3-2009_G-THOF

AIRCRAFT ACCIDENT REPORT 3/2009
Air Accidents Investigation Branch
Department for Transport
Report on the serious incident to

Boeing 737-3Q8, registration G-THOF
on approach to Runway 26
Bournemouth Airport, Hampshire
on 23 September 2007
This investigation was carried out in accordance with

The Civil Aviation (Investigation of Air Accidents and Incidents) Regulations 1996
The sole objective of the investigation of an accident or incident under these Regulations
shall be the prevention of accidents and incidents. It shall not be the purpose of such an
investigation to apportion blame or liablility.
© Crown Copyright 2009
Published with the permission of the Department for Transport (Air Accidents Investigation Branch).
This report contains facts which have been determined up to the time of publication. This information
is published to inform the aviation industry and the public of the general circumstances of accidents and
serious incidents.
Extracts may be published without specific permission providing that the source is duly acknowledged.
Published 21 May 2009
Printed in the United Kingdom for the Air Accidents Investigation Branch
ii
RECENT FORMAL AIRCRAFT ACCIDENT AND INCIDENT REPORTS
ISSUED BY THE AIR ACCIDENTS INVESTIGATION BRANCH
THE FOLLOWING REPORTS ARE AVAILABLE ON THE INTERNET AT

http://www.aaib.gov.uk
1/2008 Bombardier CL600-2B16 Challenger 604, VP-BJM January 2008

8 nm west of Midhurst VOR, West Sussex
on 11 November 2005.
2/2008 Airbus A319-131, G-EUOB January 2008

during the climb after departure from London Heathrow Airport
on 22 October 2005.
3/2008 British Aerospace Jetstream 3202, G-BUVC February 2008

at Wick Aerodrome, Caithness, Scotland
on 3 October 2006.
4/2008 Airbus A320-214, G-BXKD February 2008

at Runway 09, Bristol Airport
on 15 November 2006.
5/2008 Boeing 737-300, OO-TND April 2008

at Nottingham East Midlands Airport
on 15 June 2006.
6/2008 Hawker Siddeley HS 748 Series 2A, G-BVOV August 2008

at Guernsey Airport, Channel Islands
on 8 March 2006.
7/2008 Aerospatiale SA365N, G-BLUN October 2008

near the North Morecambe gas platform, Morecambe Bay
on 27 December 2006.
1/2009 Boeing 737-81Q, G-XLAC January 2009

Avions de Transport Regional ATR-72-202, G-BWDA, and
Embraer EMB-145EU, G-EMBO
at Runway 27, Bristol International Airport
on 29 December 2006 and 3 January 2007.
2/2009 Boeing 777-222, N786UA April 2009

at London Heathrow Airport
on 26 February 2007.
iii
Farnborough House
Berkshire Copse Road
Aldershot
Hampshire GU11 2HH
April 2009
The Right Honourable Geoff Hoon

Secretary of State for Transport
Dear Secretary of State
I have the honour to submit the report by Mr K Conradi, an Inspector of Air Accidents, on
the circumstances of the incident to Boeing 737-3Q8, registration G-THOF, on approach to
Runway 26, Bournemouth Airport, Hampshire on 23 September 2007.
Yours sincerely
David King
Chief Inspector of Air Accidents
iv
Contents
Synopsis............................................................................................................................. 1
1 Factual Information............................................................................................... 3
1.1 History of the flight......................................................................................... 3
1.2 Injuries to persons........................................................................................... 6
1.3 Damage to aircraft........................................................................................... 7
1.4 Other damage.................................................................................................. 7
1.5 Personnel information..................................................................................... 7
1.5.1 Commander....................................................................................... 7
1.5.2 First Officer....................................................................................... 7
1.6 Aircraft information........................................................................................ 8
1.6.1 Leading Particulars........................................................................... 8
1.6.2 Autothrottle system........................................................................... 8
1.6.3 Autothrottle disengagement............................................................ 10
1.6.4 Built-in Test Equipment (BITE) fault history................................. 11
1.6.5 Aircraft fault history....................................................................... 11
1.7 Meteorological information.......................................................................... 11
1.7.1 Forecast........................................................................................... 11
1.7.2 Meteorological information received en route................................ 11
1.8 Aids to navigation......................................................................................... 12
1.9 Communications........................................................................................... 12
1.10 Aerodrome information................................................................................. 12
1.11 Flight Recorders............................................................................................ 13
1.11.1 FDR/CVR....................................................................................... 13
1.11.2 Radar Recording............................................................................. 13
1.12 Wreckage and impact information................................................................ 13
1.13 Medical and pathological.............................................................................. 13
1.14 Fire................................................................................................................ 13
1.15 Survival aspects............................................................................................ 13
1.16 Tests and research......................................................................................... 15
1.16.1 General............................................................................................ 15
1.16.2 Autothrottle system testing............................................................. 15
v
1.16.3 Autothrottle computer testing......................................................... 15
1.16.4 Autothrottle operation statistics...................................................... 16
1.16.5 Aircraft manufacturer’s QAR performance analysis...................... 17
1.17 Organisational and management information............................................... 18
1.17.1 ASR processing............................................................................... 18
1.17.2 The Mandatory Occurrence Reporting Scheme.............................. 20
1.17.3 MOR scheme CAP 382 - items to be reported............................... 20
1.18 Additional information.................................................................................. 21
1.18.1 Similar events................................................................................. 21
1.18.2 EASA CS-25 certification requirements......................................... 21
1.18.3 UK CAA Paper 2004/10 Flight Crew Reliance on Automation..... 23
1.18.4 Flight crew training and manuals ................................................. 23
1.18.4.1 Operations Manual Part B . ........................................ 23
1.18.4.2 The operator’s Quick Reference Handbook (QRH)
procedure for approach to stall recovery...................... 24
1.18.4.3 The operator’s QRH procedure for upset recovery...... 24
1.18.4.4 Flight Crew Training Manual....................................... 25
1.18.4.5 ‘Airplane’ upset recovery training aid......................... 26
1.18.5 Pitch power couple.......................................................................... 27
2 Analysis.................................................................................................................. 29
2.1 Introduction................................................................................................... 29
2.1.1 Autothrottle disengagement............................................................ 29
2.1.2 Disengagement alerting requirements............................................ 30
2.2 Crew reactions............................................................................................... 31
2.3 Aircraft pitch excursion................................................................................ 32
2.4 ASR . ............................................................................................................ 35
2.5 The MOR Scheme......................................................................................... 37
3 Conclusions........................................................................................................... 39
3.1 Findings......................................................................................................... 39
3.2 Causal factors................................................................................................ 41
3.3 Contributory factors...................................................................................... 41
4 Safety Recommendations..................................................................................... 42
vi
Appendices
Appendix A Fault History

Appendix B Operator Procedures
Appendix C Extracts from: Manufacturer’s Upset Recovery Training Aid
vii
GLOSSARY OF ABBREVIATIONS USED IN THIS REPORT
AAIB Aircraft Accidents Investigation ft feet

Branch hrs hours (clock time as in 12:00 hrs)
aal above airfield level ILS Instrument Landing System
AFCS Automatic Flight Control Systems IMC Instrument Meteorological
AMM Aircraft Maintenance Manual Conditions
ASR Air Safety Report JAA Joint Aviation Authories
ATC Air Traffic Control kt knot(s)
ATE Automatic Test Equipment m metres
ATIS Automatic Terminal Information mb millibar(s)
System MCP Mode Control Panel
ATP Automatic Test Procedure MHz Megahertz
BITE Built-In Test Equipment MOR Mandatory Occurrence Report
CAA Civil Aviation Authority N1 engine fan or LP compressor speed
CAS Calibrated Air Speed nm nautical mile(s)
CS Certification Standards NPA Notice of Proposed Amendment
CVR Cockpit Voice Recorder OFDM Operational Flight Date Monitoring
CWS Control Wheel Steer PF Pilot Flying
°M,T magnetic, true PM Pilot Monitoring
DME Distance Measuring Equipment PNF Pilot Not Flying
EASA European Aviation Safety Agency QAR Quick Access Recorder
EICAS Engine Indication and Crew Alerting QNH pressure setting to indicate elevation
System above mean sea level
FAA Federal Aviation Regualtion (USA) QRH Quick Reference Handbook
FCTM Flight Crew Training Manual TAF Terminal Aerodrome Forecast
FDM Flight Data Monitoring TOGA Takeoff or Go-around
FDR Flight Data Recorder TRTO Type Rating Training Organisation
FFS Full Flight Simulator UTC Co-ordinated Universal Time (GMT)
FMC Flight Management Computer VREF Reference airspeed (approach)
F/O First Officer
FORCE Flight Operations Research Centre of
Excellence
viii
Aircraft Accident Report No: 3/2009 (EW/C2007/09/12)
Registered Owner and Operator Thomsonfly Ltd
Aircraft Type Boeing 737-3Q8
Nationality British
Registration G-THOF
Place of Incident On approach to Runway 26 at Bournemouth Airport,

