≈ 120 m off the RWY22 threshold
along the RWY left edge looking forward, reference position: 47º15′54.7″ N, 039º49′43.8″ E. The
main wreckage scatter area is located at the distance of 150-400 m off the RWY22 threshold to
the left (the major portion of the wreckage) and to the right of the RWY centerline (Fig. 37).

Fig. 37. First point of impact with ground (the left-to-right direction of flight)
The main wreckage scatter zone is located at the distance of 150-400 m off the RWY
threshold on the left side (the larger portion of wreckage) and on the right side off the RWY
centerline (Fig. 38).

Fig. 38. The general view of the accident scene (the arrow indicates the flight heading)

The impact resulted in the crater formation of about 10-12 m length, 2.5 m width and up to
1.5 m depth. After the impact, the flash of the fuel-air mixture occurred with the subsequent fire.
By the pattern of the wreckage scatter, it can be concluded that the aircraft impact with
ground had occurred in a steep nose-down attitude while rolling to the left.
The largest aircraft elements, retrieved at the accident site, rested in the following sequence
as per the flight heading8:
 the fragments of the fuselage nose section (with the cockpit), the centerwing, left wing
console, left half of the stabilizer, left half of the elevator, nose landing gear leg and the fragments
of the left engine (the compressor case with the elements of the compressor control system) rested
inside the crater;
 the right section of the stabilizer with the attached to it right half of the elevator at the
azimuth ≈ 235° at the distance of 175 m;
 the fragment of the fuselage RH half section (from FR927 up to FR986.5) (the area of
the aft service door) and the aft service door at the distance of 148 m on the RWY;
 the fragment of the upper portion of the fuselage (from FR694 up to FR727А) with the
ADF antennas at the distance of 160 m on the RWY;
 the damaged forward cabin door at the azimuth of ≈ 230° at the distance of 169 m;
 the disc of the fan stage of the right engine with the fragments of the blades at the
azimuth ≈ 222° at the distance of 325 m;
 the disc of the fan stage of the left engine at the azimuth of ≈ 217° at the distance of
264 m;
 the fragment of the LH fuselage section (from FR578 up to FR610) with the fragment
of the overwing emergency exit at the azimuth ≈ 223° at the distance of 203 m;
 left main landing gear at the azimuth ≈ 224° at the distance of 238 m;
 right main landing gear at the azimuth ≈ 216° at the distance of 259 m;
 the APU with the exhaust duct at the azimuth ≈ 237° at the distance of 147 m;
 the fragments of the right engine (compressor case with the elements of the compressor
control system and the engines rotor front support) at the azimuth ≈ 217° at the distance of 255 m;
 the fragments of the elevator control system (the input rods, the control PCUs and the
autopilot electrohydraulic actuator) at the azimuth ≈ 235° at the distance of 73 m;
 the ailerons autopilot electrohydraulic actuator at the azimuth ≈ 225° at the distance of
300 m;
8
All the magnetic azimuths and distances are referred off the RWY first point of impact.
 the assemblies of the stabilizer trim system (the drum with the fragments of the cable
linkage, the gearbox with the jackscrew separated in two parts, the trim motor apart from each
other with the azimuth ≈ 226° at the distance of 73 m.
The origin of fire was located at the area of the crater. However, the fuselage fragments
from FR500С up to 1156Н, the right half of the stabilizer (having rested at the distance of 175 m
off the crater) and the fin do not demonstrate the signs of fire or the thermal effect both on the
outer and inner skin. Based on that it may be concluded that initially at the aircraft impact with
ground, there occurred the fuselage destruction with the separation of the stabilizer and the fin and
then there developed the fuel flash and fire.
The investigation team has drawn out a wreckage map (the pattern of the wreckage scatter
afield) (Fig. 39).
Fig. 39. The pattern of the wreckage scatter on the accident site
The main wreckage scatter zone was located inside the А-Б-В-Г-Д polygon.
The perimeter distance between the points:
 А-Б amounts to 445 m;
 Б-В amounts to 148 m;
 В-Г amounts to 216 m;
 Г-Д amounts to 530 m;
 Д-А amounts to 385 m.
The reference position of the scatter zone points:
 point А – 47º15.844′ N, 039º39.49.411′ E.;
 point Б – 47º16.009′ N, 039º49.667′ E;

 point В – 47º15.975′ N, 039º39.49.7738′ E.;

 point Г – 47º15.896′ N, 039º39.49.890′ E;
 point Д – 47º15.687′ N, 039º49.49.611′ E.
Only two smaller fragments of the fuselage skin were discovered out of the main wreckage
scatter zone (the points 116 and 117 on Fig. 39), situated at the azimuth ≈ 57º at the distance of
≈ 474 m and, most probably taken away by the wind due to their sailing capacity/windage. The
actual weather observed at the Rostov-on-Don aerodrome after the alarm signal declaration as of
00:55: «…surface wind magnetic 230⁰ 16 m/s, gusts 22 m/s…».
1.13. Medical and pathological information

At the facilities of the Russian Federation Ministry of Health Center of Forensic Medical
Expertise, FSBI there was performed the DNA identification profiling of the human remains.
4389 fragments of the human remains were submitted for the examination. 3906 fragments
out of them (89%) were determined appropriate for profiling, whereas 11% were stated unsuitable
to be examined. Thus, 3906 fragments subject to analysis were identified (63 individual bodies).
At the Rostov Region Ministry of Health Forensic Medical Expertise Office, FSI, there
was performed the forensic medical expertise of the PIC. At the examination of the submitted
materials it was determined that in the accident sequence the PIC’s organism had been affected by
the huge loads far beyond the structural properties of the human body. He had been subject to the
rough multisystem injury with the total destruction and massive fragmentation of the body.
The specification of the flight crewmembers’ working posture at the moment of the
accident is based on the elaboration and the analysis of the specific (primary and secondary)
injuries that the crewmembers could have sustained having operated inside the cockpit.
In this very case due to the total destruction of the PIC’s body into a great number of smaller
parts, the presence of the multiple irreparable bones, soft tissues, internal organs defects it turns to
be impossible to sort out the primary and secondary injuries of the flight crewmembers.
Consequently, it is hardly possible to identify what the PIC’s working posture was at the moment
of the accident.
The forensic chemical analysis of the soft tissues out of the PIC dead body was performed,
based on which the following was concluded (Conclusion No 2441/276-пк, the expertise was
carried out within 18.05.2016 – 09.06.2016)9: «In the muscular tissue out of the PIC’s dead body
the ethanol, isopropanol, butanol, acetone, lower aliphatic carboxylic acids, aliphatic saturated
hydrocarbons (the hexane and octane) were detected. The ethanol concentration amounts to
9
Hereinafter in the cited documents the author’s wording is retained.
0.35 mg/g in the muscular tissue; the isopropanol content is 0.12 mg/g in the muscular tissue. The
narcotic drugs, psychotropic substances were not detected».
At the forensic chemical analysis of the muscular tissue of all the other deceased
crewmembers, the similar combination of chemicals was detected (incorporating ethanol,
isopropyl, butyl alcohol, as well as acetone, lower aliphatic carboxylic acids and saturated aliphatic
hydrocarbons (hexane, octane) with the similar ethanol and isopropanol quantitative distribution).
As for the evaluation of the presence and quantity of the detected ethanol, the potential
newly constituted ethanol in the muscular tissue should be taken into account. This ethanol
formation is the result of the putrefaction at the joint impact of several factors, justified by the
nature of the injury and the remains exposure to the environment that conduce to the rapid
development of the putrefactive processes…
By reference to the features of the detected substances, as well as taking into consideration
that due to the massive injuries (the open wounds, the bodies fragmentation) the organism of the
deceased people represented the open system, exposed to the entry of the chemicals from the
outside, the presence of the ethanol in the deceased people’s bodies should be interpreted as
justified by the pollution of the bio objects with the mechanical fluids at the point of the aircraft
fuselage destruction or due to the remains location at the aerodrome surface, polluted with such
fluids.
Based on the above stated the detection of the chemicals (ethanol, isopropanol, butanol,
acetone, lower aliphatic carboxylic acids and saturated aliphatic hydrocarbons – hexane, octane)
in the PIC’s (name, surname) muscle is estimated by the experts as being justified by the injuries
nature and the remains exposure to the environment. This very result is not the evidence that the
PIC had been in a state of alcohol or another toxic (including the drug one) intoxication».
The results of the forensic medical and chemical expertise of the F/O, performed within
the Rostov Region Ministry of Health Forensic Medical Expertise Office, FSI, are similar to those
of the PIC.
1.14. Survival aspects

In the progress of the flight, the flight crewmembers were occupying their duty stations at
the flight deck. The passengers were allocated at the passenger cabin.
Fig. 40 represents the presumed passenger seating allocation.

Fig. 40. Passengers’ seating allocation
Due to the air accident all the people aboard were killed.
According to the conclusion of the complex forensic medical expertise, drawn up by the
Russian Federation Ministry of Health Center of Forensic Medical Expertise, FSBI, the death of
55 passengers and 7 crewmembers occurred at the result of the multisystem injury with the total
disintegration and massive fragmentation of the bodies. These injuries were the outcome of the
huge loads application on the organism far beyond the structural properties of the human body.
They might have been generated at the point and at the circumstances of the aircraft impact with
the RWY on 19.03.2016 with the subsequent fuselage disintegration. The injuries in question
inflicted the grievous bodily harm to the crewmembers and passengers, having been classified as
the threat to life and directly correlated to the cause of death.
In studying the issue of the actual total number of passengers aboard the aircraft and
specifically the number of passengers, the forensic medical DNA profiling by the Rostov Ministry
of Health Forensic Medical Expertise Office, FSI was collated with the passenger manifest data,
provided by the Flydubai airline. The results of the expertise in question allowed determining that
the remains belonged to 63 deceased individuals. 62 persons out of them complied with the list of
the crewmembers and passengers having been aboard the accident aircraft. The sixty-third
identified deceased individual was a female fetus of a deceased passenger. As of the moment of
the flight, the passenger had been pregnant.
1.15. SAR and firefighting operations

The air accident occurred at 00:42.
The ground control unit officer, having seen the bright flash and the subsequent fire at the
area of TWY Delta, declared an alarm.
The firefighting operation by the Rostov-on-Don airport fire team was launched at 00:44.
At 00:45, the city fire teams were warned of the accident, at 00:48 the city emergency
medicine center was notified of it.
At 00:50, the fire was isolated.

The Rostov region Federal Firefighting Service units’ workforce and special vehicles were
deployed to the accident site, including the firefighting personnel of the Crisis Management
Center.
At 01:10, the Rostov region and Rostov-on-Don Federal Firefighting Service units’
workforce and the special vehicles arrived to the accident site. The accident scene by the time of
the arrival represented the fragmentary fire at the surface of more than 1000 sq. m.
The data on the involved workforce and special vehicles:
 the rescue teams in all – 72 persons and 21 special vehicles;
 12 EMERCOM teams and 12 firefighting vehicles;
 the Rostov region Russian EMERCOM General Directorate Crisis Management Center
response group – 3 persons, 1 vehicle;
 the Don rescue center (accident rescue vehicle-2, fire-extinguishing tanker-1) – 10
persons, 3 vehicles;
 the fire and rescue brigade response group – 2 persons, 1 vehicle;
 the Rostov region EMERCOM General Directorate psychological support team – 3
persons and 1 vehicle;
 the Rostov region Russian EMERCOM General Directorate Crisis Management Center
fire control and SAR service – 4 persons, 1 vehicle;
 the fire and rescue unit – 36 persons, 10 vehicles;
 the special fire and rescue unit – 44 persons, 6 vehicles;
 the firefighting coordination directorate – 3 persons, 1 vehicle;
 the Rostov region Russian EMERCOM General Directorate emergency response team
– 9 persons;
 the Russian EMERCOM Southern regional center response group – 3 persons, 1 vehicle;
 the Russian EMERCOM Southern regional center emergency response team – 14
persons;
 the Southern regional center emergency psychological assistance center – 11 persons, 2
vehicles;
 10 emergency medicine ambulance cars.
One Ural-4352 and two Kamaz-63501 firefighting vehicles were involved in a direct fire
suppression.
At 14:00 on 20.03.2016, the grouping was enhanced up to 900 persons and 200 vehicles.
Seven SAR activities areas were allocated within a perimeter of 250 m x 400 m with 14 groups of
rescuers out of 20 persons each.

At 06:00 on 21.03.2016 the SAR activities were completed and the airport resumed the
standard air operations.
As far as the emergency-rescue works are concerned there were no shortcomings revealed
that could have influenced the survivability of the crewmembers and passengers.
Emergency locator transmitter
The ELT is totally destroyed. As per the COSPAS-SARSAT coordination center data, the
ELT had not been activated in the accident sequence.
1.16. Tests and research

1.16.1. Wreckage layout
The identification of the fragments of the aircraft structure had been carried out, the surface
layout of the wreckage inside the hangar was performed afterwards (Fig. 41).
Fig. 41. The surface layout of the wreckage in the hangar
At the analysis of the wreckage map (Section 1.12) and the layout the signs of the inflight
aircraft destruction prior to the impact with ground were not detected. All the damage of the
airframe structure, engines and systems, as well as the disintegration of the avionics occurred due
to the application of the huge loads in the accident sequence far beyond the tensile strength.

1.16.2. The examination of the stabilizer jackscrew

The jackscrew elements (Fig. 42) in an assembly were retrieved from the aircraft wreckage
and transferred to the NTSB Material lab, where on 20.10.2016 they had been examined to perform
the physical and chemical analysis of the fractured areas and the developed bend with the
measurement of the inner diameter and the safety rod wall thickness. In addition, the works were
carried out on the identification of the metal samples of the jackscrew and safety rod with the aim
to determine whether they met the material specification requirements (AMS 6265 and
AMS 6411/6427). The examination of the jackscrew of the stabilizer trim mechanism
(p/n 07322P000-05, s/n 1847) was performed at the participation and under the guidance of the air
accident investigation team representatives, as well as with the participation of the AAIS experts.

Ballnut
Ballscrew
Gearbox
Stop end
Safety rod
Drive end
Fig. 42. The jackscrew assembly with the fractures (the fractures are pointed with arrows)
At the outcome of the examination the following was established:

 all the fractures are characteristic of the overstress loads application. The fracture
surface does not demonstrate the apparent preexisting cracks or any other anomalies;
 the fracture planes on both halves of the jackscrew were inclined to the longitudinal axis
by 45° and exhibited a comparatively rough surface;
 the fracture through the safety rod exhibited similar features to the jackscrew fractures;

 the features of the both parts material meet the requirements;

 the wall thickness of the jackscrew and safety rod as well as the inner diameter of the
safety rod meets the drawing requirements;
 the chemical composition of the jackscrew and safety rod material met the requirements
of the AMS 6265 and AMS 6411/6427.
It should be noted that the position of the jackscrew gimbal is consistent with the value of
the stabilizer position as per the FDR data.
1.16.3. Stabilizer trim control switch of the F/O control wheel
On the accident scene the stabilizer trim control switch (ensuring the “power supply”)10,
integrated to the F/O control wheel, had been discovered (Fig. 43).
Fig. 43. Stabilizer trim control switch of the F/O control wheel
Upon the investigation team decision on 24.10.2016 the trim switch (s/n 306-4305, Boeing
p/n 10-60705-1 Y) was examined at the facilities of the manufacturer Esterline Mason in the town
of Sylmar (California). The F/O control wheel trim switch examination was performed at the
participation and under the guidance of the air accident investigation team representatives, as well
as with the participation of the NTSB and AAIS experts.
The following was stated on the results of the examination:
 the trim switch sustained significant damage;
 the testing of the trim switch against the Acceptance Test Procedure is impossible;
 the continuity of the electric circuit is disrupted.
10
See also Section 1.18.2 of the present Report.
Throughout the service history of the stabilizer trim control switches, there were no events
of its short circuit.
1.16.4. The examinations of the elevator control system PCUs
The examinations of the Boeing 737-800 A6-FDN aircraft elevator PCUs had been
conducted.
PCU Boeing p/n s/n Date of manufacture Overhauls
Left 251A2160-2 14254 November 2010 -

Right 251A2160-2 14257 November 2010 -
The elevator control system PCUs examinations were conducted on

25.10.2016 – 28.10.2016 at the facilities of Parker Aerospace being their designer (the town of
Irvine, California, United States of America) and within Parker Aerospace (the town of Ogden,
Utah, United States of America) being their manufacturer (the further examinations of the control
valves are referred to in the case). The elevator control system PCUs examination was performed
at the participation and under the guidance of the air accident investigation team representatives,
as well as with the participation of the NTSB and AAIS experts.
The following works were carried out throughout the examination:
Visual inspection
Left PCU
The external surfaces of the PCU are significantly dirtied.
Left PCU (Fig. 44) sustained considerable damage in the accident sequence as follows:
Fig. 44. The left PCU external view
 the unions (supply, drain) are damaged, severed up to the retainers;

 the primary input arm is partly disintegrated, the portion of the arm with the bearing is
missing;
 the secondary input arm is partly destroyed – the housing of the secondary input arm is
disintegrated;
 the tailstock and the piston, the piston link and the actuator sealing bushings are missing.
Right PCU
The right PCU (Fig. 45) had been submitted for the examination with the elements of the
input pogos. The external surfaces of the PCU are significantly dirtied.
Fig. 45. The external view of the right PCU
The PCU sustained considerable damage in the accident sequence as follows:

 the actuator tailstock is deformed (bent);
 the secondary input arm housing is destroyed.
Computed tomography scanning
Prior to the teardown and bench test the CT scanning of the PCUs was performed by Varian
Medical Systems, Inc in Chicago, Illinois. The images were then examined, processed, and
analyzed by the NTSB in Washington, DC., USA with the participation and under the guidance of
the air accident investigation team representatives. The examination aimed at the evaluation of the
condition of the internal parts and assemblies of the PCUs in order to state the absence of the
damaged components and no event of the off-design configuration. The CT scanning results in the
PCUs 3D images with a capability to acquire the 2D images along any necessary cross-section.
The CT scanning results analysis was the evidence of the following:

over the left PCU:

 the internal components of the PCU do not exhibit any visible damage;
 all the internal parts and assemblies of the PCU are normally positioned at their standard
locations;
over the right PCU:
 the internal components of the PCU do not exhibit any visible damage;
 all the internal parts and assemblies of the PCU are normally positioned at their standard
locations.
The PCUs testing against the manufacturer ATP
The PCUs were tested to evaluate:
 the forces, required to move the input arms;
 the forces, required to move the one input arm against another one fixed;
 the full travel of the input arms;
 the bypass valve actuation pressure (the test was performed with the use of the
MIL-H-083 hydraulic fluid).
The PCUs parameters meet the requirements.
The PCUs teardown. The parts and assemblies visual inspection
The PCU were subject to the partial teardown in order to evaluate the condition of the
internal parts and assemblies, test the control and bypass valves against the manufacturer ATP.
The left PCU
The internal cavities were unsealed due to the partial destruction of the PCU. The internal
cavities were exposed to environment.
In the progress of the teardown the following parts and components were disassembled:
 the input filter F1: the condition of the filter screen is satisfactory, no external ruptures
were detected. On the bottom part of the filter frame, the corrosion traces were discovered11.
 the leverage: the primary input arm is partially destroyed. The retainer sustained
damage. The bearing unit of the primary input arm is in satisfactory condition – the bearing races
rotate smoothly without seizure;
 the bypass valve: the bypass valve parts exhibit traces of corrosion;
 the control valve: the parts of the control valve exhibit the traces of corrosion. The
primary slide crank input slot had a burr on one side with the raise of the material.
11
The investigation team is of the opinion that the traces of corrosion on this and the other parts were developed after
the accident due to their exposure to the environment.
Prior to the bench testing of the control and bypass valves their parts were soaked in Derust
PA HD250 solution to remove the corrosion deposits and then cleaned using a combination of
manual brushes and an ultrasonic cleaner.
The filter was flushed with isopropyl alcohol. The alcohol was collected in glass bottles
and retained for further examinations.
The right PCU
In the progress of the teardown the following parts and components were disassembled:
 the input filter F1: the condition of the filter screen is satisfactory no external ruptures
were detected;
 the leverage: the leverage system parts are in satisfactory condition;
 the bypass valve: the bypass valve parts are in satisfactory condition;
 the control valve: the control valve parts are in the satisfactory condition;
 the actuator: the piston rod and tailstock are bent. The sealing bushings are in the
satisfactory condition.
The filter was flushed with isopropyl alcohol. The alcohol was collected in glass bottles
and retained for further examinations.
The testing of the control valves against the manufacturer ATP
To perform the flow tests the left and right PCU control valves were assembled to the
required level. The assembled control valves were alternatively placed on the bench in the test
fixture that simulates the function of the control valve inside the PCU. The test was carried out
with the use of the MIL-H-6083 hydraulic fluid.
The control valves parameters meet the requirements, determined for the new assemblies.
The test of the bypass valves against the manufacturer ATP
The test was carried out with the use of the MIL-H-6083 hydraulic fluid.
The left and right PCUs bypass valves parameters meet the requirements, determined for
the new assemblies.
Borescope inspection
The following particles were subject to the borescopic inspection:
 the control valve secondary slide of the right PCU;
 the control valve secondary slide of the left PCU;
 the control valve sleeve of the right PCU;
 the control valve sleeve of the left PCU;
 the bypass valve sleeve of the right PCU;
 the bypass valve sleeve of the left PCU;
 the actuator cylinder of the right PCU;
 the actuator cylinder of the left PCU.