Hampshire
Date and Time 23 September 2007 at 2250 hrs

(All times in this report are UTC)
Synopsis
The Air Accidents Investigation Branch was notified by the operator on the 5 October 2007
of an unstable approach and stall during a go-around by a Boeing 737-300 aircraft,
G-THOF, at Bournemouth Airport. The event had occurred 12 days previously on the
23 September 2007.
The following Inspectors participated in the investigation:
Mr K Conradi Investigator-in-charge
Mr A Blackie Operations
Ms A Evans Engineering
Mr P Wivell Flight Data Recorders
The Boeing 737-300 was on approach to Bournemouth Airport following a routine passenger
flight from Faro, Portugal. Early in the ILS approach the auto-throttle disengaged with the
thrust levers in the idle thrust position. The disengagement was neither commanded nor
recognised by the crew and the thrust levers remained at idle throughout the approach.
Because the aircraft was fully configured for landing, the air speed decayed rapidly to a
value below that appropriate for the approach. The commander took control and initiated a
go-around. During the go-around the aircraft pitched up excessively; flight crew attempts
to reduce the aircraft’s pitch were largely ineffective. The aircraft reached a maximum pitch
of 44º nose-up and the indicated airspeed reduced to 82 kt. The flight crew, however, were
able to recover control of the aircraft and complete a subsequent approach and landing at
Bournemouth without further incident.
1
Although the commander reported the event to the operator the following morning, his
initial Air Safety Report (ASR) contained limited information and the seriousness of the
event was not appreciated until the Quick Access Recorder (QAR) data was inspected on
4 October 2007.
G-THOF was not subjected to an engineering examination to ensure its continued

airworthiness and remained in service throughout this period.
The investigation identified the following causal factors:
1. The aircraft decelerated during an instrument approach, to an airspeed

significantly below the commanded speed, with the engines at idle
thrust. Despite the application of full thrust, the aircraft stalled, after
which the appropriate recovery actions were not followed.
2. The trimmed position of the stabiliser, combined with the selection of

maximum thrust, overwhelmed the available elevator authority.
The investigation identified the following contributory factors:
1. The autothrottle warning system on the Boeing 737-300, although

working as designed, did not alert the crew to the disengagement of the
autothrottle system.
2. The flight crew did not recognise the disengagement of the autothrottle
system and allowed the airspeed to decrease 20 kt below Vref before
recovery was initiated.
Three Safety Recommendations have been made.
2
1 Factual Information
1.1 History of the flight
This section is based on information gathered during crew interviews and data
retrieved from the QAR. The Cockpit Voice Recorder (CVR) and Flight Data
Recorder (FDR) recordings had been overwritten and therefore the incident
flight information was unavailable.
G-THOF was planned to operate from Faro in Portugal to Bournemouth,

Hampshire on a scheduled night passenger flight with the First Officer (F/O)
as pilot flying (PF). Before departing Faro the crew discussed the weather
at Bournemouth, uplifted additional fuel to permit two approaches and
decided on a full flap (flap 40) landing. As part of the pre-flight actions the
PF programmed the Flight Management Computer (FMC) for a final approach
commencing at 10 nm. During the cruise the PF briefed the approach and
reiterated his intention to use flap 40 for the landing. Both pilots confirmed
the flap 40 reference speed (Vref 40) of 129 kt and the intended final approach
speed of 135 kt (Vref 40 +6). The flight was routine until the approach at
Bournemouth.
There was no other traffic in the area so the crew were cleared to self-position
for the ILS approach to Runway 26. At 2245 hrs the aircraft was 11 nm from
Bournemouth, level at 2500 ft, with a Calibrated Air Speed (CAS) of 180
kt and flap 5 set. The autothrottle was engaged in speed mode1, with N12
averaging approximately 60%. Autopilot B was engaged in CMD mode with
VOR-LOC and Altitude Hold modes engaged.
The aircraft was level at 2,500 ft for 90 seconds and at seven nm DME the
autopilot captured the glideslope. The PF asked for the landing gear to be
lowered, flap 15 to be selected and the landing check list. The commander
carried out the actions although neither pilot could recall who had moved the
speedbrake handle, which was placed at 12º, slightly beyond the armed setting
of 9º. The PF then selected a lower speed on the mode control panel (MCP)
and, as expected, the autothrottle retarded the thrust levers to idle to reduce to
this speed.
The aircraft started to descend on the glideslope, and about 20 seconds

later, with the thrust levers still at idle the autothrottle disconnect warning
1 Speed mode will attempt to fly the airspeed selected by the flightcrew on the mode control panel.
2 N1 is the rotational speed of the first stage of the engine compressor. Expressed in a percentage it is used as a guide
to the amount of power the engine is producing.
3
was triggered and the autothrottle disengaged.3 This disengagement was not
recognised by the flight crew and no manual disconnect was recorded4; the
thrust levers remained at idle throughout the remainder of the approach.
The autopilot remained engaged and continued to track both the localiser and
the glideslope. The aircraft’s speed decayed at about one knot per second, in
line with the PF’s expectations for the approach. As the speed decreased below
150 kt, flap 25 was selected. The autopilot tracked the glideslope accurately,
gradually increasing the pitch of the aircraft to minimise glideslope deviation
and adjusting the stabiliser angle to keep the aircraft in trim. Temporary
reductions in pitch were evident during flap position transitions.
The PF increased the illumination of his maplight to check the placard speed
for the selection of flap 40, turned his light back to its previous level and called
for flap 40. The commander moved the flap lever and the PF then selected
135 kt on the MCP. The commander observed the flaps move to the flap
40 position and then completed the landing checklist by calling “flaps”. The
PF checked the flap gauge showed 40 and responded “flap 40 green light5”.
Recorded data shows the flaps had reached the flap 40 position when the
airspeed was 130 kt (Final Approach Speed -5 kt) and the aircraft was slowing
at approximately 1.5 kt per second. The commander stowed the checklist
on top of the instrument panel and when he looked down he saw an IAS of
125 kt. He called “speed”, the PF made a small forward movement with the
thrust levers and the commander called “I have control”. The commander
moved the thrust levers fully forward and called “go-around flap 15 check
thrust”.
Recorded data shows that, at a CAS of 110 kt and an altitude of 1,540 ft, the
autothrottle manual disconnect was pressed and the thrust levers moved forward
slightly. Within 1.5 seconds the stick-shaker (stall warning) activated and in
the following two seconds the thrust levers were advanced to the full forward
position. The autopilot mode changed from localiser and glideslope to Control
Wheel Steer (CWS) pitch and CWS roll6. The aircraft pitch attitude which had
been steadily increasing under the influence of pitch trim, reached 12º nose-up.
The automatic pitch trim stopped at 4.9º (7.9 units) of stabiliser trim.
The commander moved the control column forward to counteract the expected
3 The methods of recording these events are different and it is plausible that they occurred in the reverse order.
4 Manual disconnect of the autothrottle is sampled eight times a second.
5 Referring to the leading edge lift devices.
6 With CWS mode engaged the autopilot manoeuvres the aircraft in response to manually applied pressure on either
pilot’s control column.
4
pitch-up moment from the increased thrust, which arrested the increase and
reduced the pitch to 5º nose-up. The stick-shaker operation stopped and the
minimum airspeed was 101 kt. A small, apparently unintended application of
right aileron induced a right roll.
Four seconds after the thrust levers reached the fully forward position, with
airspeed increasing and N1 on both engines increasing through 81%, the TOGA
mode became active. The autopilot disengaged, the pitch attitude started to
increase again and the stick-shaker reactivated. A corrective roll input was
made to bring the aircraft wings level, and although the control column was
positioned fully forward the nose-up pitch increased to 22º.
The airspeed increased to 118 kt CAS. The pitch attitude appeared to stabilise
at 22º nose-up and the angle of attack started to decrease. Both engines were
producing 96-98% N1, which was in excess of the rated go-around thrust of
94%. The first officer selected the flaps to 15 and looked at the N1 reading
which he recalls as being 95%. The stick-shaker ceased but, as the flaps
retracted past the flap 25 position, the nose of the aircraft began to pitch up
at an increasing rate and a small continuous left rudder input started a left
roll. As the flaps reached flap 15 the pitch angle was increasing through 27º
and left roll was increasing through 7º. The stick-shaker reactivated, full
nose-down elevator was still being applied and the airspeed began to decay.
The first office called “high pitch” and the commander responded “I have
full forward stick”.
The F/O, although he was now the Pilot Monitoring (PM)7, also held full
forward stick; both pilots reported that they had no pitch control authority.
The airspeed had decreased rapidly but neither pilot was fully aware of exactly
what the airspeed was.
As aircraft pitch increased above 36º nose-up, the TOGA mode disengaged, the
left roll increased beyond 13º and the CAS decreased below 107 kt. A small
sharp right rudder input recovered the roll from a maximum of 22º left wing
down to wings level as the aircraft stalled with a peak pitch of 44º nose‑up.
With no change in elevator position the pitch rate reversed from positive to
negative, although angle of attack continued to increase as the aircraft started
to descend. Despite reducing pitch, the airspeed continued to decrease for a
further five seconds to a minimum recorded CAS of 82 kt when the pitch was
33º nose-up.
7 This role was previously known as Pilot Not Flying (PNF). PM is now common as it more accurately describes the
role.
5
The pitch angle reduced to 20º over 10 seconds, airspeed began to rise rapidly
and five seconds after the minimum recorded speed, the thrust was reduced to
86%. The pitch-down rate increased with the pitch reducing a further 15º in two
seconds.
The aircraft then stabilised in a 5º nose-up attitude, the speed increased and the
commander regained control of the aircraft. The TOGA mode was re-engaged
as the CAS reached 147 kt.
The commander initially levelled the aircraft at 3,000 ft before climbing to

4,000 ft and self positioning for a second approach. While downwind the
F/O told the commander that “the autothrottle did not capture it”. The
commander remained PF throughout the second approach which was conducted
with the autopilot and the autothrottle engaged. Both auto systems performed
normally throughout the second approach. The autopilot and autothrottle
were manually disengaged at 1,200 ft and 800 ft respectively and the aircraft
landed at 2301 hrs.
After the engines were shut down on stand, the commander spoke with the
operator’s base engineer. He told the engineer that, although he thought that
the aircraft was serviceable, there had been an incident and the company
would want the flight data. No defects were entered in the technical log.
The engineer assured the captain that the Operational Flight Data Monitoring
(OFDM) information was sent from the aircraft by an automatic mobile
telephone based datalink. The commander and F/O discussed the incident
before going off duty.
The next morning the commander returned to the airfield and telephoned the
operator’s safety department to advise them of the incident. He completed
an operator Air Safety Report (ASR) using the company’s internet based
system. This ASR was not reported to the Civil Aviation Authority until
4 October 2007.
1.2 Injuries to persons
Injuries Crew Passengers Others

Fatal 0 0 0
Serious 0 0 0
None 5 132 0
6
1.3 Damage to aircraft
None.
1.4 Other damage
None.
1.5 Personnel information
1.5.1 Commander
Male, Age 56 years

Licence: Airline Transport Pilot’s Licence
LPC/OPC renewed: 12 May 2007
Line check renewed: 28 July 2007
Medical certificate: JAA Class 1 issued 20 March 2007
Flying experience: Total 11,280 hours (of which 420 were on type)
Last 90 days 190 hours
Last 24 hours 5 hours
Previous rest period: 15 hours 45 mins
The commander had previously spent approximately 17 years on the operator’s

Boeing 757/767 fleet as a first officer and had completed a combined type
conversion and command upgrade course in 2006.
1.5.2 First Officer
Male, Age 30 years

Licence: Commercial Pilot’s Licence
LPC/OPC renewed: 12 June 2007
Line check renewed: 19 April 2007
Medical certificate: JAA Class 1 Issued 6 August 2007
Flying experience: Total 3,170 hours (of which 845 were on type)
Last 24 hours 0 hours
Previous rest period: 36 hours
The first officer had previously flown DHC Dash 8-300 aircraft for a different
company before joining the operator and converting to the Boeing 737 in 2006.
7
The crew, who were both based at Bournemouth, commented that flap 40
landings were not normally conducted by Bournemouth based crews as only
one of their normal destinations required it’s use for performance reasons.
Neither pilot reported any medical or fatigue issues which they considered
could have affected their performance.
1.6 Aircraft information
1.6.1 Leading Particulars
Registration G-THOF
Type Boeing 737-3Q8
Serial No 26314
Year of Manufacture 1995
Airframe life at time of incident 37,061 hours on 23/9/07
Engines 2 x CFM56-3C1 (22,000lbs thrust each)
Serial Numbers L/H 858302
R/H 858245
Hours/Cycles L/H 32,621 hrs/18,465 cycles
R/H 34,842 horus/19,667 cycles
(at 16/10/07)
The aircraft held a current EASA Standard Certificate of Airworthiness, valid

until 27 March 2008
The last maintenance carried out was a 1A check on 20 September 2007.
There were no relevant deferred defects or technical log entries.
The aircraft weight and balance was within the prescribed limits.
The autothrottle computer (Boeing Part No 10-62017-30, Manufacturer’s P/N

735SUE10-12, S/N 5593) had been fitted to G-THOF on 24 February 2002. It
was removed following another autothrottle disengagement on 22 October 2007
described in paragraph 1.18.1 of this report (see page 21).
1.6.2 Autothrottle system
The autothrottle system automatically positions the thrust levers to maintain

a computed engine thrust level and forms part of the Automatic Flight
Control Systems (AFCS). It comprises a digital autothrottle computer, two
autothrottle servo mechanisms, two thrust lever synchros and various pilot
8
selection controls. The servomechanisms (for left and right engines) are the
interface between the system and the throttle operating linkages; thrust lever
position is fed back via the synchros. The autothrottle is selected, or ARMED,
by operating a two-position, solenoid-held paddle switch located on the left
side on the glareshield panel (Figure 1). Once ARMED the switch is held in
position unless disengaged. The autothrottle can be disengaged manually
either by moving the paddle switch to the OFF position, or by pressing either
of the push buttons that are located on the outboard end of each thrust lever. A
green annunicator light beside the switch is illuminated when the autothrottle
is engaged. Disengagement extinguishes the green light and illuminates a
flashing red A/T P/RST autoflight status annunciator on the instrument panel
(Figure 2). This can be cancelled by a second press on the buttons either side
of the thrust levers, or by pressing the switch light itself.
Figure 1
Autothrottle switch
Figure 2
Autothrottle red warning of disengagement
9
The AT/WARN autoflight status annunciator will illuminate with a flashing
amber light (Figure 3) if the autothrottle is not holding the target speed. There
are three conditions when this light will flash; the airspeed is 10 kt above the
target speed and not decreasing, the airspeed is 5 kt below the target speed
and not increasing, or the airspeed has dropped to alpha floor (a factor of 1.3
above the stall speed) during a dual channel autopilot approach. On approach,
with the aircraft decelerating, the caption will routinely flash for extended
periods.
Figure 3
Autothrottle amber off speed caution
Flight Mode Annunciators on the primary flight display also show the
autothrottle status and mode.
1.6.3 Autothrottle disengagement
Disengagement of the autothrottle will result from any of the following

conditions:
1. Movement of the autothrottle ARM switch to OFF
2. Activation of one of the autothrottle disengage switches on the

thrust levers
3. Detection of an autothrottle fault within the computer by its

Built-in Test Equipment (BITE), or in the interfacing systems
4. Two seconds having elapsed following touchdown
5. Thrust levers becoming separated by more than 10 degrees

during a dual channel autopilot approach after FLARE ARMED is
annunciated
10
6. Autopilot roll control requiring significant spoiler deployment
and thrust levers becoming separated, when flaps are less than
15º8 and the autothrottle is not in take-off or go-around mode
Autothrottle disengagement results in the ARM switch releasing to OFF and the
red autothrottle disengage lights flashing unless it has disengaged automatically
after touchdown when the lights will not illuminate.
Manual positioning of the thrust levers does not normally cause autothrottle
disengagement.
1.6.4 Built-in Test Equipment (BITE) fault history
The autothrottle computer contains a BITE program to monitor its function

and non-volatile memory to store fault information for the last 10 flights. Due
to the delay in notification of the incident the aircraft had completed more
than 10 flights and therefore the fault history information from the incident
had been overwritten.
1.6.5 Aircraft fault history
A recent fault history for the aircraft is included in Appendix A.
1.7 Meteorological information
1.7.1 Forecast
As part of the flight brief the crew received forecast weather for Bournemouth
before departure from Faro.
The weather provided was the 1807 hrs TAF valid between 1900 hrs and
0100 hrs. It forecast a surface wind of 200°/14 kt, 6,000 m visibility with
cloud broken at 800 ft. Temporary between 1900 hrs and 0100 hrs the
visibility was expected to reduce to 2,000 m in light rain and mist with cloud
broken at 300 ft and a 40% probability that temporarily, between 1900 hrs
and 0100 hrs the surface wind would be 200°/15 kt gusting to 25 kt.
1.7.2 Meteorological information received en route.
The last weather report recorded by the crew as they approached Bournemouth
was ATIS ‘H’ recorded at 2220 hrs. It reported Runway 26 in use with wind
8 This is monitored at the flap and not at the cockpit control. Therefore it is possible for the mode to be active with the
flap lever in the Flap 15 position if the flaps have not reached 15° deployment.
11
220°/14 kt, 4,000 m visibility in light rain, cloud overcast at 400 ft, temperature
and dewpoint plus 17ºC. The QNH was 1011 mb. The weather at the time of
the incident was the same.
Throughout the incident the aircraft was operating in Instrument Meteorological

Conditions (IMC) at night.
1.8 Aids to navigation
At the time of the incident the aircraft was established on the ILS to Runway 26
at Bournemouth Airport.
Runway 26 was equipped with a category one ILS DME transmitting on a

frequency of 110.5 Mhz with an approach gradient of 3º. The ILS was orientated
on the runway track of 258º.
1.9 Communications
Radio communications throughout the approach were routine.
Following the go-around the crew informed ATC that they had conducted a
go‑around but no mention of the incident was made.
1.10 Aerodrome information
Bournemouth (Hurn) Airport (EGHH) is located on the south coast of England

some 3.5 nm NNE of the city of Bournemouth. It has a single runway orientated
26/08 of 2,271 meters length by 46 meters wide. The runway in use at the
time of this incident was Runway 26 which was equipped with a CAT 1 ILS
with full night lighting including high intensity approach lighting (HIALS)
and 3º PAPI. The LDA for Runway 26 was 1,970 m. The standard missed
approach for Runway 26 was a climb straight ahead to 4 miles DME then a
left turn to return to the BIA locator at 3,000 ft. At the time of the incident a
temporary restricted area was in effect over the city of Bournemouth which
required a missed approach climb to 4,000 ft to remain clear. The aircraft did
not enter this restricted area.
12
1.11 Flight Recorders
1.11.1 FDR/CVR
This incident was reported to the AAIB 12 days after it occurred. This resulted
in the loss of Flight Data Recorder (FDR) and Cockpit Voice Recorder (CVR)
recordings.
The operator was running an Operational Flight Data Monitoring (OFDM)

programme using onboard equipment. This sent a copy of the QAR data to a
server at the operator’s headquarters, via a mobile phone service, whenever the
aircraft terminated a flight in the UK. This data was made available to the AAIB
and extracts are shown in Figure 4.
1.11.2 Radar Recording
Pease Pottage radar recorded the aircraft track, as well as a number of transmitted
parameters, at a rate of approximately once every 6 seconds. This information
was not as comprehensive as the QAR data so it was used for checking the FDM
data only.
Details from the recorded data were used in the construction of Section 1.1,
History of the flight (see page 3).
1.12 Wreckage and impact information
Not applicable.
1.13 Medical and pathological
Not applicable.
1.14 Fire
Not applicable.
1.15 Survival aspects
Not applicable.
13
Figure 4
Quick Access Recorder data depicting the accident
14
1.16 Tests and research
1.16.1 General
Both pilots reported that they had not seen the autothrottle disconnect warning.
The autothrottle system was functionally tested. The AAIB became aware that
there had been a number of other events possibly with similar precursors, ie an
autothrottle disconnect followed by a long period of warning without crew
recognition. Consequently, the efficacy of the autothrottle warning became of
interest to the investigation.
1.16.2 Autothrottle system testing
The autothrottle computer (P/N 735SUE10-12, S/N 5593) from the incident
flight was re-fitted to G-THOF and the Aircraft Maintenance Manual
(AMM) Task 22-31-00-735-060 autothrottle System Test was carried out
on 15 November 2007. The purpose of this test was to confirm the correct
functioning of the autothrottle system, including the autothrottle warnings.
No faults were found with the system during these tests. The autothrottle
computer was subsequently removed for further testing.
1.16.3 Autothrottle computer testing
The autothrottle computer was subjected to a full Automatic Test Procedure