The traces (marks) were noted on the surfaces of the examined parts that may be
characteristic of the normal operation, as well as the marks of corrosion on the inner surfaces of
the left PCU. The parts that were subject to the borescope inspection do not exhibit any specific
marks of the mechanical jamming, seizure or the abnormal operation of the PCUs.
The diameter clearance measurement
Prior to the measurement the left and right PCU control valves parts surfaces were
thoroughly cleaned out from the hydraulic fluid flow marks and any other deposits.
As informed by the PCU designer the performance (flow) tests are the major criterion of
acceptance. However, there is a 0.00015 inch requirement to the minimum diameter clearance as
far as the primary slide – secondary slide assembly of the control valve is concerned.
The primary slide – secondary slide assembly of the right PCU does not meet the minimum
diameter clearance requirement. The results of the measurement are the evidence that this very
assembly had been manufactured with a deviation from the drawing or in the progress of the
manufacture it might be case of the deficient technological dimensional stabilization. The reduced
diameter clearance may cause chocking of the control valve slide due to the sleeve deformation,
being affected by different loads throughout the operation. At the same time all the parameters of
the control valve meet the requirements of the ATP. The control valve, having been integrated into
the PCU, had been operated for a long time against different external impact. Thus the mentioned
discrepancy of the minimum diameter clearance against the technological documentation
requirement had not anyhow affected the PCU serviceability.
1.16.5. The evaluation of the condition and the serviceability of the electric
stab trim motor
The examinations of the condition and the serviceability of the electric stab trim motor
were carried out within 25.10.2017 – 26.10.2017 at the Eaton facilities in the Grand Rapids, United
States of America. The NVM microchip and the stab trim motor p/n 6355C0001-01, s/n 2062,
were tested on the bench. The stab trim motor examination was performed at the participation and
under the guidance of the air accident investigation team representatives, as well as with the
participation of the NTSB and AAIS experts.
The NVM with the U8 integrated circuit, mounted to the 6355-0230-13 PWA, is
incorporated into the motor assembly. The NVM only records latched faults (faults that result in
the stabilizer trim motor becoming inoperative until the next power cycle) and corresponding
operational data. The NVM also records one initial cold boot record. The NVM can store up to
5459 faults. The faults are stored in order of occurrence. The non-latched faults are not stored in

the NVM. In case of repair performance at the manufacturer facilities (Eaton) the NVM is
downloaded and then cleared as a part of release to service.
The integrated circuit was soldered onto a verified good 6355-0230-13 PWA. The PWA
was then connected to an engineering test bench.
There were no faults recorded in the NVM.
The brushless DC motor, p/n 6355-0210-05 (Rev. C), s/n AM0039 was installed into a
known good 6355C0001-0l stabilizer trim motor assembly and then an assembly level acceptance
test was carried out.
The stab trim motor assembly with the accident motor installed was subjected to an
acceptance test per Eaton document CMM 27-40-10 Rev 7, paragraph 1G. The results were
recorded on an Eaton Acceptance Test Data Sheet.
The assembly passed all performed portions of the bench test.
1.16.6. The assessment of the language proficiency level of the approach
control unit officer
To assess the level of the language proficiency «The analysis of the assessment adequacy
of the overall and air English language proficiency level based on the results of the ATC officer
testing of 21.06.2013 and 15.03.2016» had been carried out. The activities were performed by the
Consulting and Analytic Agency for Flight Safety experts upon the investigation team enquiry. At
the outcome of the works, it was noted that in both cases the submitted rater assessment protocol
cannot be considered as the documented evidence of the ICAO Level 4 language proficiency. The
protocols incorporate the quotes from the ICAO Language Proficiency Requirements that describe
Level 4 but in the first case (of 21.06.2013), they do not contain any examples of the speech pattern,
indicative of its compliance with these descriptions. As for the second case of 16.03.2016, only
some mistakes in pronunciation, grammar structures and vocabulary out of speech pattern are
noted.
Based on the analysis of the audio recording of the both tests the following was concluded:
the concerned speech pattern does not comply with the ICAO Level 4 proficiency description.
The rater conclusion based on the results of both tests on the ICAO level 4 general and
aviation English language proficiency had been drawn up unjustified.
1.16.7. Engineering simulation (Mathematical modelling)
With the aim to determine the stability and controllability performance of the aircraft in
the progress of the accident flight against the performance of the type aircraft, to evaluate the range
of the potential external disturbances, which the aircraft had been affected by, as well as to
compute the non-registered parameters the engineering simulation of the flight has been performed
together with the kinematic consistency of the recorded data.

The kinematic consistency has been applied to correct the potential reciprocal imbalance
of the recorded data that is often observed due to the different (sometimes, insufficient) sampling
rate, the record of the same parameters by different means and the instrument errors. The kinematic
consistency allows integrating the recorded acceleration values to check the compatibility of the
primary inertial parameters: altitude, ground speed and drift angle. The mutually corresponded and
adjusted set of parameters is the output that allows computing the wind disturbances together with
the other parameters such as AOA and sideslip, TAS, etc.
The simulation was set up for about last 90 s of the flight (the simulation was stopped 3 s
before the impact with ground due to the model constraints). At the simulation the Boeing 737-800
six-degree-of-freedom mathematical model, adjusted as per the results of the flight tests, has been
used. The similar mathematical model is used into the Level D FFS.
The initiation (balance) of the mathematical model was carried out with the estimated
values of weight and CG, as well as for the actual parameters of flight (airspeed, etc.), the controls
deflection and the engines power mode, recorded by FDR. At the simulation, the estimated values
of the horizontal components of the wind were taken into account. The vertical component of the
wind, including the gusts, was disregarded as per the calculations it was inconsiderable.
The position of the high-lift devices, stabilizer and landing gear were taken according to
the actual data as per the FDR record.
In the progress of the simulation with the use of the “mathematical autopilot” the small
biases were applied to the recorded deflections of the control surfaces and controls to ensure the
correspondence of the estimated and recorded values of the pitch, bank and heading angles with
the verification on the other parameters, Fig. 46, Fig. 47 present the results of the simulation. On
figures the signs of the parameters comply with the convention, enacted in USA.

Fig. 46. The results of the engineering simulation (the pitch channel)

Fig. 47. The results of the engineering simulation (the lateral and directional channels)
On the results of the engineering simulation, the estimated data match the FDR ones very
closely. It means that the aerodynamic performance of the aircraft in the accident flight complied
with the performance of the type aircraft, the airplane motion was determined by the control
surfaces deflection, the engines power mode and actual wind disturbances. The airplane was not
anyhow affected by the other factors, for example by the icing.
The flight leg within the time of 0:41:14 – 0:41:21 (42068 – 42075 on the plots above), on
which the rapid decrease of vertical G from ≈ 1.35 g to ≈ 0.4 g with its subsequent restoration up
to ≈ 1.25 g occurred, has been analyzed apart. The load factor change in question was accompanied
by the “emotional” exhale of one of the crewmembers. It was found that this very change was
entirely due to the control inputs.
The crew capability to recover from the current state of the airplane by the control column
deflection only was assessed as well (while keeping current positions of all other controls including
stab trim, throttles and flaps). The simulation of the scenario, at which the control column full pull
position (the corresponding forces amount to about 125 lb (55 kg)) occurred at the point of time12,
concurrent to the stabilizer trim stop, when the aircraft in descent had been passing the true altitude
of ≈ 2000 ft (610 m), was the evidence that the aircraft recovery from the descent had occurred at
the true altitude of about 500 ft (150 m) with the further transition to climb (Fig. 48a).
As per FDR immediately prior to the aircraft impact with ground (IAS ≈ 340 kt) at the full
aft position of the control column the elevator deflection amounted to 5⁰. According to the
materials, submitted by the aircraft manufacturer, because the maximum output force of the PCUs
is limited, the maximum elevator deflection for the given conditions (the altitude of the flight, the
stabilizer position, AOA and sideslip angle values) is dependent on the airspeed.
The maximum elevator deflection vs airspeed dependency Fig. 48 represents the notional
dependency of the maximum available elevator deflection vs IAS. The numerical values of the
scales are not given as they constitute the Boeing Company trade secret. In the progress of the
investigation, the investigation team had been provided with the stated dependency, comprising
the numerical values. The convergence of the factual and calculated elevator deflection is
satisfactory.
12
As per the FDR data, at that very moment the control column had been momentarily deflected to a pull position at
approximately 2/3 off the neutral.
Fig. 48. The maximum elevator deflection vs airspeed dependency
Fig. 48a. Simulation of airplane recovery

1.16.8. On the control column forces

The pilots command the elevator, deflecting left or right control columns, which via the
system of cables, rods, pogos and the input torque tube transfer the control input to the PCUs input
rods. Mechanically, the PCUs with the pressure of the power fluid via the output torque tube move
the elevator. The aerodynamic forces of the elevator are not transferred to the control columns.
The Elevator Feel and Centering Unit is designed to provide feedback “on forces”. The force value
depends on the columns deflection, the stabilizer position and IAS of the flight. The forces,
generated at the control columns movement, are measured with the pitch control wheel steering
force transducers, which, in their turn, transfer the signals, proportionate to the forces value, to the
FCC. The separate force sensors are integrated to each of the control columns. The actual position
of the left and right control columns is measured by the control column position sensors. The
values, measured by the sensors in question, are recorded by the FDR. It is important to understand
that the measured forces values are the forces on sensors and not the forces on the control columns
themselves (i.e. these are not the forces that the pilots sense). For example at the application of
forces on the left column only, because of the mechanical coupling of the linkages, as normal both
columns will move simultaneously, with that the specific forces will be measured both with the
right and the left sensor, still in general their value will be different. If one pilot is on the controls,
the forces value “by his sensor” will be greater.
The control linkage integrates the control column breakout mechanism, which is activated,
when the forces by either sensor exceed 66 lb (30 kg). With that, the control column may be
actually deflected by only one pilot, whereas another pilot may be off the controls. At the decrease
of forces less than the threshold value, the linkages are again “coupled”.
Based on the FDR record (Fig. 49)13, in the accident flight until the time of 00:41:32 (42088
on the plot below), the forces, measured by sensors had not exceeded 66 lb and the breakout had
not occurred. After the indicated point of time the breakout of the control columns occurred, in
this throughout the entire interval the active control was carried out from the PIC’s duty station.
13
On figure the signs of the parameters comply with the convention, enacted in USA.
Fig. 49. The control column breakout mechanism activation and the “dual” pitch control
Within the time of 00:41:44 – 00:41:47 (42098 – 42101 on Fig. 49) the forces by the left
and the right sensor have the opposite signs that is the evidence of the divergent actions of the
pilots, of the PIC - to nose down, the F/O – to nose up. The indicated interval correlates to the
activation of the EGPWS alert.
Following the request of the investigation team, the aircraft manufacturer evaluated the
forces on the PIC’s control column itself. The predicted data are valid only if the control
is exercised from the one duty station. The computation results are given on Fig. 50. On figure the
signs of the parameters comply with the convention, enacted in USA.

Fig. 50. The estimated forces on the PIC’s control column
The results inside the crosshatched region may be unreliable as the control was exercised
with both pilots. The computations were the evidence that at the moment of the stabilizer trim
initiation to nose down the pilot flying, for the considerable time, had been applying the “pushing”
forces of about 50 lb (23 kg)14.
1.16.9. The reconstruction of the Head-Up Display/HUD readings
According to Rockwell Collins, the HUD manufacturer (USA), as of the moment of the
examination the method to recover the HUD readings, based on the data, recorded by FDR,
through the stream video was quite complicated and time-consuming (inclusive of the limitation
of the sampling rate of the parameters, recorded by FDR) and allowed obtain the result within the
limited time interval only. Following the investigation team request the manufacturer
reconstructed the HUD images at specific time instants. The reconstructed images and their
description are presented in the Analysis section of the Report.
The HUD manufacturer emphasized that the represented images were the reconstruction of
the indication only (and not the accurate depiction), among others implemented through the
interpolation of the available FDR data. The presented images should not be considered “the
14
Hereinafter the “pushing” (“pressing”) forces mean these that are generated on the control column at its forward
deflection (to nose-down). At the aft deflection of the control column (to nose-up) the “pulling” forces are generated.
complete and accurate reproduction of what the pilot was actually seeing”. They (the images) are
the most accurate one out of possible approximations of what should have been displayed on HUD
as per the available FDR data. The reconstructed images do not contain the data on a range of the
flight parameters, of which there is no information, as well as on the settings of the display itself
(for example the brightness) or on the possible impact of the outside environment (for example
clouds) on the display picture.
The comprehensive description of the HUD indication symbology is stated in the HGS
Model 4000 Pilot Guide for the Boeing 737 Series.
1.17. Organizational and management information

The postal address of the Rostov-on-Don airport: 344009, Russia, Rostov-on-Don, 270/1,
Sholokhov avenue. The airport is in the area of responsibility of the Southern FATA Interregional
Territorial Department.
The Dubai Aviation Corporation, trading as Flydubai.
Legal address: PO Box 353.
Dubai Aviation Building, Ittehad Road 353, Dubai, United Arab Emirates.
The UAE GCAA carries out the supervision of the corporation activities.
1.18. Additional information

1.18.1. On the fatal accident to the Ilyushin Il-86 RA-86060 aircraft at the
Sheremetyevo airport
On July 28, 2002 at 11:20 UTC, at daytime in VMC after takeoff the fatal air accident
occurred to the Ilyushin Il-86 RA-86060 aircraft, owned by the Pulkovo airline, FSUAE. The
aircraft was performing a ferry flight with no passengers aboard. The air accident resulted in the
death of 14 crewmembers, two persons sustained serious injuries.
The investigation team concluded the following15: The accident occurred because the
aircraft reached the critical AOA at the climb out and entered the stall owing to stabilizer trim to
the full nose up position (-12°) at the rate, consistent with the function of 4 hydraulic motors. The
created pitch-up moment could not have been counteracted by the pilots with the elevator.
The discrete signal “The press of the stabilizer primary control switches”, recorded by the
MSRP-256 FDR is the evidence of the stabilizer trim control input by one of the pilots in 2-3 s
after the aircraft liftoff from the runway”.
As the result of the examination of the survived structure elements, the stabilizer control
system failures had not been detected. The analysis of the schematic and structural design of the
15
This section cites the excerpts out of the investigation team conclusion on the investigation of the Il-86 RA-86060
aircraft fatal accident, which help to figure out the analysis and conclusion of the investigation team on the fatal
accident to the Boeing 737-8KN A6-FDN aircraft.
Il-86 aircraft stabilizer control system had not identified the potential failures which could have
led directly to the uncommanded stabilizer movement with a probability greater than “practically
improbable”.
Fig. 51 presents the takeoff parameters. It is notable that the stabilizer control switches
remained pressed 15 s more after the stabilizer setting reached the limit position. The release of
the switches is most probably related to the intensive control column input that could have led to
the change of the control wheel “grasp”.
Fig. 51. The takeoff parameters of the Ilyushin Il-86 RA-86060 aircraft on 28.07.2002
Legend to Fig. 51 (colors of filled dots):

black external radio communications
red GPWS warning
bright green pressing of the horizontal stabilizer main control switches
blue leading-edge slats are extended
azure leading-edge slats are deployed to 25 degrees
violet the crew readiness for takeoff/landing
yellow weight-on-wheels
khaki AOA too high
dark blue pressing of the landing gear retraction button
dark violet nose landing gear control by pedal ON
purple radioaltimeter failure
Legend to Fig. 51 (colors of lines):
first graph:
red horizontal stabilizer position

violet flaps' position
bright green vertical G
blue longitudinal G
second graph:
azure control column position
violet LH elevator (degrees)
bright green pitch attitude angle (degrees)
red true angle-of-attack
third graph:
bright green IAS (km/h)
blue radioheight (meters)
red pressure altitude (meters)
In the progress of work, the investigation team identified the following hazards that
correlate with the circumstances of the Boeing 737-8KN A6-FDN fatal accident:
 at the flight legs that imply the increased workload of the crewmembers or at their non-
optimal working condition, the time, within which the fact of the stabilizer trim is identified, is
significantly increased;
 throughout the operation the crewmembers get used to the stabilizer trim alert (as for
the Il-86 aircraft it is the callback signal on every 0.5⁰ of the stabilizer trim) and it is not “alarming”
for them;
 the stabilizer position indicator is not very good readable;
 at the long press of the stabilizer control switches the feeling of feedback forces (on a
finger that presses trim switches) may be lost as these are little and, consequently, the uncontrolled
stabilizer trim (the prolongation of the earlier started action) may occur, for example, at the
distraction of the pilot flying or at his non-optimal working condition (in the progress of the
investigation the occurrences of the switches “keeping be pressed” up to 34 s had been revealed16);
 the crewmembers consider the stabilizer control system reliable and do not anticipate its
failures that degrade the in-flight stabilizer position monitoring;
 not all the pilots have the necessary knowledge on the forces trim principle as for the
aircraft with the trimmable stabilizer.
1.18.2. On the stabilizer control at the Boeing 737 -800 aircraft
According to FCOM, Volume 2, page 9.20.7, the horizontal stabilizer is positioned
(controlled) by a single electric trim motor controlled through either the stab trim switches on the
PIC and F/O control wheels (on the left horn of the control wheel at the PIC’s duty station and on
the right one at the F/O’s duty station) or autopilot trim.
16
The “keeping be pressed” occurrences had been noted with both the crewmembers of the accident aircraft, and the
other crewmembers of Pulkovo airline.