(ATP) using the manufacturer’s Automatic Test Equipment (ATE). The
computer failed two of these tests which were related to analogue inputs
and outputs. These failures appeared to be small tolerance errors that could
have been attributable to the ATE, so the computer was then tested on a
manual rig where each individual test could be performed and observed. The
analogue inputs were checked and the results were within the specification
requirements. The analogue outputs were found to be out of tolerance at the
extreme negative (-10 Volts) end of the output range (the measured voltages
were up to -10.24 Volts). These results indicated that the analogue output on
the Output Interface Card was the likely source of the high voltage.
An internal visual inspection of the computer did not show any signs of damage
or failure. The Output Interface Card was interchanged with a serviceable card
and the analogue outputs were rechecked and were within specification. The
original card was returned to the computer but attached to an ‘extender’ card
allowing adjustment of the -10 Volt output to within the specified limits. With
the adjustment made the computer was then subject to a further ATP which it
passed successfully.
15
The Output Interface Card output voltage is used as a reference for the Digital
to Analogue Converter circuit, which generates the output signals to the
throttle servomechanism amplifiers and the flight director outputs. The same
reference signal is used for the autothrottle servo demand signals on both
engines. Another analogue output is used by the autothrottle itself as part of
the built-in test and if a fault is detected, would record a Digital to Analogue
Converter fault code within the fault history.
For this over-voltage to have caused an autothrottle disengagement, the thrust

levers would be required to move asymmetrically to generate a ‘throttle split’.
This is unlikely as the reference voltage is applied to both engine systems. Any
other internal faults which could have caused an autothrottle disengagement
are not dependant on the Digital to Analogue Converter reference voltage.
1.16.4 Autothrottle operation statistics
During the investigation, anecdotal information indicated that there may

have been problems with the autothrottle alerting system on the Boeing 737
aircraft. The information was not considered to be precise enough alone for
conclusions to be drawn. Therefore a more statistically based review of the
function of the autothrottle and its warnings was generated.
UK operators of large aircraft are required to run Operational Flight Data

Monitoring (OFDM) programmes. These programmes gather data on each
flight and analyse it for trends or events that can lead to improved safety
strategies. The analysis is largely in the form of setting thresholds of parameter
combinations that, if exceeded, highlight the flight for further review. Following
an approach from the AAIB a number of operators implemented the following
four triggers designed to see how prevalent Boeing 737 autothrottle problems
were:
1. An autothrottle warning preceding an autothrottle disconnect
2. An un-commanded autothrottle disconnect
3. An unacknowledged autothrottle disconnect warning
4. An unrecognised and un-commanded autothrottle disconnect
In order to have a high degree of confidence in the data it was decided that these
events should be implemented in such a way so as to ensure minimum nuisance
triggering, even at the cost of missing some genuine events. The triggered
16
events were broken down by phase of flight and only events associated with
the descent, approach and final approach were included in the final results,
generating a database of 5,284 sectors.
The algorithm implemented to capture the autothrottle warning preceding an

autothrottle disconnect did not capture any other similar events. It is of note
that whilst it would have triggered on the incident flight, it would not have
triggered on the second occasion when the aircraft autothrottle disconnected
itself (see 1.18.1, page 21).
0.4% of sectors flown had an autothrottle disconnect without recording

any manual disconnect switch activity. However, the autothrottle can be
disconnected using the switch on the mode control panel. This switch is
not recorded on the FDR, which makes the action indistinguishable from an
uncommanded autothrottle disconnect.
The “unacknowledged autothrottle disconnect warning” algorithm simply

highlighted any occasion when the autothrottle warning was active for more
than nine seconds. This was taken to be an indication that the warning had not
been noticed. It was established that extended warnings occurred on 2.5% of
sectors. [Note that this was based on a subset of 2,354 sectors.]
0.3% of sectors had the combination of the autothrottle disconnecting without

the use of the manual disconnect switches and the resultant warning lasting
for more than 9 seconds.
The study was a ‘first look’ and as such did not yield sufficiently robust results
to provide definitive occurrence rates.
1.16.5 Aircraft manufacturer’s QAR performance analysis
The aircraft manufacturer was provided with the recorded data in order to
conduct an engineering simulation of the approach and go-around with various
configurations and power selections.
The simulation determined that during this event the nose-up pitching moment,
generated as the engine thrust increased and by the stabiliser’s trimmed
position, overwhelmed the elevator until the recovery after the stall.
The simulation considered the amount of elevator deflection required to trim

the aircraft during the go-around manoeuvre. The initial configuration was
the baseline recorded data for a point shortly after initiation of the go-around
17
and with engines at 98.4% N1. This showed that the elevator required to
trim the aircraft at that stage was near the maximum nose-down deflection
available for the 737-300. Therefore there was little or no nose-down pitch
capability available.
Other simulations included having the stabiliser set for the intended approach
speed or the use of the rated go-around thrust of 94% N1. Both of these
configurations allow additional pitch authority (6.1º and 1.7º of elevator
deflection respectively).
A kinematic analysis was performed which showed no significant evidence of

external influences such as windshear.
The manufacturer was asked to consider the effect of changes in the flap
setting. They confirmed that there was a nose-up pitching moment associated
with changing the flaps from flap 40 to flap 15. This pitching moment created
by the flap change would require approximately 0.1 degrees of stabiliser to
maintain a trimmed condition. The trim change due to the flap change was
therefore seen to be negligible compared with the trim change due to the thrust
increase.
The manufacturer further commented that had the flaps remained at flap 40 and
all other parameters remained as they were throughout the event, the aircraft
would still have stalled. However it may have taken slightly longer to reach
the maximum nose-up pitch.
1.17 Organisational and management information
1.17.1 ASR processing
The commander’s Air Safety Report was received by the operator’s flight safety
office the day after the event occurred. All reports received, including this
one, were initially assessed for severity by a member of the safety team. Part
of this assessment was to decide if a Mandatory Occurrence Report (MOR)
was required to be filed with the CAA. To determine if this was required the
airline safety office referred to CAP 3829 which has a list of events that require a
MOR (see 1.17.3). If additional information was required in order to assist the
decision, ASRs were flagged electronically to the OFDM analyst.
This process left the ASR file open on the safety department computer system.
The system would then automatically generate an alert if the ASR was going to
9 Civil Aviation Publication 382 – The Mandatory Occurrence Reporting Scheme.
18
breach the company’s internal time limit for closure. At the time of the incident
this was set at 30 days.
At the time of the incident the OFDM analyst was located in the same building as
the safety department but on a different floor. The OFDM analyst would receive
an electronic alert to analyse a specific flight, review the data and respond back
to the safety office.
The OFDM analyst was not a pilot and, when an aircrew opinion was required
he referred events to one of the company’s pilot representatives (also known
as an ‘honest broker’). This system is common UK industry practice.
The pilot representative’s role was to view de-identified flight data safety
trends and advise the operator’s safety department. The pilot representatives
also liaised with the analyst to follow up ASRs and to place recorded data in
context with the ASR for the analyst. In events identified solely from OFDM,
they provided the link between the aircrew involved and the management.
The pilot representative role was part-time and in addition to his normal flying
commitments. At the time of the incident a pilot representative was available
in the OFDM office on average once every ten days. A pilot representative
was in the OFDM office the day following the incident but due to workload
he did not review the data downloaded from the aircraft. The next pilot
representative was on duty 11 days after the incident and he reviewed the data
and realised the severity of the event. The event was then progressed through
the safety system and the AAIB was alerted by the operator.
The company OFDM programme followed the guidance of CAP 73910 and
included an agreement (known as Schedule J) with the pilot’s union regarding
the conduct of OFDM and access to the data. Appendix 2 of that agreement
states:
‘Where a pilot has submitted a Flight Safety Report (FSR) or similar

and an incident follow-up is initiated, the crew and aircraft are
known. In these circumstances, relevant FDM data could prove
invaluable in improving Flight Safety and the Company Council
has agreed that the associated OFDM data does not need to be
de-identified.’
10 Civil Aviation Publication 739 – Flight Data Monitoring.
19
1.17.2 The Mandatory Occurrence Reporting Scheme
The CAA MOR scheme is outlined in CAP 382. The objective of the MOR
scheme is:
‘to contribute to the improvement of air safety by ensuring that

relevant information on safety is reported, collected, stored,
protected and disseminated.’
The requirements of the scheme were highlighted in the operator’s Flight