Note: The stabilizer trim is only possible with the simultaneous press of two
independent switches on the either control wheel. One of the switches ensures
the direction of trim, whereas another one ensures “power supply”.
Stabilizer trim switches on each control wheel actuate the electric trim motor through the
main electric stabilizer trim circuit when the airplane is flown manually. With the autopilot
engaged, stabilizer trim is accomplished through the autopilot stabilizer trim circuit. The main
electric and autopilot stabilizer trim have two speed modes: high speed with flaps extended and
low speed with flaps retracted. If the autopilot is engaged, actuating either pair of stabilizer trim
switches automatically disengages the autopilot.
The STAB TRIM MAIN ELECT cutout switch and the STAB TRIM AUTOPILOT cutout
switch, located on the control stand, are provided to allow the main electric trim and/or autopilot
inputs to be disconnected from the stabilizer trim motor.
There are also control column actuated stabilizer trim cutout switches. These switches
ensure the stop of the stabilizer trim through the main electric stabilizer trim circuit or autopilot
stabilizer trim circuit at the actual deflection (position) of the control column by more than the
threshold value (≈ 4) to the direction, opposing the stabilizer trim.
Note: FCOM in the description of the stated function reads the following: «Control
column actuated stabilizer trim cutout switches stop operation of the main
electric and autopilot trim when the control column movement opposes trim
direction». The investigation team notes that the applied term “movement” is
ambiguous. For example, at the full push of the control column and press of the
trim switches in the same direction the stabilizer will move. If now, without the
release of the stabilizer trim switches, start the control column aft movement the
stabilizer will not stop. The stabilizer movement will only stop when the control
column will pass through neutral and will be positioned to a pull by more than
the threshold value.
When the STAB TRIM override switch is positioned to OVERRIDE, electric trim can be
used regardless of control column position.
Manual stabilizer control is accomplished through cables, which allow the pilot to position
the stabilizer by rotating the stabilizer trim wheels (two of these trim wheels are installed, one for
PIC and one for F/O). The stabilizer is held in position by two independent brake systems.
The stabilizer trim wheels rotate (making the hum), when the stabilizer control (trim) is
carried out on the either circuit: main electric or autopilot one. Manual rotation of the trim wheels
can be used to override autopilot or main electric trim. Grasping the stabilizer trim wheel will stop

stabilizer motion. With that, the effort required to manually rotate the stabilizer trim wheels may
be considerable under certain flight conditions.
The presence of the cutout switches, which are actuated at the control columns deflection,
establishes another feature at controlling the stabilizer with the trim switches on control wheels.
At the control columns position close to neutral the stabilizer movement, commanded by one of
the pilots (for example to nose-down) may be stopped by the other pilot by pressing “his own”
trim switches to the opposite direction (to nose-up). At the control column deflection and the cutout
switches actuation the stop of the stabilizer movement (which, as mentioned above, in this case is
only possible in the direction of the control column deflection) by another control wheel is not
possible. All the cases of the possible response of the stabilizer to the trim switches actuation at
the forward deflection of the control columns are given in the table below:
The control column The trim The trim
position (no switches position switches position
The stabilizer movement
mismatch as for on the PIC’s on the F/O’s
their position) control wheel control wheel
To nose-up To nose-up To nose-up
< 4 «forward» No movement (due to the control
To nose-up To nose-down
(the cutout switches commands mismatch)
are not actuated) No movement (due to the control
To nose-down To nose-up
commands mismatch)
To nose-up To nose-up No movement (due to the cutout
switches actuation)
To nose-down (the control
> 4 «forward»
To nose-up To nose-down command to nose-up does not
(the cutout switches
pass (cuts out))
have been actuated)
To nose-down (the control
To nose-down To nose-up command to nose-up does not
pass (cuts out))
1.18.3. On the PFD monitoring at the forward deflection of the control column
In the progress of the investigation team activities, a number of pilots noted that at the
correct posture at the seat with the gradual forward deflection of the control wheel the view of the
PFD indication is degraded (Fig. 52). At the deflection by ¾ of control column full travel the
significant part of the PFD “is blocked” to the pilot. The investigation team is of the opinion that,
as for the accident flight, this specific factor had no significant impact, as the PIC had been
controlling the aircraft by HUD, whereas the F/O had the correct concept of the aircraft attitude
and had granted the appropriate prompts to the PIC. At the same time, these features should be
additionally analyzed and briefed to the pilots in order to monitor the relevant risks.
Control column full forward position Control column ½ forward position

Fig. 52. The pictures of the PFD visibility at different positions of the control column
1.19. Useful or effective investigation techniques

Within the investigation, the reconstructed images of the HUD were used. Refer to the
1.16.9 and Analysis sections for details.

2. Analysis
2.1. Description of the flight
At the overnight into 19.03.2016 the Flydubai airline flight crew, consisting of the PIC and
F/O, was performing the round-trip international scheduled passenger flight FDB 981/982 en route
Dubai (OMDB) – Rostov-on-Don (URRR) – Dubai (OMDB) aboard the B737-8KN A6-FDN
aircraft.
The FDB 981 flight departure was scheduled for 17:45 (21:45 local time). Reportedly, the
crew arrived for the preflight check one hour before STD, which met the airline OM provisions.
The preflight preparation was carried out by the crew themselves in full under the supervision of
the PIC.
The PIC and F/O had never flown together prior to the accident flight. Furthermore, it was
the first flight to the Rostov-on-Don aerodrome for both crewmembers. As of the day of the
accident the PIC had performed 14 flights to the RF aerodromes, eight of which as a PIC. The F/O
had not performed any flights to the RF aerodromes before.
Note: The crew formation so composed had been consistent with the UAE GCAA and
the airline OM requirements. Specifically, OMA item 5.2.10.2 allows for the
performance of flight to the aerodrome, both flight crewmembers are not
familiar with, if there are the instrument approach charts for this aerodrome and
they are available to the crew at the flight preparation.
At the preparation and performance of the flights, the airline uses the Lufthansa Systems
LIDO aeronautical information. Every crewmember has an iPad with the installed software.
At the preparation of the flights to the RF aerodromes, the crewmembers undergo
additional briefings on the specific aspects of such flights performance.
The airline OM Part C incorporates specific sections on all the destination aerodromes. As
for the Rostov-on-Don aerodrome (Section 16.B.136), there is a warning on the application of the
specific restrictions “FLYDUBAI RESTRICTIONS APPLY”. Specifically it is determined that
the approach, landing and takeoff is performed by the PIC only (“CAPTAIN ONLY approach,
landing and take-off”). The non-standard glideslope angle of 240’ and the necessity to confirm
the setting in use, QNH or QFE, are noted as well. Additionally, there is a notice that the crew
should be aware of possible turbulence and windshear on final.
The investigation team considers that the airline OM reflected the specific features of the
flights both to RF in general and to the Rostov-on-Don aerodrome in particular.
According to the airline standards, the crewmembers receive their monthly flight schedule
before 25th of the previous month that ensures (if necessary) the sufficient amount of time for self-
study (including the study of the specific features of the destination aerodromes).

As per the available information (Section 1.5.1), the crewmembers held the valid pilot’s
licenses, have undergone all the necessary training and were ready to operate the flight.
The crewmembers were issued valid medical certificates.
Note: As it is indicated in Section 1.13, as per the results of the forensic medical
expertise of the PIC and F/O, some chemical substances, including ethanol, were
detected. With that, the expert team concluded that the presence of the chemical
substances was associated with the nature of the injuries and the remains
location environment and that it is not the evidence that the PIC and F/O, as of
the moment of the accident, had been in a state of alcohol or another toxic
(including the drug one) intoxication. Thus, the investigation team did not reveal
any signs of the consumption of alcohol and another psychoactive substances by
the PIC or F/O.
The analysis of compliance of the work and rest schedule within a record period (28
consecutive days) did not identify any violations. The crew had a sufficient amount of the preflight
rest. As per the submitted data, the Fatigue Management System is implemented in the airline. The
system encourages the fatigue-related confidential reports by the crewmembers for any stage of
the flight operations (the preflight, in-flight, post-flight one). For a number of quantitative
indicators the system goes beyond the national aviation legislation (that is it ensures the improved
conditions for the crewmembers). Since 2009, the airline has accumulated 450 000 flights with a
total flight time of more than 1 million hours. Within the period, 70 fatigue-related confidential
reports were submitted. The majority of them were proactive by nature – as the crewmembers
reported the fatigue presence and were removed from duty until they felt fit for flight operations.
The forecast and actual weather en route, for the Rostov-on-Don destination aerodrome
and the alternative aerodromes of Trabzon (LTCG) and Volgograd (URWW) did not impede the
performance of the flight mission and met the IFR. The crew took the appropriate decision to
depart. At the same time the weather package, that the crew were provided with, did not contain
SIGMET No 6 into the Rostov-on-Don FIR that incorporated the information on the forecast
severe turbulence to the south of 48° N and to the west of 48° E from GL up to FL150.
In the progress of the preflight check there were no anomalies detected as for the aircraft
systems and equipment functioning. As of the moment of departure, there were no MEL A and B
deferred defects. There was one C defect, deferred until 25.03.2016: «F/O FMC Alert Light
remaining ON upon self test». The defect in question is covered by MEL Section 34-36-02-01A.
The aircraft is valid for flight provided the respective FMC is not used for the autopilot guidance
during approach. There were another 7 D defects. The indicated deferred defects did not anyhow
affect the outcome of the flight. On the results of all the performed works, including the conducted

examinations, the investigation team did not identify any evidences of the technical failures in the
accident flight that could have affected the flight outcome.
Based on the data stated in the loadsheet and the computations, made by the investigation
team, the aircraft TOW at the departure from the Dubai aerodrome amounted to 68 tons with the
CG of 17.3% MAC, which did not exceed the limitations determined by the AFM (79015 kg and
10 -31% MAC respectively).
The aircraft was refueled with a sufficient amount of fuel to perform the flight along the
approved route with the consideration of the selected alternative aerodromes.
At 18:20 ABOT, the aircraft commenced movement after the passengers’ boarding was
completed. The delay of 35 minutes against STD was due to the late arrival of the aircraft from
the previous flight. At 18:37, the crew took off from the Dubai aerodrome.
At the altitude of ≈ 2700 ft (820 m) the engage of the right autopilot (autopilot B) is
recorded that may be indicative of the active control by F/O on this stage of the flight. The A/T
was engaged prior to takeoff. Further, it was an automatic flight.
At 18:59:30 FL360 was reached. The cruise flight was proceeded as assigned at the IAS
≈ 260 kt.
At 21:19, the crew disengaged the right autopilot and engaged the left one. Most probably,
from this point of time the PIC functioned as PF.
At 21:35:00 the FDR recording of the discrete signal “HUD IN USE COMBINER POSN”
(“Not stowed”) stopped. From the beginning of the flight up to this moment, the HUD functioned
in a primary mode/PRI. According to the airline SOP (Supplement D, Section 1.2) the use of HUD,
if it is operative is mandatory throughout the entire flight. All the airline flight personnel undergoes
the HUD initial and recurrent training (see Section 1.5.1). At 21:50:53, the HUD was again
reactivated in an IMC mode. It has not been possible to determine the reasons for such
“manipulations” with the HUD. The investigation team did not reveal any signs of the anomalies
of the HUD operation in the accident flight. There were no HUD operation anomalies, detected by
the crews that performed the previous flights either.
At 21:40 at the LAPTO reporting point the PIC contacted the UAMTS Rostov area center
В-1/В-2 combined sector radar control officer: «Rostov-Control, Good evening. SkyDubai 981,
flight level 360»17. The controller relayed to the crew that the aircraft had been identified.
As requested by the crew at 21:51:30 the controller relayed the actual weather at the
Rostov-on-Don aerodrome and the information about the active runway: «Actual weather for
17
The crew communication with the ATC and between them was conducted in the English language. In the Report
the translation into the Russian language, made by the investigation team with the consideration of the circumstances
of the flight, is given as well.
Uniform - Romeo - Romeo - Romeo at 21:30 Zulu time: wind – 250 degrees, 9 meters per second,
gusting – 15 meters per second, visibility – 5 kilometers, light shower rain, scattered clouds,
ceiling – 390 meters, broken clouds, cumulonimbus – 900 meters, overcast clouds, 3000 meters
ceiling, temperature + 6C degrees, dew point +3C degrees, runway in use – 22, braking action –
good, temporary wind – 250 degrees, 13 meters per second, gusting – 18 meters per second,
visibility – 1000 meters, shower rain, mist..».
At 22:12, the aircraft was transferred under the control of the UAMTS Rostov area center
Р-3/Р-6 combined sector. In contacting the Р-3/Р-6 sector controller the crew reported: «Rostov-
control, good evening or morning, Sky Dubai 981, maintaining FL 360, inbound KULOM».
The controller informed the crew that a SIGMET was in force: «Sky Dubai 981, and for
your information, we have SIGMET in our area from surface up to FL 150 for severe turbulence».
The crew asked to specify the area with a severe turbulence, for which the controller
communicated that SIGMET was valid into almost the entire area, but no aircraft had reported
turbulence.
At 22:17 the crew requested and was cleared by ATC for descent to FL190, after which
started the descent.
The CVR stored the record of the crewmembers communication within little more than
2 hours of flight18. With that, due to the long flight in the holding area (see here below) there was
no record of the landing briefing stored. In view of this the investigation team is not aware what
features of the approach had been discussed by the crewmembers (the communication on what go-
around procedure would be applied and under what conditions, namely).
At 22:24, while passing FL220, the crew requested clearance to continue descent, for which
the controller instructed the crew that they contacted Rostov Approach.
At 22:24:25 the Rostov Approach controller, after the crew reported approaching the
ER/YEGORLYKSKAYA reporting point and having ATIS UNIFORM that incorporated the
information on the moderate windshear, issued clearance for descent to FL60 on the ER 22A
STAR (Fig. 30).
At 22:33, in the progress of reaching FL60, the aircraft was transferred under the control
of the Rostov Radar officer. At 22:33:14, the crew contacted the Rostov Radar controller, reported
passing FL60 and requested further descent: «Control, good evening! SkyDubai 981. Passing
FL 60, further descending, FL 60». The controller issued clearance for further descent on ER 22A
STAR to 600 meters, QFE 990 hPa. According to the approach chart (Fig. 32 and Fig. 33) the QFE
altitude 600 m corresponds to 2250 ft QNH.
18
The record duration meets the established requirements.
The crew at their callout reported the descent to 800 m, but not to 600 m, however the
controller did not correct the crew that is inconsistent with FAR-362 item 2.13.3.
Note: FAR-362 item 2.13.3:
If the aircraft crew repeated the clearance or instruction incorrectly, then the
controller relays the word “correction” (“wrong”), followed by the content of
the correct clearance or instruction.
At this flight stage, the crew set 3300 ft19 (1000 m) as the target altitude. The flight
parameters of the first approach are given on Fig. 53. On the plot the parameter “relative altitude
(adjusted for baro correction)” stands for QFE altitude (the altitude above the aerodrome level).
19
According to the approach chart (Fig. 30) the altitude of passing of one of the fixes (magnetic azimuth 051⁰ ,
distance 23.8) is 3240 ft QNH (900 m QFE).
Fig. 53. The flight parameters at the first approach
At 22:35:42, the crew set the pressure of 990 hPa on the pressure altimeters. At this time
in descent, the aircraft was passing FL44 (with the established transition level 50).
The set value of the atmospheric pressure was consistent with QFE. The airline aircraft are
not equipped for the flights on QFE. Consequently, the airline OM prescribes the flights on QNH
only with the use of altitude in feet. As for the flights to countries with the metric system and/or
to those that use QFE, the crews undergo specific training on the use of the conversion tables. As
mentioned above, the crew clarified the QNH value (1000 hPa) with the ATC. All the further
actions by the crew are the evidence that the altitudes set on MCP and reached (including the go-
around altitude) were set based on QNH, that is the 990 hPa pressure was an incorrect setting. The
difference in pressure of 10 hPa is consistent with the difference in altitude of 280 ft (85 m) that
corresponds to the aerodrome elevation. That is to say until the glideslope interception the crew
was proceeding the flight on altitudes by a certain amount higher, than the target one. The error of
the pressure setting did not affect the outcome of the accident flight. Still the investigation team
notes that the use of QFE at the RF airports instead of QNH that is used practically in all the other
countries, presents extra risks.
At 22:36:11, the Rostov Radar controller contacted the crew and issued clearance for an
ILS approach to RWY 22.
At 22:37:44 the aircraft was transferred under the control of the Rostov Tower controller:
«Sky Dubai 981, contact Rostov Tower, 119,7. Have a nice landing! See you later».
At 22:38:0020, the F/O contacted the Rostov Tower controller and reported the descent to
2600 ft21 and the readiness to capture the localizer.
At 22:38:26, the aircraft reached the altitude of 2600 ft (800 m)22. The PIC informed
the F/O that 2300 ft (700 m) would be the next target altitude23.
The controller relayed to the crew the actual wind data: «wind 240º, 11, gusts 15mps» and
issued the clearance to land to RWY 22.
At 22:38:43, the localizer was captured, at 22:39:00 so was the glideslope. By the moment
of the glideslope capture the flaps were extended to 10 (the crew incrementally extended flaps to
1, 5 and 10 at the respective speeds). At the moment of the glideslope capture the distance to
runway amounted to ≈ 16.5 km (the glide path interception point is located at the distance of
12.52 km (6.76 nm)). The aircraft was proceeding at the altitude of 2600 ft (800 m) in a horizontal
flight, with that the altitude of the glideslope capture is 600 m.
20
It is with this sentence that the CVR record begins.
21
According to the approach chart (Fig. 30), the altitude of passing of one of the fixes (magnetic azimuth 040⁰ distance
19.8) is 2580 ft QNH (700 m on QFE).
22
Hereinafter at the description of the first approach and go-around, the altitudes are given in relation to the runway
level (on QFE).
23
According to the approach chart (Fig. 32) the altitude of the G/S capture is 2250 ft QNH (600 m QFE). According
to the CVR record, the PIC intentionally increased the target altitude by 50 ft.
After the glideslope was captured, the F/O, commanded by the PIC, set the go-around
altitude, equal to 2300 ft24.
After the initiation of the glide path descent the flaps were incrementally extended to 15,
25 and 30, the landing gear was extended as well. By the point of time of 22:40:17, at the altitude
of 2050 ft (625 m) at the distance of 12.2 km, the aircraft was in landing configuration.
The current weight was equal to ≈ 58.5 t. The Vref as for the approach with flaps 30⁰ is
140 kt. After the flaps extension to 30 the crew set the approach speed of 150 kt on MCP, that is
the correction to Vref amounted to +10 kt (according to the cockpit communication it is these Vref
values and corrections that the crew had calculated). The last weather information, relayed to the
crew by the Rostov Tower controller: wind 240 degrees25 11 m/s gusts 15 m/s. According to
FCT 737 NG (TM), the Command Speed section, pages 1.11-1.12, when calculating the approach
speed, the half stable headwind component is counted plus the full value of gusts that exceeds the
stable headwind component. With that, the maximum correction should not exceed 20 kt. For the
actual conditions the stable headwind component value amounted to: 11*1.94*cos(240-
218) ≈ 20 kt, the exceedance of the gust value against the stable component: (15-11)*1.94 ≈ 8 kt.
Thus, the recommended correction amounted to 20/2+8 = 18 kt, that is the speed, selected by the
crew, was even 8 kt less against the recommended one.
After setting of the approach speed, the crew performed the LANDING section of the
Checklist.
Up to the altitude of 1850 ft (560 m), the approach was performed with the A/P and A/T
engaged. At 22:40:40, the PIC (PF) made the decision to proceed the approach in a manual mode
and disengaged the A/P and A/T, having called it out to the F/O. At the point of disengage the A/P
ensured the glide path descent (the flight, driven by the localizer and glideslope signals), the A/T
maintained the speed, set on MCP. There were no deviations off the beam, the IAS altered within
a range of 145 –155 kt, with that from the altitude of ≈ 2600 ft (800 m) the aircraft entered the area
of turbulence (the vertical G altered within a range of 0.8 –1.25 g).
It has not been possible to clearly determine the reasons for the A/P disengage. Probably,
it was caused by intensified turbulence. After the A/P disengage the PIC spoke out that “It should
be a bit bumpy and then later should calm” and then that he is “flying the cue the speed will come
back slowly”.
Note: In the communication, the PIC used the term «cue». This sentence context, as
well as with the consideration of the airline OM provisions and the HUD, being
active (in working condition) (as per the FDR data), the investigation team
24
According to the missed approach procedure, the pattern altitude is 2250 ft QNH (600 m QFE).
25
The landing magnetic heading was equal to 218 ⁰.
makes the conclusion that the PIC was controlling the aircraft, following the
HUD indications. At this flight stage the Guidance Cue was obviously used. In
this case, the pilot's task is “to overlay” the Guidance Cue by the Flight Path
Symbol (Fig. 54).
WINDSHEAR WARNING
SSymbollPath Поме
стите
Guidance Cue
П
Flight Path Symbol
здесь ваш
П текст
Fig. 54. The HUD reconstructed image at the point of time of 22:42:02 (the windshear warning trigger) 26
It should be noted that the airline OM encourages the autopilot use on all stages of flight,
besides takeoff, go-around (accordingly, as per the OM in progress of the approach the use of two
autopilots is prohibited) and the low visibility approach. There were no conditions to follow the
low visibility procedure. Thus the A/P disengage on this very flight stage was only due to the
respective decision by the PIC.
The PIC performed further approach controlling the aircraft manually with the use of HUD
with good accuracy of the flight path hold on localizer and glideslope. The F/O was monitoring
the flight with the use of conventional instruments. The investigation team notes that the use of
HUD on this very segment of the flight, probably, had a positive effect as for the quality of the
manual control, as the HUD supplies information on the deviation off the ILS beams in a larger
26
Hereinafter the pictures with the HUD reconstructed images represent only the parameters, the reconstruction of
which had been possible with a sufficient degree of accuracy based on the FDR record.
scale against PFD and ND. For instance, the scale of the signal of the localizer beam deviation is
increased by 6 times against the conventional indication.
At 22:41:50 on the altitude of ≈ 1240 ft (380 m), the PIC told that he could see the runway.
The aircraft was then at the distance of 7 km off the runway. According to the ATIS information
at that point of time: cloud scattered at 480 m.
At 22:42:02 on the altitude of 1100 ft (335 m) the aural «GO-AROUND, WINDSHEAR
AHEAD» warning was triggered. There are two windshear detection systems on the airplane:
Predictive Windshear (it is a predictive alarm that resides in the Weather Radar system and informs
the pilot when the radar has detected a windshear event ahead of the airplane, ≈ 10-60 seconds
ahead typically) and Reactive Windshear (this very alarm resides in the EGPWS and informs the
pilots when they actually are in a windshear event). The analysis revealed that the warning was
triggered by the predictive alarm. Concurrently, the respective information was displayed on HUD
(Fig. 54), PFD and ND. It should be noted that the Reactive Windshear warning had never been
triggered either on this very segment of flight, or in the progress of the subsequent flight.
It was an almost instant response by the crew, about 1 s from the aural message trigger!
This is the evidence of the crew having been prepared for such kind of events and well trained on
the simulator as for the practice of actions in windshear. Most likely, the HUD had a positive effect
at that, too, as the windshear warning is very vividly noticeable.
22:42:02,127 22:42:04,3 AV Go around windshear
ahead.
22:42:03,2 22:42:03,7 F/O Windshear.
22:42:03,6 22:42:04,9 CPT Windshear. Go around.
The PIC made the decision to initiate Windshear Escape Maneuver, that is to perform go-
around with the aircraft configuration unchanged (without the retraction of landing gear and flaps)
at the maximum thrust (the maneuver is described in QRH D6-27370-8KN-JXB, page MAN.1.10).
Note: QRH D6-27370-8KN-JXB р. MAN.1.9:
Predictive windshear warning during approach (“GO–AROUND,
WINDSHEAR AHEAD” aural) perform the Windshear Escape Maneuver, or, at
pilot’s discretion, perform a normal go–around.
It should be noted that the operational documentation (for example, OM and FCOM) does
not determine the criteria, which the pilot could follow when making a decision to choose between
the standard go-around and the Windshear Escape Maneuver. Upon that if in actual windshear, it
is no alternative but for the pilot to perform the Windshear Escape Maneuver.
27
Hereinafter, as for the tables that contain the CVR transcript, in the first column the time of the sentence beginning
is given, in the second – the time of the sentence end.
At 22:42:04 the TO/GA mode was activated and the maximum thrust was established as
prescribed by QRH («Aggressively apply maximum thrust»). The actual N1 was 101-102% and
was considerably higher than N1, required for GA that could have been set at the A/T engaged.
Note: On this airplane, the function was incorporated of the redиced go аrоuпd thrust.
The concept of the function is that at the go-around with no need of maximum
thrust of the engines, at the first press of the TO/GA button the A/T (if engaged)
sets some estimated (based on the actual flight conditions) thrust that is
sufficient for climb with the target gradient (the estimated RPM value is not
recorded by FDR but the pilots may refer to N1 Limit Page or CDS). Another
press of the TO/GA button sets the relevant thrust value, indicated by the green
bug on the Thrust Mode Display (as for the accident flight – 97 % of maximum
thrust). The pilots can set the maximum available thrust only manually by
advancing the thrust levers full forward, which had been actually done by the
PF.
After the TO/GA mode activation the HUD automatically switched from IMC to PRI mode
(Fig. 55), at the same time the Guidance Cue had become “filled” that is took the form of
Windshear Guidance Cue, which is a signal of the Windshear Escape Maneuver carry out.
Windshear
Guidance Cue
SSymbollPath
Поме
стите
здесь
Fig. 55. The HUD reconstructed image at the point of time of 22:42:04 (after the TO/GA mode activation)