Operations Manual Part A, General/Basic section 11.3.7. This required that,
unless prevented by exceptional circumstances, reports were to be despatched
within 72 hours of the event.
The Air Navigation Order (ANO) specifies the categories of persons (or
organisations) who are required to report occurrences. These include:
‘a) the operator and the commander of a turbine-powered aircraft

which has a certificate of airworthiness issued by the CAA’
1.17.3 MOR scheme CAP 382 - items to be reported
Under Part 1 Section 1.1 Operation of the aircraft is a list of specified conditions
which are to be reported. Of relevance to the G-THOF incident are:
‘E) Loss of control (including partial or temporary) regardless of

cause
G) Go-around producing a hazardous or potentially hazardous

situation
H) Unintentional significant deviation from airspeed, intended

track or altitude (more than 300 ft) regardless of cause.
X) Operation of any primary warning system associated with

manoeuvring the aircraft e.g. config warning, stall warning
(stick-shaker), over-speed warning etc.’
A search of the CAA MOR database was conducted for the period 1 Oct 2002
to 10 Oct 2007 for “Stick Shaker” and “Stall Warning” events. This search
returned three occurrences submitted by this operator including the incident
under investigation.
20
A search of the operator’s ASR database for similar events returned 11 reports
in the twelve months before the incident to G-THOF.
In the twelve months following the G-THOF incident, 18 ASRs were received,
9 of which were reported by the operator as MORs. Events regarded as false,
that cause no crew difficulties, are not required to be reported to the CAA as
an MOR and this accounts for the difference in the numbers.
The CAA commented that the Flight Operations Inspectorate routinely

monitor operators’ ASR processes to ensure incidents are reported and tracked.
Sample checks of ASRs are carried out during audits to ensure that MOR
reporting requirements are satisfied.
1.18 Additional information
1.18.1 Similar events
In June 2007 a Boeing 737-300 had an uncommanded and unrecognised

autothrottle disconnect during the initial stages of an approach to Belfast
Aldergrove Airport. The disconnect occurred with the thrust levers at idle
and the aircraft decelerated below its commanded speed of 170 kt. At 112 kt
with a 16° nose-up attitude the crew advanced the engines to 96% N1 and a
rapid change of pitch to 22° ensued. The aircraft lost some 300 feet before
recovering to manoeuvre speed and level flight.
On 22 October 2007, G-THOF was subject to a similar autothrottle

disconnect event. In this case the autothrottle warning was shown as
active for 31 seconds. No manual disconnect was recorded at the time the
warning stopped, although a manual disconnect was recorded approximately
10 seconds later. During the event, speed decreased to a minimum of
128 KIAS. The crew on this flight applied approximately 75%N1 and the
aircraft recovered without the pitch attitude exceeding 8° nose-up. They
also used two small applications of forward trim.
1.18.2 EASA CS-25 certification requirements
The original certification basis for the B737-300 was Federal Aviation Regulation
(FAR) 25 which did not require an indication for autothrottle disconnect.
Current requirements also do not require any aural indication of autothrottle/

autothrust disengagement. Certification Standard (large aeroplanes) CS-25
Amendment 4, effective 27 December 2007, contained a revision to CS 25.1329
21
for Flight Guidance Systems as proposed by Notice of Proposed Amendment
(NPA) 18/2006. It included requirements for an autopilot and autothrust
disengagement caution:
‘(j) Following disengagement of the autopilot, a warning (visual

and aural) must be provided to each pilot and be timely and
distinct from all other cockpit warnings.
(k) Following disengagement of the autothrust function, a caution

must be provided to each pilot.’
Definitions of ‘Warning’ and ‘Caution’ are in AMC 25.1322 ‘Alerting

Systems’:
‘Warning: Immediate recognition and corrective or compensatory

action by the crew is required.
Caution: Immediate crew awareness is required and subsequent

crew action will be required.’
CS-25 AMC 25.1329 included some explanatory text:
‘8.3.2 Autothrust Disengagement
Autothrust disengagement should not cause any unsafe condition

(e.g., pitch attitude, pitching moment, or significant thrust transient),
to show compliance with CS 25.1329(d), and the disengagement
should not preclude, inhibit, or interfere with timely thrust changes
for go-around, landing, or other manoeuvres requiring manual
thrust changes. The autothrust normally should be designed to
preclude inadvertent disengagement during activation of autothrust
modes of operation. Following disengagement of the autothrust
function, positive indication of disengagement should include at
least a visual flight crew alert and deletion of autothrust ‘engaged’
status annunciations (to show compliance with CS 25.1329(k)).
For automatic disengagement, visual indications should persist
until cancelled by flight crew action. For manual disengagement,
if an aural is provided, visual indications should persist for some
minimum period. If an aural is not provided, the visual indications
should persist until cancelled by flight crew action. For aural
indication, if provided, an aural alert of sufficient duration and
volume should be provided to assure that the flight crew has been
22
alerted that disengagement has occurred. An extended cycle of an
aural alert is not acceptable following disengagement if such an
alert can significantly interfere with flight crew coordination or
radio communication. Disengagement of the autothrust function is
considered a Caution alert.’
1.18.3 UK CAA Paper 2004/10 Flight Crew Reliance on Automation
In 2004 the CAA published a paper relating to the use of automation. This paper
reviewed literature related to flight deck automation, and in Section 2.2.1 the
paper states:
‘Detection of automation failures was poor under constant

reliability automation, even following a catastrophic failure.
However, monitoring was efficient under variable-reliability
automation. These effects do not significantly alter following
training.’
The CAA had been funding studies by the Flight Operations Research Centre
of Excellence (FORCE) at Cranfield University into flight crew performance
and interaction with automated flight decks. These studies are being concluded
as responsibility in this area has passed to the EASA. The work that has been
completed is expected to be finalised and published as a CAA research paper
by 2010 timeframe and should produce further insights in an increasingly
important field.
1.18.4 Flight crew training and manuals
Both pilots attended a Type Rating Training Organisation (TRTO) and completed
the approved syllabus. They had subsequently completed simulator renewal
checks. During the TRTO course the pilots were taught approach to stall
recovery and unusual attitudes on the Boeing 737 Full Flight Simulator (FFS).
1.18.4.1 Operations Manual Part B
The initial actions in the operators Boeing 737-300/500 Operations Manual

Part B - Go-around procedure is set out in Appendix B1. The drill requires the
PM to:
‘Verify that the thrust is sufficient for the go-around or adjust as

needed’
23
1.18.4.2 The operator’s Quick Reference Handbook (QRH) procedure for approach to
stall recovery.
This drill is to be accomplished at the first indication of stall buffet or stick‑shaker.

It requires the pilot flying to select maximum thrust, not to change configuration
(flap or landing gear). The drill does not mention the use of pitch trim. The
operator commented that:
‘All of these techniques (go-around, approach to stall recovery, stall

recovery, upset recovery) - come direct from the manufacturer and
are included within the QRH (non-normal manoeuvres section) and
are not modified.’
The complete QRH section is set out in Appendix B2.
1.18.4.3 The operator’s QRH procedure for upset recovery
The operator’s QRH procedure for upset recovery is set out in Appendix B3.
The ‘upset recovery’ QRH defines an upset as unintentionally exceeding a pitch
angle greater than 25º nose-up. The procedure states:
‘If needed, use pitch trim sparingly.’
The procedure continues:
‘These techniques assume that the aircraft is not stalled. A stall

condition can exist at any attitude and may be recognised by
continuous stick-shaker activation accompanied by one or more of
the following:
●● Buffeting which could be heavy at times
●● Lack of pitch authority and/or roll control
●● Inability to arrest descent rate
If the airplane is stalled, recovery from the stall must be accomplished

first by applying and maintaining nose down elevator until stall
recovery is complete and stick-shaker activation ceases.’
The upset recovery procedure requires the pilot flying to carry out certain actions
including:
24
●● Apply as much as full nose-down elevator
●● Apply appropriate nose down stabilizer trim
●● Reduce thrust’
1.18.4.4 Flight Crew Training Manual11
The manufacturer produced a type-specific, operator generic training manual

that the operator included within its document package which is disseminated
to all of its pilots.
The FCTM preface included the caveat that in the event of a conflict, the
procedures published in the flight crew operations manual take precedence over
information presented in the FCTM.
The FCTM defines the stall recovery in two parts. The approach to stall recovery
is intended to be accomplished at the first indication of the impending stall –
either the stick-shaker or pre-stall buffet; effectively it is a manoeuvre to be
used when the aircraft has not yet stalled. Once the aircraft has stalled, the
‘stall recovery’ is a more aggressive manoeuvre intended to return the aircraft to
controlled flight as quickly as possible. The FCTM states:
‘A stall must not be confused with the stall warning that alerts
the pilot to an approaching stall. Recovery from an approach
to a stall is not the same as recovery from an actual stall. An
approach to a stall is a controlled flight manoeuvre, a stall is an
out of control but recoverable condition’
The manufacturer’s FCTM ‘approach to stall recovery’ includes the following

advice:
‘To assist in pitch control, add more nose down trim as the thrust
increases’
The ‘stall recovery’ also refers to the use of trim:
‘the use of some nose-down stabalizer trim should provide

sufficient elevator control to produce a nose down pitch rate…
care must be taken to avoid using too much trim’
11 The FCTM is a manufacturer produced document. It includes guidance on operating techniques that are not tailored
to a specific airline.
25
FCTM advice for upset recovery includes a section similar to the operator’s
QRH regarding initially recovering from the stall. However, FCTM advice also
states:
‘Under certain conditions, it may be necessary to reduce some