ваш
In the PRI mode, the HUD “director” indication duplicates the PFD command bars readings
that are computed by FCC. Fig. 56 presents the HUD reconstructed image at the point of time of
22:42:08.5, when the position of the pitch command bar was maximum to nose-up. To maintain
the target climb path the pilot needs to adjust the aircraft symbol to the TO/GA pitch target line.
Alternatively, the pilot may proceed the aircraft control by overlaying the Guidance Cue and Flight
Path Symbol.
TO/GA Pitch
Target Line
SSymbollPath П Aircraft
Reference symbol
омести мести
те те
здесь здесь
Fig. 56. The HUD reconstructed image at the point of time of 22:42:08.5
ваш
Initially in 10 sec, the PIC reached the pitch of 14…15º, following the HUD commands
ваш
(Fig. 57) and the FCOM and QRH guidance. To counteract the excessive pitch-up moment,
generated at the engine thrust increase to maximum, the PIC was forced to move the control
текст
column significantly forward.
текст

Fig. 57. The HUD reconstructed image at the point of time of 22:42:11
At 22:42:08, the F/O on the PIC’s command reported go-around. The Rostov Tower
controller acknowledged the information and instructed the crew that they contacted Rostov Radar
121.2.
The go-around altitude, set on MCP, was 2300 ft, at 1000 ft below that altitude (at
22:42:13), as per the AFDS logic, the ALT ACQUIRE mode was activated (Fig. 58).

Fig. 58. The HUD reconstructed image at the point of time of 22:42:13 (the ALT ACQ mode activation)
The crewmembers did not speak out the activation of this mode, maybe because the F/O
was then occupied with the communication with ATC. According to the airline within the flight
personnel training the attention is drawn to the need to follow the AVIATE-NAVIGATE-
COMMUNICATE principle. Probably such an early report on go-around “to the detriment” of the
flight parameters monitoring was related to the necessity to request further climb (higher of the
target go-around altitude) as the windshear warning was still displaying.
Indeed, to reach the target go-around altitude the aircraft should have climbed 1200 ft
(365 m) only. Taking into account high thrust-to-weight ratio of the aircraft (especially with the
maximum thrust of the engines) and high rate of climb it had been difficult to be established on
the target altitude without significant reduction of thrust. The Low Altitude Level Off - Low Gross
Weight of FCT 737 NG (ТМ) incorporates the recommendation as follows: if full go-around
thrust is used, reduce to climb thrust earlier than normal. The operational documentation does
not contain information that is more specific: at what stage, up to what value, based on what flight
parameters should the thrust be reduced. In case of ALT ACQ activation and active windshear
warning the recommended actions are not specified either. According to the information,
provided by the airline, as for the flight personnel training, the above-indicated aircraft
manufacturer documentation is used. The OM does not contain any “strict” quantitative

instructions. In the progress of the simulator training, the flight instructors assess the quality of the
exercises performance in a subjective manner, and, if required, another exercise is arranged.
After the activation of the ALT ACQUIRE mode, without being cleared for further climb,
the PIC started to reduce pitch with the additional deflection of the control column to “nose-down”
(up to 8.5° with the maximum deflection available of 13.75) and, consequently, with the
associated application of additional “pushing” forces. The actions in question with the engines on
maximum thrust resulted in the increase of airspeed. The F/O drew the PIC’s attention to the need
of monitoring the speed.
(Rostov Ra...) check the
22:42:28,9 22:42:31,1 F/O
speed.
22:42:32,7 22:42:33,7 CPT Speed checked.
At the moment the Flight Path Symbol on HUD was higher than Guidance Cue (Fig. 59
and Fig. 60), that is to maintain the target flight path the additional control inputs to “nose-down”
were required. With the control column position unchanged the PIC, in addition, pressed the trim
switches on the control wheel for about 4 s to nose-down (the angle of deflection of the stabilizer
was changed for 1.8º) that resulted in the pitch reduction up to 3-4 and additional acceleration.
Fig. 59. The HUD reconstructed image at the point of time of 22:42:27.5 (the first press on the stabilizer trim
switches)

Fig. 60. The HUD reconstructed image at the point of time of 22:42:31.5 (the last press on the stabilizer trim
switches)
As the result the IAS exceeded the limit value (with flaps, extended on 30°, Vfe is 175 kt)
and TE Flap Load Relief was activated. The flaps automatically retracted to 25 (the flap handle
remained at the 30 detent). It should be noted that at the initiation of go-around IAS was equal to
160 kt that is the margin to the limit (with no “flap load relief”) at the performance of Windshear
Escape Maneuver amounted to 15 kt.
At 22:42:34, the aircraft reached the target altitude of 2300 ft and, despite all the actions
by the crew, continued to climb. The PIC, without being cleared for further climb, reduced the
thrust (down to 83 % N1), although the windshear warning was still displayed. The reduction of
the engines thrust led to the reduction of the pitch-up moment. The control column was returned
to an almost neutral position.
The IAS had continued to increase for a while (the maximum reached value was 182 kt),
what the F/O drew the PIC’s attention to again.
22:42:47,7 22:42:48,2 F/O Check the speed.
22:42:48,9 22:42:49,3 CPT Checked.
Actually, due to the activation of Flap Load Relief, the speed exceedance with the extended
flaps was low and short in time (as per FCOM D6-273 70-8KN-JXВ page 9.20.19, with the flaps
set to 30° the function is activated at the speed of 176 kt, i.e. 1 kt more than Vfe). However, the

exceedance itself and the further climb with the ТЕ FLAP LOAD RELIEF activated within about
a minute (up to the moment of flaps resetting to 15° at 22:43:33) cannot be ignored. In the progress
of flight in the holding area, the pilots discussed the occurrence of the overspeed
(23:25:22…23:26:32).
How was the go around,
23:25:22,4 23:25:24,6 CPT
was it a mess or was ok?
23:25:24,8 23:25:25,3 F/O That's ok.
The only thing is that the
... the speed went to
23:25:25,6 23:25:32,5 F/O Barberpole but it's normal,
because we have this wind
shear.
23:25:27,6 23:25:28,2 CPT The speed.
23:25:32,9 23:25:33,3 CPT Yea.
What do you mean on the
23:25:33,8 23:25:35,0 CPT
go around?
23:25:35,6 23:25:36,8 F/O No, after the go around.
We were climbing did you
23:25:36,9 23:25:39,8 F/O note that the speed going
up?
23:25:37,0 23:25:37,4 CPT Yea.
23:25:40,5 23:25:40,9 CPT Ok.
23:25:40,7 23:25:42,0 F/O To the… to the up.
23:25:42,5 23:25:43,5 F/O We call it bad (illeg).
23:25:43,9 23:25:44,8 F/O Very good, very good.
23:25:45,5 23:25:46,8 F/O Did you notice anything?
So, (illeg) went into the
23:25:46,9 23:25:48,0 CPT
reds?
23:25:48,6 23:25:50,3 F/O Went almost into the red.
23:25:50,4 23:25:51,4 CPT Almost (illeg).
Maybe... maybe one or
two knots but it's nothing.
23:25:50,9 23:25:56,2 F/O
You have a fif… fifteen
knots to get into red.

23:25:56,9 23:25:57,6 F/O So, it was ok.

At placard speed, we were
23:25:58,4 23:26:01,6 F/O
with flaps thirty.
23:26:02,0 23:26:03,1 F/O That moment.
Or was it flaps thirty, one
23:26:03,2 23:26:06,2 CPT
seventy five?
23:26:06,2 23:26:06,5 F/O Yea.
Did we do one seventy
23:26:06,5 23:26:08,1 CPT
five or no?
Eh, I don't think so,
maybe... Maybe a few
23:26:08,4 23:26:14,1 F/O
seconds unable hold
altitude.
23:26:14,5 23:26:15,0 CPT Ok.
I think it was ok... the go
23:26:28,1 23:26:30,2 F/O
around was ok.
23:26:30,6 23:26:30,9 CPT Yea.
23:26:31,9 23:26:32,5 F/O Was ok.
The pilots do not mention the ТЕ FLAP LOAD RELIEF activation. It is not clear out of
the conversation, if either of the pilots even notice the activation of the function.
The PIC, apart of the above-given dialogue with the F/O, discussed the overspeed even
with the cabin attendant (at about 00:09). Among other things, he told that because of the
overspeed that would be a need for maintenance inspection.
Note: 1. The discussion of the issue with the cabin attendant and the context of the
concerned conversation show that, most probably the PIC, after landing, was
going to report the overspeed on extended flaps, this is in favor of the
confidential reporting system efficiency in the airline.
2. According to the airline, in his flying experience, the PIC, back when having
performed the flights as a F/O, had one overspeed occurrence. The F/O did not
have such occurrences before.
At 22:42:34, the crew contacted Rostov Radar controller and reported go-around, with that
requested further climb: «Rostov Ra... A… approach. Eh… This is SkyDubai niner eight one in go
around runway two two. Request to climb higher», as the respective windshear warning was kept
displayed (up to 22:42:47). Initially, the crew requested clearance for further climb without

specifying the altitude value. The controller cleared to reach the altitude of 900 m (2950 ft). After
the clarification of the altitude, up to which the controller cleared the climb, the crew requested to
reach the altitude of 5000 ft (1525 m), however was cleared on the second attempt, after informing
the controller about windshear.
In the progress of climb, at 22:43:14, with the actual setting of flaps to 25º, at the altitude
of 3750 ft (1150 m) and speed of 176 kt, the PIC engaged A/P and A/T.
At 22:43:32, the crew initiated landing gear and flaps retraction to 15°. Within about a
minute, the flaps were fully retracted. The go-around was complete.
Thus, the crew actions in response to the activation of the Predictive Windshear Warning
and in the progress of go-around, in general, had not created any significant risks to the flight
safety. At the same time, the crew run into one of the most unfavorable scenarios: the necessity of
the go-around due to the possible windshear encounter from quite a high altitude (with a little
margin as far as the target go-around altitude) on a low weight aircraft. The investigation team is
of the opinion that the scenario in question as for the manufacturer and operator documentation
needs the further elaboration in terms of recommended detailed procedures.
At 22:44:30, the Rostov Radar controller instructed the crew that they maintained FL50
with a heading to «Bravo - Alpha»28. In 15 sec, he relayed to the crew that they were number two
for landing: «SkyDubai niner eight one, for your information, you are number two to land. And
report reason for go around». In response to the controller request on the reason for go-around the
crew reported: «Wind shear was the reason of our go around, SkyDubai niner eight one».
At 22:45:15, the crew set the standard pressure of 1013 hPa. The aircraft flight path after
the go-around is given on Fig. 61. On the figure the only crew-ATC communication is stated,
which is referred to in the further text.
28
The BA (Bagayevskiy) NDB (Fig. 30).
Fig. 61. The flight path after the first go-around and in the holding area

On the Rostov Radar controller question at 22:45:20 if they need vectors for «Bravo -
«Alpha» the crew responded: «We are now avoiding weather on heading one four three. SkyDubai
niner eight niner... niner eight one».
The crew had not immediately performed the turn, prescribed by the go-around procedure,
but first proceeded with the landing heading up to reaching the altitude of 5000 ft and then with
heading ≈ 145 in attempt to avoid the area with “bad weather”, having reported it to the Radar
controller. In parallel, with the use of the available instruments the crew was analyzing the speed
and direction of the moving of the “bad weather” area. At 22:47:26, the crew requested climb to
FL080 in order to leave the moderate icing area.
Ah… Negative, sir.
Request ah… to maintain
this heading and we
22:47:26,7 22:47:35,7 CPT request to climb flight
level eight zero ah…, we
are experiencing ah…
moderate icing.
The Radar controller cleared the climb to FL080 and instructed the crew that they contacted
the Approach controller.
FL080 was reached at 22:49:34.
After weather avoidance was completed, at 22:50, the PIC performed the left turn directly
towards the BA NDB. Apparently, the PIC had not made the decision yet by the moment, whether
to perform another approach immediately or to proceed to holding area. Concurrently, the Aeroflot
SSJ-100 (AFL 1166) aircraft had been approaching. The PIC intended to wait until the information
on the result of the AFL 1166 flight approach would be available.
I want to see what this guy
22:52:22,6 22:52:24,3 CPT
will do.
SkyDubai niner eight one,
22:52:31,9 22:52:36,0 APR turn left heading three two
zero.
Negative we want to find
information the previous
22:52:36,1 22:52:40,9 CPT
aircraft if it landed or go
around.

We prefer... we prefer
proceed to Bravo Alfa to
hold there and then then
22:52:40,0 22:52:50,5 F/O get the information from
the preceeding traffic
about the weather.
SkyDubai niner eight one.
In this way, the crew actions on this stage of the flight and the decisions made by the PIC
were logical and reasoned.
At 22:53:31, the approaching AFL 1166 flight reported go-around and the presence of
windshear at the altitude of 600 m (1970 ft) to the Rostov Radar controller.
Proceeding heading BA, the crew was relayed by the controller that AFL 1166 performed
go-around due to windshear.
for information, previous
22:54:04,3 22:54:13,6 APR traffic ah… went go
around... went around due
to wind shear.
Copy information of go
around of the preceeding
22:54:14,0 22:54:20,2 F/O traffic, SkyDubai niner
eight one. Going to Bravo
Alfa to hold.
This information was important for the further decision making and was the evidence that
the conditions of windshear on final had been preserved. The PIC decided to proceed the flight in
the holding area for indefinite time and further on, depending on circumstances, make the decision
to undertake another approach or divert to an alternate aerodrome. The fuel on board (nine tons
approximately) was quite sufficient to proceed in a holding pattern for more than two hours.
Note: The UAE aviation legislation stipulates that it is the PIC, who always makes the
decision on the quantity of fuel aboard the aircraft. The only exception to this
rule is when the airline requests additional refuel based on its cost at the
destination aerodrome. In the accident flight, the aircraft had been additionally
fueled due to the higher fuel cost at the Rostov-on-Don airport compared to the
Dubai airport. The Sabre FPM that is used by the airline computes when the

additional refueling is cost-effective. This information is available to the flight

crewmember and the Dispatch. However, even when the additional refueling is
being requested, the PIC has the right to cancel it, depending on the specific
operating conditions (for instance if the increased landing weight affects the
landing performance of the aircraft).
At the same time it should be noted that within 00:11– 00:11:30, when the PIC had got out
of the cockpit, the F/O in talking to the cabin attendant, had expressed concern about the extended
flight at the holding area. The following utterances had been recorded, in particular:29
Todos los aviones se han All the aircraft have left.
marchado. Solamente We are the only one left
00:11:08,4 00:11:12,5 F/O
quedamos nosotros aquí here doing nonsense.
haciendo el tonto.
00:11:12,4 00:11:13,4 C/A ¿Y a dónde se fueron? Where did they go?
Pues no sé… porque era I do not know… because
un Aeroflot y el otro no sé there was an Aeroflot
qué era… y se fueron por and… I do not know about
00:11:14,0 00:11:23,3 F/O ahí. Se han ido a sus otros the other…..they left.
destinos. Nosotros They went to their other
tenemos suficiente destinations. We have fuel
combustible. enough.
Pero no creo que … con el But I do not think that…
tiempo como sigue, si with such a weather, if it
00:11:24,2 00:11:28,8 F/O
sigue así de tocho no keeps being bad it is not
merece la pena. worthy.
And then later:
La verdad es que no Actually I don’t

entiendo cómo pueden understand why they plan
poner este tipo de vuelos a this type of flights to this
este sitio de Rusia por la Russian place at night,
00:13:14,5 00:13:22,5 F/O
noche, cuando saben que when they already know
el tiempo es una mierda ya during the daytime that
por el día ¿lo ponen por la there is a shit of weather,
noche? they plan it at night?
29
The conversation was conducted in Spanish. In the text the translation into English is given, undertaken by CIAIAC.
At 22:55, the PIC transferred the controls to the F/O, then spoke out the reason for the delay
of landing due to safety reasons and offered apologies to passengers.
At 22:57, the crew initiated right turn at FL80 in order to enter holding area over Bravo
Alpha, of which they communicated the ATC at 22:57:24.
Further on, to avoid hold into “bad weather” in the BA area, the PIC tried to coordinate
with the Approach controller the clearance to proceed a holding pattern to the north of the
aerodrome, for which first was cleared to make holding pattern “at convenience”.
Ah… Just to ask you
Madam. Ah… Do you
have any published hold
ah…north of the airfield
cause ah… from the wind
22:57:35,1 22:57:53,6 CPT
direction ah… we can see
that the ah… weather is
coming towards ah… our
point so we prefer to ah…
hold north of the airfield.
22:58:40,8 22:58:46,1 APR making holding pattern at
convenience.
Ok. Thank you very much.
ah… We're gonna start
heading north and we will
do a present position hold
22:58:47,0 22:58:58,2 CPT ah… when we find a place
where is no icing.
We let you know. Thank
you.
At proceeding the flight with heading 310 in order to find the suitable weather conditions
for holding, the crew at 23:06:35 was relayed new information that the AFL 1166 flight in the
progress of another approach, performed go-around again due to windshear.