thrust in order to prevent the angle of attack from continuing to
increase.’
The FCTM goes on to say:
‘In a situation where the airplane pitch attitude is unintentionally

more than 25 degrees nose high and increasing, the airspeed
is decreasing rapidly. As airspeed decreases, the pilot’s ability
to maneuver the airplane also decreases. If the stabilizer trim
setting is nose up, as for slow-speed flight, it partially reduces the
nose-down authority of the elevator. Further complicating this
situation, as the airspeed decreases, the pilot could intuitively
make a large thrust increase. This causes an additional pitch
up. At full thrust settings and very low airspeeds, the elevator,
working in opposition to the stabilizer, has limited control to
reduce the pitch attitude.’
The full FCTM advice on stall recovery and upset recovery is set out in Appendix
B4‑7.
1.18.4.5 ‘Airplane’ upset recovery training aid
The manufacturer, in conjunction with other manufacturers, pilot associations,

training organisations, operators, government agencies and suppliers, produces
an “airplane upset recovery training aid”. This training aid is intended to be a
comprehensive package that airlines can present to flight crews in a combination
of classroom and simulator sessions. It provides a structured baseline tool
to be incorporated into existing programs or to be customised to individual
operators’ needs. The basic package includes a DVD with two short videos
and a powerpoint presentation. Extracts from this training aid are presented at
Appendix C. The operator commented that:
‘With regard to the Boeing CD “Upset Recovery” training aid

we believe that the contents of the CD have not been verified to
be the correct technique for all events and does not form part of
their officially reviewed training material, therefore it has not
been issued as a Thomson Airways training document. Pilots
26
are trained according to requirements in early recognition and
counter measures on approaching stall in all configurations and
recovery from full stall or after activation of stall warning device
in all phases of flight, in all configurations.’
1.18.5 Pitch power couple
Turbofan aircraft with underslung engines will tend to pitch nose-up as the
thrust is increased as the thrust line is below the centre of gravity of the aircraft.
Conversely, as power is reduced the aircraft will pitch nose-down.
Maneuvering in Pitch
Lift
Wing-body
moment
Drag
Tail lift
Tail distance Thrust
Wing distance Engine
distance
Weight
Total
(Moment) + (Moment) + (Moment) + (Moment) = pitching
Tail Lift Thrust Wing-body moment
(Moment) Total
Tail
lift
Tail
* distance + Wing
lift
Wing
* distance + Thrust
*
Engine
distance
+ Wing-body = pitching
moment
Picture courtesy Boeing Upset Recovery Training Aid.
Figure 3-B.47
The application of go-around thrust on the Boeing 737-300 will cause a
nose‑up pitching moment. This is counteracted by applying nose-down
elevator. The stabiliser on the Boeing 737, as is common on large transport
aircraft, is a moveable surface, the position of which is selected by the trim
system. When the autopilot is engaged it automatically trims the aircraft for
the present speed using the stabiliser. As the aircraft decelerates, the autopilot
applies more nose-up trim to keep the aircraft in a trimmed state. However
as the elevator is attached to the trailing edge of the stabiliser, this increasing
trim effectively reduces the amount of pitch authority available. The size of
the elevator is carefully calculated to be as small as possible while still giving
sufficient control authority at go-around thrust.
27
The Boeing 737-300 engines are not thrust limited; if a pilot selects the thrust
levers to the forward stop the engine will produce the maximum thrust available
given the altitude and temperature. On days colder than standard12 this could
be in excess of the maximum rated thrust of the engine. The N1 to generate
rated thrust is automatically calculated and displayed on the N1 gauges; during
a go-around pilots are expected to select that N1.
As the thrust increases during a go-around a large nose-up pitch couple is

produced. This needs to be counteracted by nose-down elevator deflection
to maintain level flight. If the elevator authority is reduced because of the
amount of nose-up trim applied and the go-around thrust is greater than the
rated thrust there may be insufficient elevator authority remaining to counteract
the nose-up pitching moment.
12 International Standard Atmosphere Sea Level Temperature +15C Pressure 1013 millibars.
28
2 Analysis
2.1 Introduction
This serious incident occurred during a routine ILS approach at night in

IMC. The investigation focussed on three areas: the incident itself, the
operator’s response and a review of the efficacy of the autothrottle warning
on the Boeing 737.
2.1.1 Autothrottle disengagement
The autothrottle disengaged during the approach. There are six possible
conditions where the autothrottle system would disengage as detailed in
paragraph 1.6.3 (page 10). The QAR information showed that there was no
split between the thrust levers. The aircraft had not touched down. The crew
did not report disconnecting the autothrottle at this stage and the QAR did not
record a manual disconnect. The only condition that remained was an internal
fault within the autothrottle computer.
The autothrottle system continued to operate normally on the subsequent

approach and during flights following the incident. It was comprehensively
tested on the aircraft and no faults were found. After removal from the aircraft
the autothrottle computer did not pass the ATP. However, it was demonstrated
that this fault would not necessarily have caused an in-flight disengagement
of the autothrottle system.
An uncommanded disconnect, however, is not an unusual event. The analysis

conducted on the OFDM data indicated that 0.4% of sectors flown had an
autothrottle disconnect without any manual disconnect switch activity. The
maintenance history on G-THOF’s autothrottle system recorded five defects
in the nine months prior to this incident; in each case the system was checked
with no fault found.
The autothrottle is a single-channel system. It is not required for flight

operations on the Boeing 737-300 and the aircraft can be flown without
the use of the autothrottle. The crew did not respond to the autothrottle’s
disconnection either because the autothrottle warning did not work, or if
it did work, the crew did not notice it. The recorded flight data shows the
autothrottle warning as active; function testing could find no fault with the
warning system. The subsequent event on 22 October and the other event
discussed in 1.18 (page 21) showed that there may be a wider issue of the
warning system not alerting flight crews.
29
The data gathering exercise carried out by the AAIB identified that 0.3% of
Boeing 737 sectors had a combination of no recorded manual disconnection
of the autothrottle and a warning lasting over nine seconds. The purpose of
this limited study was to examine anecdotal and other information in order
to establish the efficacy or otherwise of the autothrottle warning. The data
gathering, although limited in scope, returned results which appear to justify
the need for a much larger study of the autothrottle warning system on the
Boeing 737. Therefore:
It is recommended that Boeing, in conjunction with the Federal

Aviation Administration, conduct a study of the efficacy of
the Boeing 737-300/400/500 autothrottle warning system
and if necessary take steps to improve crew alerting. (Safety
Recommendation 2009-043)
2.1.2 Disengagement alerting requirements
The autothrottle warning on G-THOF was typical of its era. Many later
generation aircraft incorporate an autothrottle warning, including an audio
alert, into an Engine Indication and Crew Alerting System (EICAS). Aircraft
in general and automation technology specifically has advanced rapidly in
reliability. Pilots familiar with operating older aircraft, which had more
variable reliability, are nearing the end of their careers and there is a generation
of pilots whose only experience is of operating aircraft with highly reliable
automated systems. With this increasing reliability there is concern about
flight crew encountering the issues outlined by the CAA paper 10/2004: that
of a normally reliable system failing.
An illuminated warning should be distinct and gain the crew’s attention. The
use of a multifunction caption whereby it flashes one colour (amber) for one
reason and then another colour (red) for a different reason may reduce the crew’s
awareness of the caption. The autothrottle warning on the Boeing 737‑300
flashes amber routinely for extended periods during the approach phase of
flight. It is likely that flight crews are subconsciously filtering out what is
perceived as a nuisance message. This combined with the general high levels
of reliability of modern automation could lead to a lack of awareness of
autoflight modes.
The EASA Certification Standards (CS) set the standards for new aircraft
designs. The current CS-25.1329 requirements treat autothrottle disconnection
as a caution requiring immediate crew awareness but not immediate crew
action as set out in 1.18.2. The regulators rationale for this was that although
30
the flight crew needs to be made continually aware of the autothrust system
status, including disengagement, normal disengagement of the autothrust
would not require immediate thrust control changes. A less specific indication,
rather than a warning, is therefore allowable. The system fitted to the 737-300
would meet the current CS for a newly designed aircraft.
This incident, along with those in 1.18.1 (page 21) and the OFDM study highlight
the limitations of the Boeing 737-300 alerting system despite it meeting current
certification standards. Therefore:
It is recommended that The European Aviation Safety Agency

review the requirements of Certification Standard 25 to ensure that
the disengagement of autoflight controls, including autothrottle,
is suitably alerted to flight crews. (Safety Recommendation
2009‑044)
2.2 Crew reactions
Both pilots were qualified to operate the flight in accordance with company and
national requirements. There is no reason to believe the crew’s performance
was degraded by fatigue or for medical reasons. There is no evidence that the
aircraft or flight crew were affected by any external factor during the approach.
There also appears to be no technical failure of the autothrottle warnings or
any of the aircraft’s engine or airspeed indications.
The aircraft was on target to be configured for landing by 1,400 ft aal, within
the operator’s requirement to be stabilised1 by 1,000 ft aal, and there is no
evidence that the crew were rushing the approach.
The use of flap 40 was not common for crews operating out of the Bournemouth
base. The F/O needed to increase the brightness of his map light to ensure he
was below the flap limiting speed for flap 40 before asking for it to be selected.
This activity surrounding the selection of flap 40 was the only point where the
reported flight deck activity appeared to vary from what was usual.
Regardless of the status of the autothrottle and its warnings, both pilots appear
to have been distracted at a critical phase of flight. This lack of effective
monitoring of automated systems allowed the aircraft to enter a low energy
state following disconnection of the autothrottle.
1 Landing gear down, flaps in the landing position, on approach speed and engines spooled up.
31
The CAA funded studies by the Flight Operations Research Centre of
Excellence (FORCE) at Cranfield University into flight crew performance
and interaction with automated flight decks (section 1.18.3, page 23). It has
completed its study phase and the report into its findings is expected to be
available in the public domain in due course.
2.3 Aircraft pitch excursion
The go-around decision.
A go-around is a routine flight manoeuvre and flight crew are encouraged to