SkyDubai niner eight

one, previous traffic
23:06:35,5 23:06:44,3 APR
went around due… again
due to wind shear.
Yea. Thank you very
23:06:44,2 23:06:45,2 CPT
much.
All right, thank you very
much for the
23:06:45,9 23:06:49,0 CPT
information, SkyDubai
niner eight one.
So they went around
23:06:49,2 23:06:50,2 CPT
again.
23:06:50,3 23:06:51,2 F/O Again. So we...
It's no, it's no way we
can go in now, man.
23:06:51,3 23:06:55,6 CPT
We’ve got plenty of fuel
nine tons
23:06:55,9 23:06:56,2 F/O Yea.
Ok now it's how to
23:06:56,3 23:06:59,6 CPT manage to go out of this
icing conditions.
The stated dialogue between the PIC and the F/O is the evidence that the PIC was not eager
to make decision to perform another approach or to divert to an alternate aerodrome. Still, the crew
requested and was relayed the weather at the Volgograd aerodrome (URWW). It indicates that the
diversion to an alternate aerodrome was not excluded by the crew and they prepared in advance
for it.
Subsequently the crew contacted the Rostov Approach controller on the possibility to hold
on Sierra Bravo (SAMBEK) NDB or over the BEREG fix, but the Approach controller refused it
due to restrictions. The controller proposed to hold over the KAZAK fix (it is located eastwards
of the aerodrome), but the crew rejected it.
It should be noted that the English language proficiency of the Approach controller (see
Section 1.16.6) did not meet ICAO Level 4 requirements. It posed difficulties for the crew. The
crew was forced to explain their intentions to the controller several times, whereas the responses
by the controller were unclear, once it even required the intervention of another controller to clarify
the content of the ATC instruction. At some point the crew decided to stop trying to explain their
intention to hold to the north of the aerodrome (the PIC to the F/O at 23:13:03: «Leave it. It’s we
do whatever they say once we are not happy we just go and that’s it.».
Finally following the Approach controller guidance the aircraft was brought to holding area
over the MN/Manychskiy NDB to the south-east of the aerodrome. To leave the icing area the
crew requested and was cleared to climb to FL150 (at 23:16:10).
Meanwhile, the AFL 1166 flight performed the third go-around due to windshear with the
speed increment at the altitude of 400 to 300 m. The respective information was relayed by the
controller to the crew.
23:16:22,6 23:16:28,5 APR previous traffic went
around due to wind shear.
Ok. Thank you very much.
23:16:29,3 23:16:34,0 CPT When was the approach
made, what time?
Sky... SkyDubai niner
23:16:40,4 23:16:46,2 APR
eight one. Eh... Just now.
All right thank you very
23:16:46,5 23:16:47,8 CPT
much.
23:16:46,6 23:16:46,9 F/O Just now…
23:16:48,4 23:16:49,1 CPT Still there…
Thus, within 22:00 – 23:16 the crews of different aircraft repeatedly reported moderate-to-
strong windshear. The crew of the AFL 1166 flight performed go-around three times, and diverted
to the alternate aerodrome of Krasnodar afterwards.
Under the current circumstances with the windshear conditions that continued to prevail
for quite a long time, the PIC made the decisions as follows:
 wait for the improvement of weather in the holding area. Particularly given that, there
had been the trend towards its improvement, whereas the quantity of fuel aboard allowed
performing a long-time flight in the holding area;
 under the anticipated conditions of the improved weather (no windshear or the reduction
of its intensity to acceptable values) to perform another approach;
 if it does not work out again, to divert to an alternate aerodrome immediately after go-
around.
Such decisions seem to be logical and correct. In taking them the PIC, most probably,
was driven by the considerations as follows:

1) The improvement of weather conditions that had been monitored by the crew as far back
as the point of go-around allowed hoping for a successful approach after a certain time.
The weather looks better
23:24:09,1 23:24:13,1 CPT here it’s not the same as
before.
23:24:10,7 23:24:13,4 F/O Much better (illeg). Yea.
2) The sufficient quantity of fuel allowed to proceed the holding pattern not less than
2 hours and did not constitute the time pressure.
(illeg) they diverted, they
23:30:52,9 23:30:55,7 CPT
didn't have the fuel.
23:30:56,4 23:30:56,6 F/O Yeah.
And for us now, you know
23:31:04,5 23:31:06,5 CPT
what I'm doing?
23:31:07,1 23:31:07,3 F/O What?
23:31:07,9 23:31:08,8 CPT We are holding.
23:31:09,7 23:31:10,6 CPT We are burning time.
Yeah, we are burning a lot
23:31:11,0 23:31:14,1 F/O of time. Yeah, oh, thank
you. For me?
for your information,
23:31:14,3 23:31:23,6 APR
previous traffic went to
alternate aerodrome.
Oh (no) to divert you
23:31:15,8 23:31:18,8 CPT
mean? It’s no need.
23:31:21,0 23:31:22,5 CPT Eh… I want to fly.
Sorry, say again for the
23:31:24,5 23:31:27,1 F/O SkyDubai niner eight one,
please.
3) The actual weather and the forecast weather at the alternate aerodromes (Volgograd and
Mineralnye Vody) were favorable, which did not involve hasty actions either.
4) The aircraft was in immediate proximity to the Rostov-on-Don destination aerodrome,
with radio and navigation aids set and ready for approach.
Taking the decision to perform another approach the PIC, most probably, took into account
that in case of landing at the Rostov-on-Don airport the loss of time in the holding area
(about 2 hrs) would not be crucial. The passengers will be delivered to destination; it will still be
possible to perform the return flight, as the duty time will not exceed the allowed value. The
potential diversion to the alternate aerodrome would entail many problems: the overnight stop and
the associated significant delay of the flight, the passengers’ accommodation, the breach of the
operational schedule of the aircraft, the other expenses both for the airline and for the crew. These
circumstances could have enhanced the motivation of the PIC to perform landing right at the
destination airport (Rostov-on-Don). The PIC undertook every possible step to complete the main
task, that is to deliver passengers to destination.
It was unreasonable, from the point of view of the PIC, not to take the chance to perform
another approach. Particularly given that to his mind he could have completed even the first
approach if «GO-AROUND, WINDSHEAR AHEAD» warning were not activated and brought
him to make the decision to perform go-around. He mentioned it three times. The first was
immediately after go-around in the conversation with the F/O.
It was ok, man, but it call
22:46:32,7 22:46:35,1 CPT
"wind shear".
22:46:33,8 22:46:34,7 F/O [Sigh].
22:46:35,3 22:46:35,7 F/O Yea.
22:46:36,1 22:46:37,4 CPT It was ok to be honest.
That for me was ok, but
22:46:37,5 22:46:39,6 F/O
they call ''wind shear''.
The second and the third time was in the holding area in a conversation with a cabin
attendant.
Eh, we got a fucking wind
23:01:29,1 23:01:30,9 CPT
shear warning.
Man I could see the
23:01:31,0 23:01:34,7 CPT airfield, and we could go
in, but nevertheless...
23:01:35,1 23:01:36,7 C/A No problem (illeg).
Wind shear warning we
went around, aa.. we are
ah.. in the hold now. We
23:01:35,6 23:01:44,5 CPT
gonna move from this
direction the weather is
coming this side.

So, ah… I expect I don't

know, maybe twenty
23:01:45,6 23:01:51,5 CPT
minutes and then we shoot
another approach.
So, do you think we will

00:07:45,7 00:07:47,2 C/A
get in or not?
00:07:47,6 00:07:48,6 CPT I think we will do.
00:07:48,8 00:07:49,8 C/A We will do?
00:07:50,3 00:07:50,8 CPT Yea.
I think… Before I could
00:07:51,4 00:07:54,6 CPT
see the runway, but...
The machine called wind
00:07:55,9 00:07:59,6 CPT shear, I could control the
aeroplane very well.
But it says: go around,
wind shear, that's it, I
00:08:00,5 00:08:04,6 CPT
couldn't do the approach,
you know.
And it was correct,
because the other
00:08:05,5 00:08:08,9 CPT
aeroplane they go around
as well.
Truly, at the point of activation of the warning «GO-AROUND, WIND SHEAR AHEAD»
the PIC from the altitude of about 1100 ft (335 m) had positive visual contact with the RWY and
with the required accuracy proceeded the flight on the glide path. The current airspeed (160 kt)
exceeded the approach speed, selected by the crew, for 10 knots and actually met the recommended
one as per operational documentation for these conditions. These circumstances (the RWY in sight
and the stabilized position of the aircraft) could have reassured the PIC that if the warning had not
activated he would successfully have completed the landing under the prevailing conditions at that
moment. This fact could have had an important psychological effect as far as the subsequent
development of the situation is concerned.
Note: The investigation team notes that the long discussion by the PIC with the cabin
attendant of the “purely flying matters” (on the windshear warning activation

and the exceedance of IAS with the extended flaps), can be the evidence that the
PIC was affected by them and needed to speak them out.
When approaching the holding area the crew set the speed of 210 kt that is consistent with
the maximum endurance (minimum fuel consumption) for the actual conditions (the weight and
the altitude of flight of the aircraft) and maintained it until flown out the holding area for another
approach.
In the holding area over the MN NDB, the aircraft was flown for about an hour. The
analysis of the crew communication and the actions allows making the conclusions as follows:
 the crew was constantly monitoring the Rostov-on-Don aerodrome actual weather in
attempt to choose the best time for another approach;
 the PIC contacted the airline representative via the SATCOM satellite communication
system and was proceeding the long discussion. In the progress of the discussion the airline
supported the decision of the PIC, recommended the Mineralnye Vody aerodrome (URMM) as the
alternate with more stable weather and even filed and sent the flight plan for an alternate aerodrome
aboard (the estimated distance 270 nm, the flight time 44 min and the required fuel 1728 kg). With
that, the airline recommended that the PIC attempted to land in Rostov that could have been a
stronger motivation for the PIC to perform the task in question:
If we divert you know that
we are divert... diverting
to Minerale Vody. I'm
gonna try this approach if
I can not get in I will not
try another one, because
23:38:30,9 23:38:52,1 CPT they tried many people the
next air... the next airplane
coming in is about one
hour from now. So I will
go once if I don't manage
I will go around and I will
go to Minerale Vody, ok?

Ok, captain that is

understood. We would
like to recommend you to
hold max as possible but
23:38:54,0 23:39:04,5 FD
if you don't want to
approach take one more
approach that is
understood.
Yes hopefully if the
weather passes, because
we have the fuel we
would like to at least try to
land into Rostov instead
of diverting so that would
be our recommendation.
Eh... depends on you
23:39:12,1 23:39:38,1 FD directing weather situation
out there what you see fit.
And then, you can try
another approach if the
weather improves. But
ac... According to the fuel
that we have we will
recommend you to hold as
max as possible.
Ok, perfect, I agree with
you, we will hold then. If
you see the weather that is
23:39:38,4 23:39:53,4 CPT
forecasting to be bad in
Minerale, just give us a
call again, ok?
23:39:44,6 23:39:45,1 FD Ok.
 the crew was thoroughly preparing for a possible diversion to an alternate aerodrome.
Two aerodromes were considered as the alternate - those of Volgograd and Mineralnye Vody. The
crew, having studied the actual weather and the forecast for the immediate future and, as well,

followed the recommendations by airline, selected the Mineralnye Vody aerodrome as the first
option, calculated the remaining fuel, at which the diversion to the alternate should have been
performed, entered the route to the alternate aerodrome into FMS, studied the approach charts and
the other necessary data on the alternate aerodrome;
 the crew calculated Vref and discussed the go-around sequence if windshear warning
were activated again:
So one three four,
autobrake three, flaps
23:54:39,0 23:54:49,1 CPT thirty, wind shear warning
go around do not change
flap or gear configuration.
23:54:49,3 23:54:50,0 F/O Aha.
In case of go-around due to windshear, the PIC, as he did the first time, was going to
perform Windshear Escape Maneuver, that is to go around with the aircraft configuration
unchanged, of which he informed the F/O;
 for quite a long time the crew reviewed the possible options of the subsequent actions
in case of landing at the alternate aerodrome and estimated the amount of duty time:
I don't know, man if we
divert there… we are
23:59:21,4 23:59:27,7 CPT gonna be out of hours, we
are late for five hours,
man.
23:59:26,3 23:59:26,6 F/O Yea.
23:59:28,2 23:59:28,4 F/O Yea.
I see my future is sleeping
23:59:33,8 23:59:36,1 F/O
in the aircraft.
23:59:36,9 23:59:37,2 CPT Oie.
[laugh] And, and... If this
23:59:37,3 23:59:41,8 F/O is it, this place, having the
night there.
00:02:30,5 00:02:30,9 F/O Tired?

00:02:33,3 00:02:33,6 CPT No.
00:02:35,4 00:02:36,5 CPT What are you looking?

For the (illeg) limitations

00:02:36,5 00:02:44,6 F/O is on the... for report time
duty time or whatever.
00:02:45,1 00:02:45,5 CPT Ah…
We eleven hours or
00:02:45,8 00:02:48,6 F/O
something (illeg).
00:02:48,6 00:02:50,3 CPT Just put there FDP.
00:02:50,7 00:02:51,7 F/O Yea. I put it that.
00:02:53,3 00:02:55,1 F/O Put FDP.
00:02:55,4 00:02:56,0 CPT Yea.
Then go OMA chapter
00:02:58,7 00:03:00,9 CPT
seven.
00:03:01,3 00:03:03,7 F/O Yea, it's here (illeg).
Calculation of the flying
00:03:02,6 00:03:06,0 CPT duty periods for flight
crew.
00:03:04,9 00:03:05,1 F/O Yea.
00:03:06,1 00:03:07,0 CPT Acclimatised.
00:03:09,6 00:03:11,9 F/O So, we reported around…
Hm… This time nine
00:03:13,6 00:03:16,0 F/O
forty, yea.
00:03:13,8 00:03:15,5 CPT Nine... nine forty five.
Eleven and fifteen
00:03:18,2 00:03:20,4 F/O
minutes.
I lost the number, the
count number of how
00:03:40,5 00:03:46,2 F/O
many holds, we have done
already (illeg).
00:03:48,1 00:03:49,7 CPT We’ll count it later.
When we enter this hold
we had two hours and
00:03:56,0 00:04:03,2 CPT twenty minutes holding
time, so we are almost one
hour holding.

00:04:00,1 00:04:00,4 F/O Yea.
So how long, before out of

00:08:12,2 00:08:15,3 C/A hours so that you are not
looking into that again?
Yea, of course, we are out
00:08:15,8 00:08:18,1 CPT
of hours, don't worry.
00:08:19,5 00:08:20,9 C/A So what? Nightstop?
00:08:21,4 00:08:23,3 CPT Yea. Most probably, yea.
We are flying now for five
00:08:27,4 00:08:29,6 CPT
and half hours, man.
We know we’ve been
00:08:29,9 00:08:31,8 C/A holding for like an hour
and a half, right?
00:08:31,8 00:08:32,1 CPT Yea.
So, we gonna stay in
00:08:34,3 00:08:36,1 C/A
Minerale Vody?
00:08:36,6 00:08:38,2 C/A Or, I mean in Rostov?
If we manage to land we’ll
00:08:38,3 00:08:44,5 CPT see what the plan it is and
I smell Kuwait.
In this context, the PIC’s decision to proceed a holding pattern for another approach under
the actual circumstances had been the best possible and reasoned decision. This very decision was
coordinated with the airline. Taking into consideration the sufficient quantity of fuel to hold and
divert to an alternate aerodrome, actual weather and favorable forecast as for the alternate
aerodrome, this solution did not pose risks and did not threaten the successful outcome of the
flight. The crew undertook all the necessary steps both for the successful landing at the destination
aerodrome and the diversion to an alternate aerodrome.
However, it should be noted that the landing at the destination aerodrome was the primary
(dominant) goal for the PIC. In talking to the F/O at 00:18:40 the PIC expressed concern that even
in case of diversion to the alternate aerodrome the crew might be commanded to “refuel the aircraft
and fly back to Rostov” («… they will tell us: fill up the airplane and fly back to a … Rostov»).
At 00:20, having evaluated the weather (its shift) once again as per the available
instruments, the PIC decided to perform another approach.

Sixty five knots, a-a-a, I

think it will clear by the
time we do this approach
00:19:52,4 00:20:01,4 CPT
now, even now if we start
the approach. I think it
will be good.
00:20:01,7 00:20:02,0 F/O Will be good.
I think… I would like try
00:20:02,4 00:20:03,6 CPT
now.
00:20:02,9 00:20:04,1 F/O Let's try, let's try.
In addition, the PIC requested the weather from the ATC. The Tower controller at 00:20
relayed the weather as follows: «SkyDubai niner eight one, weather at zero zero two zero: visibility
five kilometers, ceiling six three zero meters, surface wind two three zero degrees one three
maximum one eight meters per second, light shower rain, mist, on final severe turbulence and
moderate wind shear».
From 00:22:17 to 00:22:36 the crew accomplished the DESCENT section of the Checklist.
Another before-landing briefing had been previously accomplished.
At 00:23, having got an update on weather at Rostov, the crew requested descent.
at two two: wind two three
zero degrees, one four
meters per second,
maximum one eight
meters per second,
00:22:43,6 00:23:15,8 APR
visibility six kilometers,
scattered four eight zero,
correction, six three zero
meters. Meteorological
office is not reported about
wind shear on the runway.
00:23:16,5 00:23:17,6 CPT Request descent.

Ok, copied, SkyDubai

niner eight one. Request
00:23:18,4 00:23:25,6 F/O descent for another
approach, SkyDubai niner
eight one.
It should be noted that the latest weather information, relayed to the crew, incorporated the
sentence “Meteorological office is not reported about wind shear on the runway”, which, to
certain extent, contradicted the information on moderate windshear, relayed to the crew two
minutes before. As per item 5.3.16 of Doc 9817 Manual on Low-Level Wind Shear, ATS units
should continue to transmit information on wind shear conditions until it is confirmed, either by
subsequent aircraft reports or by advice from the associated MET office, that conditions are no
longer significant for operations at the aerodrome. In the previous period, the flight crews had not
reported on the windshear presence or absence. Accordingly, the ATC officers had not informed
the meteorological office on the windshear situation. With that, there were no automated means of
windshear detection at the aerodrome. According to the data, present at that moment on the ATC
MeteoDisplay, there had been no information on the actual windshear, as there had been no reports
by the flight crews for more than 30 minutes, confirming the windshear presence. However, at the
same MeteoDisplay the information (warning) had been kept displayed: moderate windshear is
forecasted. The flight crew had not requested the windshear information clarification/update.
At 00:23:35, the crew set the target altitude of 8000 ft and started the descent.
At 00:27, the pilots once again discussed the Vref value and the correction that should have
been introduced, considering the actual wind. The Vref was determined by the crew as 133 kt,
which was consistent with the actual landing weight of 54 t. The fact that the crew introduced the
one knot correction against Vref, that had been determined by them earlier (see the text above for
the communication at 23:54:39), indicates that the crew had been in a normal working condition
and monitored the situation.
The PIC decided that the “wind” correction of Vref should have been equal to +20 kt,
whereas the approach speed Vapp = 153 kt. The latest data on the wind that the crew had been
relayed: wind 230 degrees, 14 m/s, max. 18 m/s. The value of the stable headwind component
amounted to: 14*1.94*cos(230-218) ≈ 26 kt, the exceedance of the gusts value against the
headwind component: (18-14)*1.94 ≈ 8 kt. Thus, the correction, recommended by the operational
documentation was equal to: 26/2+8 = 21 kt. With the consideration of the recommended
maximum value of the correction of 20 kt, the crew had appropriately identified the approach
speed. The crew had “reviewed” the Vfe that had been exceeded during the first go-around, too.