go-around if an approach becomes unstable. When the captain reportedly
took control, the aircraft was approximately 20 kt below the approach speed,
decelerating with the thrust at or close to idle and fully configured for landing.
The approach was unstable and action was required to recover the aircraft.
The aircraft was on the verge of a stall warning and had the stall warning been
active the commander should have carried out the ‘approach to stall recovery’
actions as a priority over the go-around. The evidence suggests that, at the
time the commander made the decision to go-around, the stall warning was
not active.
Go-around execution
The execution of a go-around involves both crew members carrying out specific
tasks. The commander advanced the thrust levers to the forward limit of their
travel. He called “go-around, flap 15 check thrust” and, in accordance with
his training, applied forward column to counteract the expected nose-up pitch.
When the commander took control, the F/O became the PM. It is the role of
the PM to position the flap lever and adjust the N1.
Immediately after the go-around was commenced the stall warning activated.
At this point responding to the stall warning became the highest priority, but
the commander initiated a large increase in thrust as required by both the
‘approach to stall recovery’ and the ‘go-around’ procedures. The ‘approach
to stall recovery’ calls for maximum thrust as opposed to ‘sufficient’ thrust
for the go-around. The crew had maximum thrust selected. The commander
attempted to control the pitch using the control column but, as shown by the
manufacturer’s analysis, there was insufficient elevator authority available to
counteract the thrust/pitch couple in the as-trimmed condition.
As the rate of pitch-up decayed briefly around 22º nose-up, the crew had
maximum thrust selected. They had not yet changed configuration and were
32
adhering to the ‘approach to stall recovery’ drill. It is therefore probable that,
whatever decision the commander made at the point he took control, the aircraft
was going to pitch-up to at least 22º. The only way to avoid this would have
been either to select a lower thrust setting or to trim forward during the thrust
application.
Thrust
The operator’s ‘approach to stall recovery’ QRH drill requires the PF to

‘Advance thrust levers to maximum thrust’. Maximum is defined by the
Oxford English dictionary as the ‘greatest amount possible.’ The ‘approach
to stall recovery’ drill has to be an instinctive and straightforward drill and
the technique of advancing the thrust levers to maximum meets these criteria.
Unlike a go-around which can occur throughout the approach speed range, an
‘approach to stall recovery’ is only likely to occur with the aircraft in a low
energy state. This drill makes no allowance for the pitch couple produced
by the engines and the subsequent risk of the engine thrust overcoming the
available elevator authority. A requirement to select less than maximum thrust
may reduce the risk of an upset similar to the one in this incident but could
result in too little thrust being used, increasing the risk of a stall. A technique
that allows the use of maximum thrust while protecting the aircraft from the
pitch couple is required.
Safety action taken by the manufacturer
Boeing stated that they:
‘recognise that there may be confusion with the wording of

“maximum thrust” in the Approach to Stall Recovery procedure.
The engine control system installed on the 737-600-900 (NG’s)
are set so that in normal operation, selection of full forward thrust
levers will result in a very small overboost and prevent exceedance
of N1 and N2 rotor speeds. However the older control system of
the 737 300-500 (Classics) will allow the engines to overboost if
maximum thrust lever movement is selected. As such we plan to
revise the definition of ‘maximum thrust’ for the Classic models
to mean go-around or takeoff thrust unless ground contact is
imminent.’
33
Trimming
Trimming is not mentioned in either the operator’s QRH ‘approach to stall

recovery’ or the ‘go-around’ drills. In this incident the commander did not
consider trimming the aircraft.
The FCTM includes advice on trimming in both the ‘approach to stall recovery’
and the ‘stall recovery’ sections. This was available to both pilots but it is
a generic document and not tailored to a particular airline. The company
operations manuals, including the QRH, were the primary documents referred
to by the flight crew for aircraft operation. Pilots are required to follow drills
and procedures outlined in the operator’s QRH and deviation from these drills
is discouraged. The FCTM itself states that in the event of a conflict, the
procedures published in the flight crew operations manual take precedence
over information presented in the FCTM. Therefore;
It is recommended that Boeing clarify the wording of the

Boeing 737 300-500 approach to stall recovery Quick Reference
Handbook Non-normal Manoeuvres to ensure that pilots are aware
that trimming forward may be required to enhance pitch control
authority. (Safety Recommendation 2009-045)
Flap retraction
Approximately 11 seconds after the thrust increased, the flaps were retracted
to flap 15 even though the stall recovery checklist specifically states not to
change the configuration. The manufacturer’s opinion is that on this occasion
it made little difference to the outcome.
Upset recovery - general
Once the nose-up pitch angle exceeded 25º, the aircraft was outside the
normal flying regime and an upset, as defined by the QRH, had occurred. At
this stage, regardless of their initial actions, the crew needed to use ‘upset
recovery’ techniques.
Upset recovery is defined in both the QRH and the FCTM. There is a caveat
included:
‘if the airplane is stalled, recovery from the stall must be

accomplished first by applying and maintaining nose down
elevator until stall recovery is complete and stick-shaker activation
ceases.’
34
The definition of a stall:
‘A stall condition..…may be recognized by continuous stick-shaker

activation accompanied by one or more of the following:
• Buffeting which could be heavy at times

• Lack of pitch authority and/or roll control
• Inability to arrest descent rate…
Under certain conditions, it may be necessary to reduce some thrust

in order to prevent the angle of attack from continuing to increase.’
By this definition the aircraft was recognisably stalled and the crew needed to
effect a stall recovery.
Upset recovery - nose-high specific techniques
The first requirement of upset recovery is to recognise that the aircraft is in

an upset condition as defined by the FCTM. For nose-high recovery the PF
should apply as much as full nose-down elevator, appropriate stabiliser trim
and reduce thrust.
Apart from holding the control column fully forward, the flight crew made
no other pitch control actions throughout the 44º nose-up excursion until the
aircraft had stalled and the nose had dropped towards the horizon. At this
stage the thrust was reduced to go-around thrust. This thrust reduction allowed
sufficient control authority to recover the aircraft and all three upset training
documents include reducing thrust as one of their techniques. The techniques
outlined in the manufacturer’s FCTM, the upset recovery training aid and
the operator’s operations manual deal with this type of upset. Had these
actions been carried out the aircraft was unlikely to have reached the extreme
angles and speeds encountered and would have recovered to controlled flight
sooner.
2.4 ASR
Company response
The initial reviewer decided not to file the ASR as an MOR. Although the
ASR included a number of factors such that an MOR would have been
appropriate, he decided to get more information from OFDM. This decision
to seek additional information from the OFDM system was an optional
35
company procedure. However once the event passed out of sight of the safety
department there was no process in place to ensure a timely follow-up of the
incident. Although the company’s system would have provided a backstop of
30 days for the incident to be highlighted, there was a risk, borne out by this
incident, that the processing of safety-critical information could be delayed
between the two office locations.
The OFDM review policy was biased to dealing with ‘de-identified’ events
where only a pilot representative reviewed the data. At the time of the incident
the pilot representative workload did not allow for timely review of the data
and when the 72 hour time limit for MOR submission was reached, there was
no alert. When the next pilot representative came on duty 11 days after the
event it was identified and appropriate action was taken.
There was no requirement for this incident to be de-identified. The commander

had already filed an ASR and spoken with the safety department about the
incident. The Schedule J, Appendix 2 agreement allowed for events that had
been identified by ASR to be reviewed by an appropriate person and set out
a process to facilitate this. Had the Schedule J, Appendix 2 process been
followed it is likely that the severity of the event would have been captured
within the MOR scheme time limits.
There were potential safety implications of the aircraft remaining in service

for an extended period following this incident. The aircraft had operated
outside the normal envelope although no damage had occurred. The flight
crew operated additional sectors whilst expecting to be contacted regarding
the incident. Had the severity of the incident been realised by the operator, the
crew would have been suspended pending an investigation by the company.
Timeliness of OFDM analysis
OFDM as a mandatory activity is a relatively recent requirement. CAP 739

talks of batching data but also of ‘ideally’ reviewing data before the next flight.
OFDM systems are valuable in the identification of trends in pilot handling
and use this information to devise training programmes which improve
standards across operators and the industry generally. A balance has to be
struck between protecting the rights of the individual against unreasonable
monitoring and the need to expedite data processing. Where an event has
occurred and the pilots themselves have identified this, a rapid review of the
data would seem appropriate.
36
2.5 The MOR Scheme
The ASR filed to the company by the captain stated ‘…..Speed decreased to
110 ……stick-shaker during initial part of G/A….’ Following the guidance
of CAP 382 this would suggest that an MOR be filed under condition
G (Go‑around producing a hazardous or potentially hazardous situation.) The
loss of at least 25 kt below the target speed of 135 kt should be reported under
section H (Unintentional significant deviation from airspeed). The activation
of the stick-shaker is covered by section X (Operation of any primary warning
system associated with manoeuvring the aircraft eg config warning, stall
warning (stick-shaker), over-speed warning etc.)
The search of the CAA MOR database showed that the operator had only
filed two previous reports under the scheme for stick-shaker events in the
five years prior to this incident. This contrasts with 11 ASRs received in the
twelve months before this incident. The purpose of the MOR scheme is to
collate statistical information to be used by regulators, designers and accident
investigators. Significant under-reporting of events reduces the availability of
safety trend analysis information to assist the industry to improve its safety
standards. The step change in reporting levels from the operator following
this incident is indicative that the company had reviewed its MOR submission
policy throughout its organisation.
Safety action taken by the operator
Following the incident the operator conducted an internal review of the incident
and their response to it. The following actions have been taken or are planned:
Pilot training
Additional training in general handling, specifically stalling and upset recovery