Ah… fifteen meters to this

00:26:54,5 00:26:57,5 CPT
like thirty knots.
(illeg). А thirty knots of a
00:26:56,4 00:26:59,4 F/O
head wind.
00:26:59,8 00:27:00,4 F/O A lot.
00:27:00,6 00:27:01,9 CPT Plus twenty we need.
00:27:01,9 00:27:02,3 F/O Yea.
I just update here, one
00:27:03,8 00:27:05,6 F/O
three three.
One three three plus
00:27:06,0 00:27:08,3 CPT
twenty.
00:27:06,3 00:27:07,2 F/O (illeg).
We will give us one five
00:27:14,0 00:27:15,8 CPT
three.
00:27:15,8 00:27:16,8 F/O One five three, yea.
And the flaps thirty it’s
00:27:16,7 00:27:18,9 CPT
one seventy five.
Yea, we have a lot of
00:27:18,8 00:27:20,8 F/O
margin with that.
At the same time the crew, having in mind the difficulties during the first go-around,
coordinated with the controller that in case of go-around they would immediately climb FL080.
But I will let them know in
case of go around at least
00:27:22,4 00:27:26,9 F/O
coming flight level eight
zero.
00:27:27,2 00:27:27,7 CPT Yea-yea.
I will inform them,
00:27:31,2 00:27:32,7 F/O
already.
Rostov Approach,
00:27:37,2 00:27:39,4 F/O
00:27:40,9 00:27:43,3 APR
go ahead.

Just for your information,

in case of next approach
and in a go around, we are
00:27:44,3 00:27:53,9 F/O
going to request climb
flight level eight zero,
during the go around.
roger, after go around
00:27:56,2 00:28:05,3 APR
climb to flight level eight
zero.
At 00:28:23, the controller issued clearance for descent to FL060. The crew set 6000 ft as
the target altitude.
At 00:29, the HUD switched to an IMC mode.
At 00:30:30 the aircraft was transferred under control of the Rostov Radar controller, to
whom the Co-pilot, at 00:30:41, reported the descent to FL060 with heading 310°. The controller
instructed for the further descent to the altitude of 600 m QFE 988 hPa (2250 ft QNH) and
requested by the crew, relayed QNH of 998 hPa.
This time the crew set QNH correctly. With that the PIC drew attention on the QNH value:
«Before we did the approach with niner niner zero». The F/O agreed to that noting the significant
difference. The fact that the crew noted the difference of pressures (although they did not catch
that as for the first time they were wrong in the setting) indicates the normal working condition
and the monitoring of the situation.
The approach was carried out under weather conditions, similar to those of the first
approach (as per ATIS as of the moment of the accident: moderate turbulence, visibility 7000 m,
scattered cloud 4 octants, cloud base 420 m).
The descent was proceeded with the A/P and A/T engaged. At 00:33:01 at the speed of
210 kt the flap handle was set to 1.
From the altitude of about 3800 ft QNH (1075 m QFE), the aircraft entered the area of
turbulence (the vertical G was alternating within 0.75 – 1.4 g).
At 00:33:41 at the speed of 190 kt the flap handle was set to 5.
At 00:33:48 the controller instructed the crew to change the heading from 310° to 290° and
asked to copy the actual weather, transmitted by the crew of the SVR 2757 aircraft that had just
departed: «… wind on six hundred meters two six zero degrees five three knots and there is also
light icing».

Note: The engineering simulation (section 1.16.7) did not reveal any signs that the
aircraft performance had been affected by the potential icing.
At 00:34:49 the crew introduced correction of QNH following the controller information
on a QNH change of 1 hPa. Concurrently the controller instructed to turn to heading 270°.
At 00:35:29, the Rostov Radar controller issued clearance for an ILS approach to RWY 22
and instructed the crew that they reported established. The crew acknowledged the information
and relayed that in case of go-around they would immediately climb to FL080.
Fig. 62 presents the flight parameters at the second approach. On the plot the parameter
“relative altitude (adjusted for baro correction)” stands for QNH altitude.

Fig. 62. The flight parameters at the second approach

At 00:36:05, the localizer capture is recorded and the aircraft automatic tracking out
initiation on final. At the moment the aircraft was flown in a level flight at the distance of 11.5 nm
(21.3 km) off the runway at the altitude of 2250 ft QNH (it corresponds to the altitude of glide path
interception) (600 m QFE), IAS – 166 kt.
At 00:37:03, the crew requested the wind information from the Rostov Radar controller.
The controller relayed «230 degrees, 12 m/s, gusts 18 m/s». The crew acknowledged the receipt
of information and reported established on localizer. After this communication, the Rostov Radar
controller transferred the crew to the Rostov Tower controller.
At 00:37:25, the F/O contacted the Rostov Tower controller and reported established on
localizer. The controller relayed the actual weather to the crew: «wind two three zero degrees one
two maximum one eight meters per second» and issued clearance to land on RWY 22. The PIC
affirmed cleared to land.
At 00:38:14, the crew had once again corrected QNH in relation to the controller
information of the pressure change by 1 hPa. The capture of the glideslope occurred at 00:38:29
in a level flight at the distance of 7 nm (13 km) off the RWY. At the point of the glideslope capture,
the crew extended landing gear and flaps on 15°. The target go-around altitude was reset to 8000 ft.
The aircraft started descent on glide path. At 00:38:55 at the altitude of 2165 ft QNH
(575 m QFE), the PIC disengaged autopilot and continued to control the aircraft with the use of
HUD. The PIC did not explain the reason for A/P disengage to F/O.
Note: The airline SOP, based on the aircraft manufacturer documentation, do not
require such explanations.
After autopilot disengage, the flaps were extended to 25.
At 00:39:17 at the altitude of 1960 ft QNH (510 m QFE) the AT was disengaged. The flaps
were extended to the landing position of 30. The LANDING section of the Checklist was carried
out afterwards.
At the point of One thousand advisory callout activation (00:40:37) the aircraft was nearly
stabilized for the approach (the flaps at a landing position 30, landing gear down, the deviations
off the beam on localizer and glideslope within tolerance), with that the PIC uttered: «Stabilizing
now», most probably, speaking about speed that was equal to 163 kt (and trended to reduce), which
was 10 kt higher than the approach speed, determined by the crew.
The aircraft was flown a little bit higher of glideslope (0.3…0.2 dots), and the PIC was
applying the corrective “pushing” movements on the control column to maintain the glide path
descent more precisely, along with that the thrust (N1) was increased from 65% to 70%. Over the
same moment the aircraft encountered wind gust. The combination of these three factors resulted
in the IAS, after decreasing to the target value of 153 kt, increase within a second for 15 kt (from

153 to 168 kt), in 2 next seconds it additionally increased up to 176 kt. In such a way the actual
speed exceeded the target one (153 kt) for more than 20 kt. This overspeed was responded by the
F/O at 00:40:49: «Check the speed». It is the overspeed for a considerable value that, most
probably, was the reason for the PIC to make decision on go-around. The PIC took the decision
right away, called it out to the F/O and similarly was responded immediately:
00:40:49,7 00:40:50,4 CPT (Ok), go around.
00:40:50,5 00:40:51,1 F/O Go around.
At 00:40:50 the TO/GA mode was activated with the power levers advanced to full thrust.
The flight parameters at the go-around are given on Fig. 63.

Fig. 63. The flight parameters at the second go-around
Fig. 64 presents the HUD reconstructed image at the point of TO/GA mode activation.

Fig. 64. The HUD reconstructed image at the point of time of 00:40:50 (the TO/GA mode activation)
At the point of go-around initiation the altitude was equal to 1100 ft QNH (250 m QFE)
with the speed of 173…175 kt, that is it was consistent with Vfe for flaps 30. As per the
information available, it cannot be clearly established if the PIC perceived the abrupt increase of
speed readings as one of the signs of the windshear encounter (according to QRH D6-27370-8KN-
JXB page MAN.1.9, the IAS alteration for 15 kt at the altitude of less than 1000 ft is applicable to
“unacceptable flight path deviations’, which in their turn are one of the signs of the windshear
encounter), or as the ordinary sign of an unstabilized approach. It may be assumed that in
advancing thrust levers to full thrust the PIC (“in his mind”) initiated, similar to the first go-around,
the Windshear Escape Maneuver. Particularly given that, if the PIC perceived the speed leap in
question as associated with the windshear encounter, this very maneuver is directly (without an
alternative) prescribed by QRH.
The SOP of some airlines, prior to initiate a range of maneuvers (TCAS, Windshear escape,
Terrain avoidance etc.) recommend that the type of the performed maneuver be spoken out
explicitly (for example, GO-AROUND, Windshear Escape Maneuver), for the other crew member
unambiguously understand what it is going to happen. Particularly if, as in the case in question,
both standard go-around and the specific maneuver were possible. Prior to the first go-around the
windshear was called out with the activation of the respective warning, that is why, most probably,

the Windshear Escape Maneuver was carried out coherently by the crew even without any
additional information.
Before the second go-around there was no windshear warning activation, as well as there
was no explicit callout by the PIC on the type of maneuver (the airline OM did not stipulate such
a specification of commands), that is why the F/O could have perceived the maneuver to come as
the standard go-around – particularly given that he had just drawn the PIC’s attention to the
deviation (on speed) of the stabilized approach criteria. It is the F/O who offered to PIC to retract
flaps to 15, as it is prescribed at the standard go-around. The PIC immediately agreed with the
F/O, although according to SOP (item 12.1.1), the decision-making and the respective callout to
retract flaps should come from the PIC himself with a simultaneous GO-AROUND callout.
00:40:51,1 00:40:51,6 F/O Flaps fifteen?
00:40:51,6 00:40:52,2 CPT Flaps fifteen.
Thus, with the consideration of the subsequent landing gear retraction following the F/O’s
report on positive rate of climb, actually the crew was performing a standard go-around procedure
at the low weight aircraft with the maximum thrust of the engines (N1=101%), as it is prescribed
by the Windshear Escape Maneuver. Taking into account that in preparation for the second
approach it is this procedure that had been talked over by the crew, the indicated factor may have
played a certain role, as later the whole range of the erroneous and inappropriate actions by the
PIC resulted in a loss of control of the aircraft. This very moment, most probably, had been a
turning point in the chain of the events.
At 00:40:54, the F/O reported go-around to the Rostov Tower controller and was instructed
to contact the Rostov Radar controller. The crew did not contact the Radar controller.
The increase of the engines thrust combined with the retraction of flaps and landing gear
led to the significant increase of pitch-up moment (more than at the first go-around); to counteract
it the substantial push of the control column was required with the application of pushing forces
on the control column of up to 50 lb (23 kg), that had been preserved for quite a long time (more
than 40 sec). The piloting - especially the precise piloting - of an out of trim aircraft (unbalanced
on forces), is always complicated and implies the increase of the pilot’s workload, including the
psychoemotional component. Indeed, the nature of the control wheel motion on pitch after the
initiation of go-around is loose (the longer movements) and abrupt (reactive) with the noticeable
delay. This very nature of the piloting is always the evidence of the flight mental mode disruption
and the absence of forecast on the further behavior of the aircraft (the pilot “is behind” the aircraft,
the aircraft flies “ahead” of the pilot).

Note: 1. It should also be taken into account that the stated forces on the control
column PIC, most probably, had maintained by one (left) hand. The other hand
was on power levers (the confirmation to this is stated here below).
2. Hereinafter the integral “flight mental mode” means not only the pilot’s
reflection of the flight process, that is matching of real aircraft attitude and mode
to the perceived ones, but also the ability of the pilot to fly “ahead of the
aircraft”, including to predict the aircraft behavior in response to the control
inputs.
3. As per the information, submitted by the airline, the PIC was well experienced
in performing go-arounds in real flights (6 go-arounds not counting the first go-
around at the day of the accident, of which 3 as a PF and 3 as a PM). The airline
could not provide the information on the reasons for go-arounds. However,
based on the data that all go-arounds having been performed at N1 less than
100 %, it could be supposed that the windshear escape maneuvers had not been
applied.
As the result all the further actions in the rapidly changing environment, that was evolving
into non-standard (non-trained) situation, were of a belated character. The analysis of the angle of
climb (pitch) value shows evidently that the PF (the PIC) had been never able to create the initial
pitch of 15º, recommended by FCOM. The actual pitch was changing in steps (6º, 12º, 14º, 9º, 14º,
18.5º, 4º). As oppose to the first go-around, the PIC failed to maintain the climb path, set on by
the pitch command bar, the position of which on PFD at the same points of time was equal to30
(4º, 3.5º, 2.5º, 7º, 3º, 1º, 13º). The HUD reconstructed image as per the points of time 00:41:03
(the pitch local maximum 14), 00:41:07 (the pitch command bar local maximum 7), 00:41:13
(the maximum reached pitch 18.5), 00:41:19 (the pitch command bar local maximum 13) is
given on the below figures (Fig. 65., Fig. 66, Fig. 67, Fig. 68).
30
These values are measured from the present pitch of the aircraft.
Fig. 65. The HUD reconstructed image at the point of time of 00:41:03 (the pitch local maximum ≈14°)
Fig. 66. The HUD reconstructed image at the point of time of 00:41:07 (the pitch command bar local
maximum ≈7°)

Fig. 67. The HUD reconstructed image at the point of time of 00:41:13 (the pitch local maximum ≈18.5°)
Fig. 68. The HUD reconstructed image at the point of time of 00:41:19 (the pitch command bar local
maximum ≈13°)
The average pitch in climb was equal to about 10º, which was obviously insufficient for
the given conditions, taking into account that:
 the thrust-to-weight ratio of the aircraft was high (the aircraft was of a relatively low
weight of 54 t. and the engines were delivering the maximum possible thrust);
 the drag of the aircraft was significantly less against the first go-around (landing gear
up, flaps 15º).
The PIC, apparently, failed to consider all these factors. At the same time the F/O,
observing the PIC, having difficulties in piloting the aircraft drew his attention to the necessity to
maintain the required pitch.
Keep it to fifteen degrees,
00:41:09,7 00:41:12,0 F/O
nose up.
The PIC briefly (for a second) increased pitch up to 18.5º, then reduced it again to 4º...5º,
having applied the disproportionately significant pushing movements on the control column (more
than ½ of travel from the neutral position to fully forward) and generated the vertical G of 0.43 g.
In reaction to such actions by the PIC, the F/O prompted once again:
Now keep it, keep it to
00:41:18,0 00:41:19,4 F/O
fifteen, keep it…
Note: At this moment, there was a minor (down to 97 % N1) decrease in engines power
with the subsequent restoration of the maximum thrust. The investigation team
is of the opinion that this very decrease was incidental and is associated with
the PIC’s hand on power levers at the above-indicated significant change of the
vertical G.
Having focused the attention on maintaining pitch, both pilots had completely lost the
awareness of on the airspeed, that was gradually increasing and at 00:41:10 reached the limit Vfe
(200 kt) for flaps 15º, after which the T.E. FLAP LOAD RELIEF was activated. The flaps were
set to 10º. In such a way, it was a repetition of the situation that occurred during first go-around.
With that, the F/O, as opposed to the first go-around, did not draw the PIC’s attention on the speed
exceedance. The engines kept on running at the maximum thrust up to the impact with ground; the
flaps handle remained at the 15º detent.
The situation when the attention of the crew was excessively concentrated on the pitch
control of the aircraft led to the substantial narrowing of perception of the other information (“the
tunnel effect”). The number of parameters, having been monitored and analyzed at a time, had
decreased sharply, down to about one or two.

The onset of the critical situation started at 00:41:30.5. In proceeding the piloting of the
out of trim aircraft (unbalanced on forces), and obviously enduring inconvenience; at the present
parameters of flight: Н ≈ 3350 ft QNH (935 m QFE), IAS – 210 kt, pitch – 10º to nose-up, the
stabilizer setting – 2.7º to nose-up, flaps 10º, the power mode of the engines – the maximum thrust,
the PIC pressed the stabilizer trim switches to nose-down. Fig. 69 presents the HUD reconstructed
image at that point of time.
Fig. 69. The HUD reconstructed image at the point of time of 00:41:30.5 (the initiation of the stabilizer trim to
nose-down)
The figure is the evidence that the system “commanded” the pitch increase (the dashed line
– TO/GA pitch target line – is higher than the aircraft reference symbol). In other words, most
probably the PIC, having acted in such a manner, tried “to relieve” the pushing forces on the
control column, rather than to transition the aircraft to nose down. However, along the stabilizer
trim, the pilot did not return the control column to neutral position, with that the stabilizer trim
switch remained pressed for an abnormally long time (12 s!!!).
Note: 1. The investigation team did not identify any signs of failure or the
uncommanded movement of the stabilizer. The direction and duration of the
stabilizer trim are consistent with the respective discrete signals registration (the
FDR records the discrete signals, associated with the commands to the nose-
down and nose-up stabilizer trim in a manual or automatic mode). The short-

time pull of the control column (at 00:41:39) beyond neutral resulted in the
stabilizer trim stop that is consistent with the inherent logic of the control system
function.
2. Most likely, by this moment of time the flight mental mode at the PIC had
been completely disrupted, whereas his emotional condition may be
characterized as “distress/a very strong stress”. It is confirmed by the fact that
from 00:41:33 and up to the end of the flight the quite dynamic pedal inputs are
recorded, with that they had not been anyhow justified by the flight situation. It
is known from the practice of the air accident investigation that when the pilots
encounter complete spatial disorientation and lose the ability to appropriately
evaluate the situation, often they start to do the involuntary (reflex) control
inputs. In this sense, as for the modern transport aircraft, the deflection of
pedals that, usually, are not used in flight, is the most indicative symptom. After
the forward deflection of the left pedal within a time of 00:41:33 – 00:41:39,
the aircraft bank was changing with a constant angular rate from 5 to the right
to 12 to the left with the practically neutral position of the control wheel that
again demonstrates the crew failure to monitor all the necessary parameters.
The actions in question (the substantial stabilizer trim to nose-down along with the keeping
and even some increase of the average deflection of the control column to nose-down) led to the
rapid decrease of pitch (the average pitch rate about 6 /sec) and the aircraft transition from the
climb to descent with the significant negative G (initially minus 0.3…minus 0.4 g, then – up to
minus 1.07 g), at that the near-zero and negative G (<+0.2 g) were present for a long time.
Note: FCT 737 NG (TM), page 7.22:
It may be difficult to know how much stabilizer trim to use, and care must be
taken to avoid using too much trim. Pilots should not fly the airplane using
stabilizer trim, and should stop trimming nose down when they feel the g force
on the airplane lessen or the required elevator force lessen.
The F/O, operating as PM, realized that the hazardous situation was emerging. Desperately,
with the increasing anxiety in voice, he tried “to return” the PIC to a control loop and rectify the
situation. From the point of time of 00:41:34 the F/O prompted the appropriate actions to the PF
(the control column pull) to prevent the situation transition to the accident. Nevertheless, most
likely, the PIC had no longer heard the F/O and had not reacted to his words – he had fully lost the
control of the situation, as well as the capability to control the aircraft.

00:41:33,8 00:41:35,4 F/O Yea. Be careful now.

00:41:35,4 00:41:36,4 CPT Oh, shit!
00:41:35,6 00:41:36,6 F/O Be careful. Be careful!
Sound effect of rising or
00:41:36,2 00:41:45,9 falling items sound in
flight deck.
00:41:36,8 00:41:37,3 F/O Be careful!
00:41:37,5 00:41:38,2 F/O No, no, no, no, no, no!
00:41:38,3 00:41:38,5 F/O No!
00:41:38,5 00:41:39,2 F/O Don't! Don't do do that.
00:41:39,9 00:41:40,4 F/O Don't do that.
00:41:40,7 00:41:41,6 F/O No! Pull it! Pull it!
00:41:42,1 00:41:42,5 F/O Pull it!
00:41:42,7 00:41:43,5 F/O My God!
The inappropriate actions by the PIC within 12 sec, when the stabilizer trim switches were
kept pressed, the aircraft entered into nose-down upset (negative pitch of about 40º, IAS – 280 kt
with the flaps 10º, the altitude of 2800 ft QNH (770 m QFE), the engines power at maximum
thrust, the stabilizer setting of 2.4º to nose-down).
Within the time interval of 00:41:42-43 it had been a momentary aft control column
deflection by ≈11 (more than 2/3 travel off neutral) with the subsequent return beyond neutral in
the forward direction. The calculations were the evidence (see Section 1.16.7), that if at that point
of time the control column were repositioned to a full pull and was kept at that position, the aircraft
could have been recovered from descent with a sufficient margin of height. Only aircraft
aerodynamic performance and capabilities are analyzed here. The crew members’ state and actions
are analyzed below and have not been taken into consideration in the engineering simulation.
Similar results were obtained as a result of the experiment at the B737-800 aircraft
simulator. The appropriate actions by the crew (the pull of the control column creating vertical G
of 2.2…2.4 with the simultaneous reduction of the thrust to idle) allowed recovering the aircraft
safely.
From the point of time of 00:41:44, the F/O attempted to pull the control column (see also
Section 1.16.8).
The last seconds of the FDR record explicitly confirms the complete spatial disorientation
of the PIC and indicate the critical disruption of the flight mental mode integrity. Apart the lack of
appropriate actions on the aircraft recovery from descent (even after the EGPWS PULL UP

warning activation) the control wheel was repositioned to practically full left in roll that led to the
intensive development of the left roll. The aircraft impacted ground with the significant speed
(≈ 340 kt), high negative pitch (≈ 50º) and the left bank ≈ 60º.
Note: The analysis showed that, most probably, even with the appropriate actions by
the pilot at the EGPWS PULL UP warning activation at that point it had not
been possible to prevent the catastrophic situation (to recover the aircraft out of
descent).
Honeywell, being the EGPWS manufacturer, upon the investigation team
request commented that the accident flight actual parameters (rate of descent
18000 ft/min (91 m/s)) had been far beyond than the maximum values
(7000 ft/min (36 m/s)), determined by the TSO C151b and DO-161A documents,
in accordance to which the system had been designed.
At the same time at that phase of flight the pilots could not have known that it
had been already impossible to recover the aircraft out of descent, for this
reason no appropriate actions in response to this warning demonstrates indeed
the critical disruption of the flight mental mode.
In the progress of the analysis of the crew proficiency and their in-flight actions the
question arises – quite naturally – how the experienced and disciplined crew, having been in a
normal working condition and monitored the flight, all of a sudden had lost control and let the
situation occur that resulted in the fatal accident?
All the actions and taken decisions (besides the setting of QFE instead of QNH during first
approach) up to the moment of initiation of the second go-around testify to a high level of training
and proficiency of the crew, and the PIC had been an indisputable leader. The point of initiation
of go-around during the second approach can be considered as the onset of the abnormal situation.
From that moment, the PIC started to make errors, take inappropriate actions and was no longer a
leader. After the Gear Up callout (00:41:00) and up to the end of the record the PIC did not give a
single command. Within this time interval, he uttered the total of four remarks:
00:41:12,2 00:41:13,5 CPT Checked [Exertion breath]
00:41:16,4 00:41:17,6 CPT Aaak [Exertion breath]
00:41:22,4 00:41:23,1 CPT Don't worry. Don't worry.
00:41:35,4 00:41:36,4 CPT Oh, shit!
These remarks were not commands. They are just another proof of the non-optimal
psychoemotional state of the PIC.
Why in a short time (several seconds) after taking the decision to go-around did the PIC
get into such a condition? As it was already stated above, in case of an unsuccessful second
approach due to windshear, the crew were prepared to perform Windshear Escape Maneuver and
immediately divert to an alternate aerodrome. However, the landing at the destination aerodrome
had been the predominant goal for the PIC. Just to achieve the goal two hours were spent in the
holding area. The PIC had been certain that he would manage landing under actual circumstances,
if the windshear warning did not activate. Most probably, it was the activation of the warning in
question that should have been that internal trigger for the PIC, which would have altered the
action plan – to perform go-around instead of landing. Instead of the activation of the windshear
warning the speed leap caused the decision to go around which under the conditions of turbulence
and gusty wind had not been as much clear sign from the PIC’s point of view comparing to the
warning activation (taking into account the PIC confidence in his capability to perform landing).
Most probably, psychologically, the PIC had never got over the impossibility to perform
landing at the destination aerodrome and accepted the need to divert to an alternate aerodrome.
Particularly given that after the first go-around he had been sure in mind that he would have
managed landing. The confidence in question at certain time kept haunting his mind and he spoke
it out even in conversation with the cabin attendant.
As the result, probably, the PIC had been stuck in “a clinch” of two opposite goals
(motives): to proceed approach for landing or initiate go-around. Even though the PIC, in
compliance with SOP, took the immediate decision to perform go-around, “the clinch” resulted in
the disruption of the previous flight mental mode (the approach with landing), whereas the new
one (the go-around) had not been formed yet. The failure to follow the Windshear Escape
Maneuver, that had been supposed to be carried out, and the loss of the initiative (the consent to
the F/O’s offer to retract flaps to 15º, that is to perform standard go-around) is the confirmation to
this. As a result, the PIC, whose actions did not allow him achieving the goal, had been in
“a mixed- up” condition and lost the ability to predict the further behavior of the aircraft (he was
“behind” the aircraft). This led to the loss of the situational awareness and psychological
incapacitation.
Thus, the erroneous actions by the PIC, when he had not been able to maintain the required
climb path for a long time with the application of the significant “pushing” forces (though to get
to the “correct” path he needed to relieve some forces on the control column), first of all are related
to his mindset as the result of the simultaneous manifest of two opposite motives at the point of
taking decision to perform go-around. Similar casual factors were revealed at the investigations of
the other fatal accidents that occurred in the progress of go-around, among them the air accident
to the B737-500 VQ-BBN aircraft, operated by Tatarstan airline, on 17.11.2013 at the Kazan
airport and to the A320 EK-32009 aircraft, operated by the Armavia airline on 03.05.2006 at the
area of the Sochi airport.

The F/O, for the most part, appropriately assessed the attitude of the aircraft and prompted
the correct actions to the PIC. The Flydubai airline OM, as the OM of the majority of other airlines,
states the actions of the PM in case of the physical incapacitation of the PF. The actions in question
are recurrently practiced on simulators. As a criterion of incapacitation (in addition to the obvious
symptoms) for the altitudes of more than 1000 ft) the airline OM states the following:
 no verbal reply to two communications by the PF addressed by the PM;
 no verbal reply to the PM’s remarks on the significant deviations of the intended flight
path;
 no reaction to the system malfunction warning.
In the case under review, the PIC verbally (though in a one word) responded to the prompts
by the F/O and there were no system failures occurred.
The events of the psychological incapacitation, when the PF verbally responds to the
communication and physically exercises the control inputs, but these are obviously inconsistent
with the current flight environment, as a rule, are not incorporated into the airlines OM. The
Flydubai airline OM did not state such occurrences either. Consequently, the PM for more than
30 s had been prompting the correct actions to PF and at a certain point even tried to intervene in
the control with prompting the correct direction of the control column deflection “by the action”.
Still, because of the continued inappropriate actions by the PIC and the specific features of the
pitch control system (Sections 1.16.7 and 1.16.8), the actions by the F/O had been unsuccessful.
Note: On the results of one of the simulator sessions, the flight instructor wrote down
a comment to the F/O “Needs to be quite a bit more assertive in what is needed
from the Captain. Tell him/her what you want done and do not wait for the
Captain to enquire with you or direct you in this regard. Need to be more
decisive in taking actions when needed.” Maybe more decisive actions by the
F/O in intervention on the control of the aircraft could have prevented the
accident, still he started to act in a formal way (as fixed by the OM provisions)
no sooner than against the inappropriate actions by the PIC at the EGPWS Pull
Up alert activation.
A major factor that led to the PIC’s spatial disorientation and his inability to take
appropriate actions was, probably, the G-force that, as the result of his control inputs, reached the
near-zero and negative values. The practice of flights with the near-zero and negative G is the
evidence that the pilots, who encounter zero G for the first time (even if fastened by the seat belts)
during the first seconds are not just incapacitated, but spatially disoriented as well. In addition at
the near-zero and negative Gs together with the unsecured objects the mud and dust, always present
in the cockpit, float up in the air. This occurs, generally, all of a sudden, with a startle effect to the

crew. Likewise, the mud and dust, as a rule, penetrate into the eyes and nose, degrading the vision
and breath. The investigation team has no data available that either of the crewmembers up until
that time had the practical knowledge of the zero-gravity condition, not to mention the respective
exercise.
The findings on the crew’s actions at the final phase of the flight
In progress of the second approach up to the point of go-around, the crew’s actions were
appropriate and complied with SOP.
The onset of the abnormal situation was at the point of initiation of go-around and, most
probably, had resulted from the inherent lack of psychological readiness of the PIC to perform go-
around that led to the loss of his leadership in the crew, the disruption of the flight mental mode
and piloting errors, which within a short time were transformed into the inappropriate actions,
resulted in the loss of control of the aircraft. The key point for the transition of abnormal situation
into the accident (emergency) one was the fact that for a long time (12 sec) the PIC kept holding
the trim switches (the stabilizer control) to nose down. With that, the stabilizer trim rate with the
extended flaps is about twice the trim rate with the retracted flaps. This resulted in the significant
pitch imbalance and, combined with the keeping of the forward deflection of the control column,
inconsistent with the actual situation, the aircraft encounter near-zero and negative Gs with the
associated spatial disorientation of the PIC. The situation itself was consistent with the
psychological inability of the PIC to control the aircraft (the pilot incapacitation); however, the
actions of the PM under the circumstances in question are not described in the airline OM.
The F/O, for the most part, was appropriately assessing the attitude of the aircraft, prompted
the correct actions to the PIC and even attempted to intervene in the control of the aircraft. Still,
the essentially correct attempts of the F/O to prompt the PIC from the hazardous situation had been
unsuccessful.
The “operational” tiredness can be attributed to the factors that could have had the negative
impact on the PIC’s condition and actions – by the moment of the accident the crew had been
proceeding the flight for 6 hours, out of which 2 hours under intense workload, related to the
performance of go-around and the necessity to take non-standard decisions; in this context the fatal
accident occurred at 04:42 as per the departure airport (Dubai) time zone – the worst possible time
from the point of view of the circadian rhythms, when the human performance is severely degraded
and is at its lower level along with the increase of the risk of errors.
2.2. On the peculiarities of the trim (relief) of forces

As it was mentioned above, the long-time press of the trim switches (the stabilizer control)
to nose down by the PIC had been a factor that contributed to the transition of the abnormal

situation into the accident (emergency) one. Probably, by this action the PIC tried to relieve the
“pushing” forces on the control column that had been preserved by him for a long time.
The presence of forces on flight controls (for example on the control column), which, in
general, increase along with the increase of their deflection, is required by the modern
airworthiness standards. With that the systems that trim (relieve) these forces are introduced into
design. There exist two fundamentally different concepts of the forces generation and relief on the
flight controls.
The first concept is nowadays integrated, mainly, to the light and ultralight aircraft, on
which the control column feel is achieved with the direct forces from the aerodynamic control
surfaces (for example from the elevator). In this case at the deflection of the elevator (the control
column) the pilot feels a feedback on changing forces (due to the change of the hinge moment on
the control surface), which is caused by the change of the pattern of the elevator flow by the relative
wind. To trim (relieve) the forces (to reduce the hinge moment) for the case the specific auxiliary
aerodynamic surfaces (trim tabs) are provided, mounted on the trailing edge of the aerodynamic
control surfaces. In deflection to the opposite direction against the deflection of the control surface
with the larger arm they relieve forces on the control surface and, consequently, on the flight
control. As for this concept, if a pilot moved the control surface to a certain position to create the
control moment and wants to preserve this moment, but to relieve forces, he should, with no change
of the flight control (control surface) position, deflect the aerodynamic trim tab to the
corresponding (opposite) direction. As the result the forces on the flight control will be relieved,
whereas the flight control itself and the control surface will remain in a required position. The
aircraft with that type of pitch control usually have fixed or adjustable (repositioned on a particular
setting) stabilizers. That means, generally speaking, that the pitch balance and control of the
aircraft are ensured by the elevator.
As for the modern commercial airliners, on which the PCUs are integrated to deflect the
control surfaces, it is not the case of the direct transfer of forces to the flight controls. The aircraft
of the type can still integrate the trim tabs, but these relieve the forces on the PCUs and are
controlled automatically, not by the pilot. With that, the control wheel loading is achieved
synthetically with the specific feel and centering unit. Generally, as for these aircraft types, the
trimmable stabilizers are mounted that may be deflected to any position (within the range of
travel), both commanded by pilot and automatically. It is this concept that was implemented on
the Boeing 737-8KN A6-FDN aircraft (and on all the aircraft of the Boeing 737-800 type). At the
same time, the aircraft has no function of trim (relief) of the forces directly on the control column.
To relieve forces on the control column it should be returned to neutral position, with that the
elevator deflection with the associated control moment is changed in proportion.

As for this very concept if a pilot created a certain control moment with the elevator and
wants to preserve it with the simultaneous relief of the forces on the control column, he should
initiate the stabilizer movement in direction along the deflection of the elevator. With the stabilizer
deflection, a pilot returns the control column to neutral, relieving the forces. Concurrently the
deflection of the elevator is reduced. At the end of the cycle the required control moment is created
by the stabilizer, whereas the control column is in neutral position. The concept implies the pitch
balance of the aircraft, ensured with the stabilizer, the control is achieved with the elevator. As
stated above, FCT 737 NG (TM) page 7.22 reads that the pilots should not fly the airplane using
stabilizer trim.
The vast majority of pilots start their career on the airplanes of the first type. At the
transition training to the aircraft of the second type the specific features in question should be
explained to them.
Upon the request of the investigation team the aircraft manufacturer, the Boeing Company,
responded that the Boeing 737 aircraft documentation does not contain the specific guidance on
the general principles of the forces trim31. Boeing is of the opinion that the indicated skills are
integral of the basic airmanship to perform flights on large transport aircraft. At the same time the
manufacturer notes that this documentation is designed based on the assumption that the customers
have had the previous flying experience on the jet multi-engine aircraft and are familiar with the
basic systems of the jet aircraft and basic flight maneuvers, common for the aircraft of the type. In
relation to this FCTM does not incorporate the background information, of which the awareness
is considered as prerequisite to familiarize with the concerned document.
Note: At the same time, at the movable stabilizer introduction into service Boeing has
issued the detailed guidance, explaining the general principles of the use of such
systems through the Boeing 707 and 720 aircraft. These materials can be found
in the Boeing Airliner magazines of April, 1959 and May, 1961 on the
myboeingfleet.com website.
As it is indicated in Section 1.5.1, the Boeing 737 type was, most probably, the first jet
multi-engine aircraft in the PIC’s career. Prior to that, the PIC had operated light aircraft. The
investigation did not manage to find the data on these airplane types. It had not been possible to
determine, whether the PIC had been ever informed of the peculiarities of the forces trim at the
transition training.
31
In this case, the general principles of forces trimming on large transport airplanes with movable stabilizer are
discussed, and not the operation of the system mounted on the specific aircraft type (Boeing 737-800).
The investigation team notes that throughout standard operation the Boeing 737 pilots
rarely run into the necessity to trim the stabilizer manually by a large amount as the aircraft balance
is exercised by the automatic systems of the aircraft.
As for the accident flight, both in progress of the first and the second go-around the PIC,
when pressing the trim switches, did not return the control column to neutral, thereby increasing
the pitch-down moment. In the course of the meetings with the representatives of the command,
flight crew and instructor personnel of the airline the view has been expressed by them that the
PIC could not have possessed the sufficient level of knowledge on the peculiarities of the forces
trim on the aircraft of the Boeing 737 type.
The investigation team also notes that in case of long-time press on the stabilizer trim
switches there is a danger of “loss” of the feedback on forces. Due to the small amount of forces
the pilot’s finger gets used to them quickly, which may lead to the “prolongation” of the earlier
started action, that is to the “holding” of the trim switches. The non-optimal working condition of
the pilot or the situation, when the pilot does not achieve the required goal (the forces relief on the
control column) with the mentioned action, can be contributive in the case.
IAC had come across the alike phenomenon in the previous investigations (see for example
the results of the investigation of the fatal accident to the Ilyushin Il-86 RA-86060 on 28.07.2002
at the Sheremetyevo airport, Section 1.18.1). To prevent such occurrences some aircraft types
integrate the discrete control of the stabilizer – one press implies the stabilizer trim to certain
amount. To trim further the trim switch is to be released, and then pressed again.
The potential risk of such a situation is also in the fact that the signs, indicative of the
stabilizer moving (resetting), (in the case of Boeing 737 it is the rotating trim wheel, enabling the
back-up control of the stabilizer), themselves (without the analysis of their duration and monitoring
of the stabilizer actual position) are not perceived by the pilots as alarming, as they are constantly
manifested throughout normal operation. According to SOP no pilot is obliged to monitor the in-
flight stabilizer setting on the indicator (it is located beneath on the central console next to the
back-up control wheel and is out of the line of sight of the pilots), with that the event of pressing
the trim switches by the finger may not be noticed by PM.
The PM has the opportunity to stop the stabilizer trim by pressing the trim switches in the
opposite direction (as for the situation of the accident flight with the control column, for a long
time had been kept by the PF at the nose-down deflection, for the PM to use “his own” stabilizer
trim switches it had been necessary to use the stabilizer override switch (see also Section 1.18.2).
2.3. On the use of HUD in progress of go -around

As it was stated above, according to the airline SOP (Supplement D, Section D.1.2) the use
of HUD, if it is operative, is mandatory throughout the entire flight.

FCT 737 NG (TM), page 1.42 reads that no restrictions are imposed to the use of HUD. At
the same time, the document does not contain the description of any additional actions by the
crewmembers as for the use of HUD in progress of go-around.
The airline SOP incorporates the information on the HUD automatic switch to the PRI
mode following the press of TO/GA button. The crew is also prescribed to adjust the Aircraft
Reference Symbol to the dashed Target Pitch Line, upon which follow the Guidance Cue.
Both FCTM and SOP encourage the use of HUD at any time (on any leg of the flight),
since it facilitates the monitoring of the aircraft flight parameters with the concurrent out-of-the-
window view. Actually, it is one of the main advantages of the use of HUD in the observation of
the outside environment with the preservation of monitoring of the primary flight parameters.
At the same time, at certain flight legs, for example, at the go-around or at the aircraft upset
recovery, especially if it is the occurrence in IMC and at nighttime, the advantage in question is
not relevant.
Indeed, under certain circumstances, such as inappropriate adjustment of the unit
brightness against specific lighting environment, the excessive focus of the pilot on the outside
environment at the potential different visual effect in cloud or fog, may contribute to the degraded
situational awareness and/or spatial orientation.
One more specific feature of the use of HUD is that the pilot only at his certain relative
head posture against HUD can observe the “complete” picture. At the considerable change of the
pilot’s head posture against HUD, for example, at the extreme maneuvers, severe turbulence, upset
encounter (notably with the reach of near-zero and negative Gs) the portion of the HUD picture
may be lost out of the pilot’s vision field. Following the investigation team request the HUD
manufacturer responded that at its certification there were no flight assessment of the unit carried
out into the entire range of the operational G (up to -1 g) for the Boeing 737 aircraft (since it is not
stipulated by the certification requirements) and the respective test pilot evaluation does not exist
either.
Should the aircraft encounter irregular attitude or upset the HUD indication changes. The
figures below (Fig. 70, Fig. 71, Fig. 72) represent the HUD reconstructed image at approximately
the time of 00:41:36, 00:41:40 and 00:41:43.


At Fig. 70, at the actual nose-down pitch of about 10, the upward pointing Pitch Chevron
is displayed on HUD (the direction of the aircraft recovery to the “regular” flight) with the top
point at the pitch value of 20 nose-down, at the same time the HUD is still operated at the PRI
mode. The Pitch Chevron appears when the HUD is switched to the compressed mode. The Pitch
Scale is compressed, when the aircraft attitude is such that the Horizon Line or the Flight Path
Symbol cannot be displayed conformably. In the compressed mode, the increment of the pitch
scale is doubled, from 5 to 10. The figure is the evidence that in this mode the proportion between
the position of Guidance Cue and Flight Path Symbol, the target and actual pitch symbols is no
longer respected. At that the Flight Path Symbol is displayed in dashed line – becomes “ghosted”
(the term, used in the HUD Pilot Guide) – this means it is no longer conformal with the real world,
though being in a qualitatively correct position against the Guidance Cue.
The respective information is stated in the HGS Pilot Guide. All the while, it is doubtful
that the PIC in the progress of the simulator sessions, with the standard go-arounds performed,
could have often observed the similar change of indication. In the case, the HUD picture becomes
as if “cluttered”, which, taking into account the increased stress of the pilot, may lead to the wrong
interpretation of readings.
One more specific feature of the display consists in the Pitch Scale Lines, which in the
positive region (to nose-up) are displayed solid-line, whereas in the negative region (to nose-down)
are dashed. The Pitch Target Line at the go-around is dashed as well. At the standard go-around,

the dashed Pitch Target Line contrasts well with the solid lines of scale and Horizon. In the
situation of the accident flight, the dashed Pitch Target Line was displayed next to the dashed lines
of the scale. Although the lines of scale are visually different (they integrate additional side
“notches”, pointing the direction of recovery to the “zero pitch”, and digital values, the dashed line
is ruptured in the middle and is of a smaller length), as for the stress situation, the pilot with the
trained skills to match aircraft reference symbol to “dashes” (see the SOP guidance above), may
misinterpret the actual indication.
Fig. 71 and Fig. 72 represent the HUD reconstructed images at the points of time, coherent
with the maximum reached negative G (-1.07) and the moment of the F/O’s interference into the
control. It is apparent that HUD switched to the display mode, corresponding to upset. The HUD
is switched to this mode, when the nose-down pitch exceeds 20. According to the HGS Pilot
Guide, this mode enhances interpretation of an unusual attitude by the pilot. All the “redundant”
information is retracted out from the HUD screen. These are only altitude and IAS scales, pitch
and roll together with the slip indicator that are displayed.
In the progress of the investigation team activities team some pilots noted that in the stress
environment the applied indication (two nonparallel lines) might be misinterpreted by the pilot, as
the indication, similar in appearance, is used to display RWY at the HUD operation in the AIII
mode.
Thus, taking into account the lack of the objective information on the HUD operation (there
were no flight tests of the unit carried out in the entire range of the operational G; the impossibility
to reproduce the real HUD readings in the progress of the accident flight, that is the image, the
pilot was really watching with the consideration of his posture in the seat, through the stream video
(see Section 1.16.9) or at the FFS) did not allow making unambiguous conclusion on its possible
impact on the flight outcome.
At the same time the investigation team believes that the further elaboration of the
methodological issues of the practicability and the use of the HUD at different segments of flight,
as well as the carry out of the additional flight tests into the entire operational range and the works
on the improvement of the data presentation (with the consideration of the opinion of the experts
in ergonomics and aviation psychology) may mitigate the risks of the erroneous interpretation of
readings in the stress situation.
2.4. On the possible impact of somatogravic illusions

Somatogravic illusions have been noted repeatedly as a contributing factor during
investigations of aviation accidents and incidents which involved go-arounds32.
32
See the BEA Study on Aeroplane State Awareness during Go-Around /ASAGA for details
(http://www.bea.aero/etudes/asaga/asaga.php).
Note: Somatogravic illusion is the common form of vestibular illusion or “false

sensation”. Somatogravic illusion may lead to spatial disorientation. The
significant positive longitudinal acceleration of the aircraft may contribute to
the nose-up pitch illusion. If it is the case, the pilot in a horizontal flight may
instinctively deflect the control column forward, tending to avoid the increase of
the “perceived” pitch. The rapid deceleration of the aircraft results in the
opposite effect: the pitch-down illusion is induced and the pilot may
inappropriately pull the control column, thereby increasing pitch.
The investigation team analyzed the possible impact of the somatogravic illusions on the
outcome of the flight.
For the preliminary assessment of the possible occurrence of the somatogravic illusions the
GIF (gravito-inertial force) action angle value is usually analyzed (the “illusory pitch”).
Note: The gravito-inertial force constitutes in the longitudinal and vertical G effect on the
pilot. The angle of action of this force may be approximately assessed according to
the following formula:
𝑁𝑥
𝐺𝐼𝐹 ≈ 𝑎𝑟𝑐𝑡𝑔 𝑁𝑦, where Nx, Ny – longitudinal and vertical G, recorded by FDR.
The formula is estimative as it has a range of limitations and assumptions. For instance, it
incorporates neither the angular rates of the aircraft motion, nor the dynamic processes, associated
with the function of the human organs of perception (the otoliths and semicircular canals). It is
also obvious that the formula is restricted on the near-zero values of the vertical G.
The results of the calculations by formula are given on Fig. 73. The plot is the evidence
that within time intervals of 00:40:52 – 00:40:56, 00:41:05 – 00:41:09, 00:41:14 – 00:41:23 and
from 00:41:33 up to the end of the flight there are substantial differences in the values of the
recorded pitch and GIF angle, with that the GIF angle is greater, that means that there are
prerequisites for the occurrence of the “pitch-up” illusion.

Fig. 73. The estimated GIF angle
In the progress of the discussion of the draft Final Report the State of the Operator
expressed the viewpoint that in 2 or 3 seconds after the initiation of another go-around the PIC
had been fallen under the influence of the pitch-up illusion and all his further actions, having led
to the accident, had been carried out affected by that illusion.
The investigation team does not agree with this position. The investigation team notes that
the presence of conditions for the somatogravic illusions (in this very case the mismatch of the
actual pitch and GIF angle) is a necessary, but not sufficient condition for its actual effect on
pilot. Actually, the illusion emerges (the pilot is affected by it) at the flight out of sight of the
natural horizon and/or ground reference and at the deficient monitoring of the flight parameters on
instruments.
At the initial phase of go-around (that is to say when on the opinion of the State of
the Operator the PIC had fallen under the influence of the illusion) the flight had been actually
proceeding in VMC. According to the KRAMS-4 weather information archive at the moment of
the accident the cloud base of the scattered (4 octants) cloud had been equal to 420 m above the
RWY level (1660 ft QNH), RVR at the touchdown and midpoint equal to 7000 m. This altitude
had been passed at 00:41:04, that is in 14 seconds after the initiation of the go-around. It is highly
unlikely that the PIC, having piloted by HUD, that is at that very phase of flight having watched
both the instruments readings and the outside environment, had fallen under the influence of the

somatogravic illusions in 2-3 seconds after the initiation of go-around. The cloud base of the
broken (5-7 octants) cloud that had been equal to 1080 m above the RWY level (3820 ft QNH)
had never been reached.
Thus, the complex analysis of all the information available, stated above, is the evidence
that, most probably, both the intense forward repositions of the control column up to the initiation
of the stabilizer trim to nose-down and the trim itself, had not been attributable to the PF (the PIC)
having been affected by the “pitch-up” illusion.
At the last time interval being considered (after the start of the stabilizer trim to nose-down
and the outbreak of the near zero and negative G) there had been the explicit conditions for the
most intense “pitch-up illusion”. However at that phase of the flight the PIC was completely
disoriented and it does not seem possible to assess the potential impact of one or another factor
(the stress, negative G, somatogravic illusions) on it. The F/O up to the point of the aircraft impact
with ground had not been subject to spatial disorientation.
Consequently, the investigation team is of the opinion that the potential occurrence of the
somatogravic “pitch-up illusion” did not have crucial importance as far as the onset of the
accident situation is concerned. In the progress of the accident situation the pitch-up illusion
might have had some impact on the long-time keeping the trim switches pressed to nose-down.

3. Conclusion
The fatal air accident to the Boeing 737-8KN A6-FDN aircraft occurred during the second
go around, due to an incorrect aircraft configuration and crew piloting, the subsequent loss of PIC’s
situational awareness in nighttime in IMC. This resulted in a loss of control of the aircraft and its
impact with the ground. The accident is classified as Loss of Control In-Flight (LOC-I) occurrence.
Most probably, the contributing factors33 to the accident were:
 the presence of turbulence and gusty wind with the parameters, classified as a moderate-
to-strong “windshear” that resulted in the need to perform two go-arounds;
 the lack of psychological readiness (not go-around minded) of the PIC to perform the
second go-around as he had the dominant mindset on the landing performance exactly at the
destination aerodrome, having formed out of the “emotional distress” after the first unsuccessful
approach (despite the RWY had been in sight and the aircraft stabilized on the glide path, the PIC
had been forced to initiate go-around due to the windshear warning activation), concern on the
potential exceedance of the duty time to perform the return flight and the recommendation of the
airline on the priority of landing at the destination aerodrome;
 the loss of the PIC’s leadership in the crew after the initiation of go-around and his
“confusion” that led to the impossibility of the on-time transition of the flight mental mode from
“approach with landing” into “go-around”;
 the absence of the instructions of the maneuver type specification at the go-around
callout in the aircraft manufacturer documentation and the airline OM;
 the crew’s uncoordinated actions during the second go-around: on the low weight
aircraft the crew was performing the standard go-around procedure (with the retraction of landing
gear and flaps), but with the maximum available thrust, consistent with the Windshear Escape
Maneuver procedure that led to the generation of the substantial excessive nose-up moment and
significant (up to 50 lb/23 kg) “pushing” forces on the control column to counteract it;
 the failure of the PIC within a long time to create the pitch, required to perform go-
around and maintain the required climb profile while piloting aircraft unbalanced in forces;
 the PIC’s insufficient knowledge and skills on the stabilizer manual trim operation,
which led to the long-time (for 12 sec) continuous stabilizer nose-down trim with the subsequent
substantial imbalance of the aircraft and its upset encounter with the generation of the negative G,
which the crew had not been prepared to. The potential impact of the somatogravic “pitch-up
33
In compliance with Manual of Aircraft Accident and Incident Investigation (ICAO Doc 9756) the factors are stated
irrespective of their priority. The determination of the contributing factors does not presume the apportion of blame
or liability.
illusion” on the PIC might have contributed to the long keeping the stabilizer trim switches
pressed;
 the psychological incapacitation of the PIC that resulted in his total spatial
disorientation, did not allow him to respond to the correct prompts of the F/O;
 the absence of the criteria of the psychological incapacitation in the airline OM, which
prevented the F/O from the in-time recognition of the situation and undertaking more decisive
actions;
 the possible «operational» tiredness of the crew: by the time of the accident the crew
had been proceeding the flight for 6 hours, of which 2 hours under intense workload that implied
the need to make non-standard decisions; in this context the fatal accident occurred at the worst
possible time in terms of the circadian rhythms, when the human performance is severely degraded
and is at its lower level along with the increase of the risk of errors.
The lack of the objective information on the HUD operation (there were no flight tests of
the unit carried out into the entire range of the operational G, including the negative ones; the
impossibility to reproduce the real HUD readings in the progress of the accident flight, that is the
image the pilot was watching with the consideration of his posture in the seat trough the stream
video or at the FFS) did not allow making conclusion on its possible impact on the flight outcome.
At the same time the investigation team is of the opinion that the specific features of the HUD
indication and display in conditions existed during final phase of the accident flight (severe
turbulence, the aircraft upset encounter with the resulting negative G, the significant difference
between the actual and the target flight path) that generally do not occur under conditions of the
standard simulator sessions, could have affected the situational awareness of the PIC, having been
in the highly stressed state.

4. The shortcomings, revealed in the investigation

The shortcomings are referred to in the text of the Report.

5. Safety Recommendations 34
It is recommended that FATA35
5.1. Inform the flight personnel, the air training centers staff and the ATC officers on the results
of the investigation in question at the subject-oriented debriefing meetings.
5.2. Consider the possibility of expediting of the Russian Federation transition to the flight
operations on QNH.
5.3. Draw the attention of organizations, engaged in the aviation personnel testing for their
compliance with the ICAO Language Proficiency Requirements, to the need for enhanced
monitoring of the raters-examiners’ activities, as well as further develop the measures to exclude
the assignment of the ICAO language proficiency level to persons, whose level in question does
not meet the subject requirements.
5.4. Draw the attention of the air navigation service providers to the need for enhanced
monitoring and responsibility of the instructor personnel, engaged in the officers’ simulator
training, inter alia of the adherence to the R/T rules in English.
5.5. In association with Roshydromet and the State ATM Corporation, FSUE organize training
with the ATC and meteorological services officers on the procedure of the information
communication on windshear to the flight crews.
5.6. In view of the position of the aircraft manufacturer that the Boeing 737-800 operational
documentation implies the presence of the pilots’ previous experience of operating the jet multi-
engine aircraft, familiarization with the basic systems and basic airmanship, assess the risks of the
pilots’ approval for the type in case it is the first jet multi-engine airplane in the pilot’s career. If
required, amend the current regulations. Assess the applicability of this safety recommendation as
for the other aircraft types.
It is recommended that the Flydubai airline36
5.7. Conduct the flight personnel training on the specific aspects of the stabilizer trim manual
operation (the forces trim).
5.8. Consider the practicability of the HUD Model 4000 changing to Model 6000
(STC ST02522SE) to equip both pilots’ duty stations.
34
In accordance of ICAO Annex 13 provisions, the safety recommendations are developed with the intention of
preventing accidents or incidents and in no case have the purpose of creating a presumption of blame or liability for
the specific air accident.
35
It is recommended that the aviation authorities of the participant states of the Agreement consider the applicability
of the Safety Recommendations based on the actual state of affairs as for them.
36
The other airlines are recommended to consider the applicability of these Safety Recommendations with the account
of the operated aircraft fleet.
5.9. In association with the aircraft and HGS manufacturers, consider the practicability of the
development of the additional instructional guidelines on the HUD use at the different stages of
flight.
5.10. Consider the practicability of the elaboration of the flight personnel training programs,
allowing for the incorporation of the practical familiarization (training) of the pilots with the upset
conditions, including zero and negative G state.
5.11. Consider the practicability to amend the airline OM with the criteria of the psychological
incapacitation and the respective recommended actions.
5.12. Consider the practicability of the elaboration of SOP in terms of specifying the type of the
next maneuver (for example Go-Around, Windshear Escape Maneuver) as far as the callout by the
PF is concerned37.
5.13. Consider the practicability to elaborate SOP in terms of monitoring of the trim duration
and the current stabilizer position.
5.14. Evaluate the possible risks, associated with the partial blocking of the PFD at the significant
forward deflection of the control column and take measures on the risks mitigation (if required).
It is recommended that FAA, Rockwell-Collins
5.15. Consider the practicability of the conduct of the additional flight tests of HGS into all the
anticipated operating conditions and the entire range of G of the aircraft with these systems
installed.
5.16. Taking into account the views of the experts in ergonomics and aviation psychology,
consider the practicability of the improvement of the HUD information presentation in order to
mitigate the risk of its erroneous interpretation.
5.17. In association with the designers and manufacturers of the aircraft, equipped with the HGS,
consider the practicability of development of the additional guidance on the use of HUD at
different stages of flight.
It is recommended that FAA, the other certification authorities (EASA, IAC Aviation
Register, FATA etc.)
5.18. Consider the practicability of the amendment of the aviation regulations with the provisions
on the mandatory flight assessment of the flight parameters indication systems to pilots into the
entire operating range of the aircraft with such systems installed.
5.19. Consider the practicability of the amendment of the aviation regulations that determine the
procedure of the STC issue for the indication systems to pilots, with the requirement to the
manufacturer of the equipment in question to have the hardware/software package available to
37
As per the information available, the Flydubai airline has implemented this safety recommendation. The safety
recommendation has been retained in the section for its consideration by the other airlines.
reproduce the indication as per the FDR data in real time and in the scope, sufficient for the
investigation of the aircraft accidents and incidents.
It is recommended that the Boeing Company
5.20. Consider the practicability of the amendment of the FCT 737 NG (ТМ) Low Altitude Level
Off - Low Gross Weight section with more detailed information on the criteria that the pilots
should follow to determine the point, when the maximum thrust should be reduced, including go-
around performance in windshear.
5.21. Consider the practicability to implement the design changes of the stabilizer control system
to reduce the risk for the pilot to set stabilizer in-flight into out of trim position38.
5.22. Consider the practicability of the elaboration of SOP in terms of specifying the type of the
Go-Around maneuver (for example Go-Around, Windshear Escape Maneuver) as far as the callout
by the PF is concerned.
5.23. In association with FAA assess the possible risks, arising due to the partial blockage of
PFD at considerable forward deflection of the control column and take measures on their
mitigation (if necessary).
5.24. Taking into consideration the information, stated in Section 1.18.2, consider the
practicability of introduction of additions and amendments to FCOM and/or FCTM, explaining
the stabilizer control sequence on the Boeing 737 aircraft under different conditions. Consider the
applicability of this safety recommendation for the other aircraft families.
5.25. Consider the practicability of the introduction of additions and amendments to FCOM
and/or FCTM, explaining general principles of the stabilizer use and forces trim, as well as the
monitoring of the current stabilizer setting.
It is recommended that ICAO
5.26. Consider the practicability of establishing a working group to study the issues of the
psychological incapacitation of the flight crewmembers and elaboration of the relevant
recommendations to provide guidance to the operators and States in the OM draw up and approval.
38
In the Comments to the draft Final Report the aircraft manufacturer suggested to remove this recommendation,
reasoning that the Boeing Company design philosophy implies the pilot can fully operate with the available deflection
of flight controls, including the stabilizer control. This may be required in a variety of non-normal situations, for
example at the total loss of hydraulic system/hydraulic circuits’ pressure. At the same time according to the
manufacturer, the aircraft design provides for reasonable amount of engineering concepts for the PM to stop the in-
flight stabilizer setting into the out of trim position by the PF. The investigation team agrees that the aircraft design
allows for it. But, the air accidents investigation practice shows that the PM, who monitors the flight management and
aircraft control actions by the PF, is not always able to promptly identify the out-of-trim stabilizer position, as well as
to detect the mere fact of the stabilizer prolonged motion. The investigation team notes that at the current level of
technological development the combination of the mentioned engineering concepts is a possible solution: the
limitation of the stabilizer deflection angles, when this could result in the adverse consequences, and the full
travel/deflection when actually necessary.

Fly Dubai Accident

Uploaded by

Copyright:

Available Formats

Fly Dubai Accident

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fly Dubai Accident

Uploaded by

Copyright:

Available Formats

This document is an English translation of the Final Report on the accident involving the

INTERSTATE AVIATION COMMITTEE

Type of occurrence Fatal accident

INTERSTATE AVIATION COMMITTEE

INTERSTATE AVIATION COMMITTEE

CDS  Common Display System

INTERSTATE AVIATION COMMITTEE

FD  flight dispatcher of the airline

INTERSTATE AVIATION COMMITTEE

IAS  indicated airspeed

INTERSTATE AVIATION COMMITTEE

INTERSTATE AVIATION COMMITTEE

s/n  serial number

INTERSTATE AVIATION COMMITTEE

INTERSTATE AVIATION COMMITTEE

1.2. Injuries to Persons

Injuries to persons Crew Passengers Others

Minor/None 0/0 0/0 0/0

INTERSTATE AVIATION COMMITTEE

1.3. Damage to Aircraft

INTERSTATE AVIATION COMMITTEE

Fig. 2. The element of the F/O control column

INTERSTATE AVIATION COMMITTEE

Fig. 4. The fragments of the RH wing panel

Fig. 5. The fragments of the LH wing panel

Wing high-lift devices (flight control surfaces)

Fig. 6. Jackscrews and single unit tracks

INTERSTATE AVIATION COMMITTEE

Fig. 8. Nose landing gear

INTERSTATE AVIATION COMMITTEE

Fig. 9. The elements of the right main landing gear leg

INTERSTATE AVIATION COMMITTEE

INTERSTATE AVIATION COMMITTEE

Fig. 12. The recovered elements of the stab trim actuator

Roll control system

INTERSTATE AVIATION COMMITTEE

The aircraft fuel system

Fig. 13. The air conditioning system outflow valve

Water supply system and catering equipment

INTERSTATE AVIATION COMMITTEE

Fig. 15. The left engine fragments

Fig. 16. The right engine fragments

INTERSTATE AVIATION COMMITTEE

Fig. 17. The HF radio set switching unit fragments

Electrical power supply system

INTERSTATE AVIATION COMMITTEE

Fig. 18. The generator rotor

Several integral units to the electrical system were recovered:

Fig. 19. The standby power control unit

 the transformer-rectifier unit (TRU) (Fig. 20);

INTERSTATE AVIATION COMMITTEE

Fig. 20. The transformer-rectifier unit

Fig. 21. The bus power control unit No 1

INTERSTATE AVIATION COMMITTEE

Fig. 22. The bus power control unit No 2

All the units sustained considerable damage.

Fig. 23. The debris of the aircraft electrical supply

Fire protection system

Fig. 24. The stab trim motor elements

The oxygen supply equipment