has been added to the next set of simulator training. All pilots were due to
have had this training by the end of June 2008. This stall recovery training
included the possibility of using nose-down pitch trim during the recovery.
The operator is trialling an eye position monitor for use during simulator
training to determine where pilots are looking during critical phases of flight.
Pilot information
The operator issued a number of operational notices to pilots highlighting the

event and the need for vigilance in monitoring autoflight modes.
37
ASR and data handling
The operator moved its OFDM office to a secure office within the safety
management team. This improved communication and integrated OFDM
within the safety department while still protecting the flight crew from
unreasonable monitoring.
The company also increased its establishment to two full-time data analysts
and now has a pilot representative available in the data office daily.
The operator intends further investment to introduce an OFDM system that

allows 24 hour access to certain functions by maintenance control to ensure
continued airworthiness of aircraft involved in incidents. Its intention is
to standardise its fleet on a telephone data link system such as was already
installed on G-THOF at the time of this event.
General
The operator notified the AAIB of their intent of carrying out a general audit
to review operational standards across all fleets by the end of 2008.
Operator comment
The operator commented that the commander should have telephoned the
duty pilot manager as soon as he landed following the incident. While, in
their view, the ASR system was robust and quick enough for routine events
it is their belief that an event as serious as this should have been notified to
the management by telephone immediately. The operator also commented
that the initial ASR submitted by the commander did not accurately report
the speed loss and did not address the degradation of pitch control. In the
operator’s opinion, had either of the above procedures been followed, it would
have prevented the issues surrounding the ASR processing.
38
3 Conclusions
3.1 Findings
Flight operations
1. The flight crew were properly licensed and qualified to conduct the flight.
They were medically fit and there was no evidence of fatigue. Their
training was in accordance with national regulations and the operator’s
requirements.
2. The aircraft was certified, equipped and maintained in accordance with

existing regulations and approved procedures. At the time of the incident
there were no recorded defects that might have contributed to the event.
3. The mass and centre of gravity of the aircraft were within the prescribed
limits.
4. The flight had been routine until the approach at Bournemouth.
5. The autothrottle retarded the thrust to idle in response to crew inputs.
6. The autothrottle disengaged for undetermined reasons.
7. No significant fault could be found with the autothrottle warning or

associated systems.
8. The disengagement of the autothrottle was not recognised by the crew.
9. The aircraft’s Indicated Airspeed (IAS) decayed in line with crew

expectations for an idle thrust approach and this constant deceleration
approach masked the disengagement of the autothrottle.
10. No external factors degraded the flight crew’s ability to monitor the
aircraft.
11. The pilots were distracted at a critical phase of flight and did not properly
monitor the airspeed.
12. The aircraft stalled and descended in a nose-up attitude and slowed to a
minimum airspeed of 82 kt.
39
13. The thrust levers remained at full thrust for 26 seconds and N1 exceeded
the target N1 for 31 seconds.
14. The flaps retraction did not materially affect the event.
15. The stall recovery techniques recommended in the manufacturer’s Flight

Crew Training Manual (FCTM) were not fully applied.
16. Forward trim was not used during the stall recovery.
17. A reduction in thrust lever position to a go-around (GA) thrust setting

occurred 40 seconds after the go-around was initiated, which allowed
sufficient nose-down elevator authority to control the pitch-up couple.
18. The speeds and pitch angles were outside the flight test envelope and
outside the validated flight modelling envelope.
Flight procedures
1. The wording of the go-around drill in the Quick Reference Handbook

(QRH) has the potential to prejudice pilots away from reducing thrust to
match the required go-around thrust.
2. The ‘go-around’ drill and ‘approach to stall’ drill in the QRH do not
mention trimming the aircraft.
3. The upset recovery techniques outlined in the QRH, FCTM and the
manufacturer’s training aid are effective and would have resulted in earlier
recovery of the aircraft.
Safety management
1. The Air Safety Report (ASR) as filed by the commander did not depict the
event accurately.
2. The ASR was received at the operator’s offices the morning after the event
but was not initially filed as an Mandatory Occurrence Report (MOR).
3. The ASR was passed to the Operational Flight Data Monitoring (OFDM)
analyst on the day after the event and was reviewed that day when the
OFDM analyst flagged the event for a pilot representative.
40
4. The flight data was not viewed by a pilot representative until 11 days
after the event. This delay in reviewing the data resulted in the loss of
information of value to the investigation.
5. The delay in reviewing the data allowed both the aircraft and the crew to
continue operating without the incident being reviewed.
6. There was no requirement in the company OFDM agreement to de‑identify

the data and the data could have been reviewed on the day after the
event.
7. The operator has undertaken significant changes in their OFDM and safety
management system following this event.
3.2 Causal factors
The investigation identified the following causal factors:
1. The aircraft decelerated during an instrument approach, to an airspeed

significantly below the commanded speed, with the engines at idle thrust.
Despite the application of full thrust, the aircraft stalled, after which the
appropriate recovery actions were not followed.
2. The trimmed position of the stabiliser, combined with the selection of

maximum thrust, overwhelmed the available elevator authority.
3.3 Contributory factors
The investigation identified the following contributory factors:
1. The autothrottle warning system on the Boeing 737-300, although working

as designed, did not alert the crew to the disengagement of the autothrottle
system.
2. The flight crew did not recognise the disengagement of the autothrottle
system and allowed the airspeed to decrease 20 kt below Vref before
recovery was initiated.
41
4 Safety Recommendations
4.1 Safety Recommendation 2009-043: It is recommended that Boeing, in

conjunction with the Federal Aviation Administration, conduct a study of the
efficacy of the Boeing 737-300/400/500 autothrottle warning and if necessary
take steps to improve crew alerting.
4.2 Safety Recommendation 2009-044: It is recommended that The European

Aviation Safety Agency review the requirements of Certification Standard 25
to ensure that the disengagement of autoflight controls including autothrottle is
suitably alerted to flightcrews.
4.3 Safety Recommendation 2009-045: It is recommended that Boeing clarify

the wording of the approach to stall recovery Quick Reference Handbook
Non‑normal Manoeuvres to ensure that pilots are aware that trimming forward
may be required to enhance pitch control authority.
K Conradi
Principal Inspector of Air Accidents
April 2009
42
Appendix A
Date Reported defect/Actions taken

(Technical Log entries paraphrases for clarity)
During take-off when pressing either left or right TOGA switches,

autothrottle would not engage and no Flight Director (FD) modes
20/9/07
displayed. Attempted to engage after 80 kts, and FD worked al-
though without pitch modes but autothrottle demanded full power
Autothrottle BITE attempted – unable to enter BITE mode. Cir-
cuit Breaker (CB) panels checks and ‘AFDS MCP DC’ CB found
tripped. CB reset and BITE carried out all SATIS…..Unable to
fault system, considered serviceable
During descent on both sectors autothrottle struggles to keep N1

30/8/07
symmetrical resulting in thrust lever stagger (up to 20% )
Autothrottle system BITE and troubleshooting carried out, unable
to fault system, all SATIS. System considered serviceable
27/3/07 On TOGA selection, autothrottle commanded full go-around

thrust, then trimmed back to reduce T/O thrust
Autothrottle BITE carried out, unable to fault on ground
Carry out rigging check and adjustment of autothrottle power

12/3/07
lever synchro.
Thrust lever stagger (up to one complete throttle knob width) when
2/1/07
setting T/O thrust causing asymmetry problem
Autothrottle rigging and range checked SATIS, unable to fault
system
Further reports of 2% N1 throttle stagger on 4/1/07 – Range and
rigging checks carried out – unable to fault system
Fault History
A-1
Appendix B
Normal Procedures -
Amplified Procedures
B737-300/500 Operations Manual
Go–Around and Missed Approach Procedure

Pilot Flying Pilot Monitoring
At the same time: Position the FLAP lever to 15 and
monitor flap retraction
• push the TO/GA switch
• call "G/A, FLAPS 15, CHECK
THRUST"
Verify:
• the rotation to go–around attitude
• that the thrust increases
Verify that the thrust is sufficient for
the go-around or adjust as needed.
Verify a positive rate of climb on the Verify a positive rate of climb on the
altimeter and call "GEAR UP". altimeter and call "POSITIVE RATE".
Set the landing gear lever to UP.
Verify that the missed approach altitude
is set.
Above 400 feet, select LNAV or select Observe mode annunciation.
HDG SEL as appropriate.
If needed, call "TUNE NAV RADIOS FOR Tune the navigation radios as directed.
MISSED APPROACH".
Verify that the missed approach route is tracked.
At acceleration height, verify Set the FLAP lever as directed.
IAS/MACH Display is blank or call Monitor flaps and slats retraction.
"SET UP SPEED".
Call "FLAPS ___" according to the flap
retraction schedule.
After flap retraction to the planned flap
setting, select LVL CHG. VNAV may
be selected if the flaps are up.
Verify that climb thrust is set.
Verify that the missed approach altitude is captured.
Set the landing gear lever to OFF after
landing gear retraction is complete.
Set the engine start switches as needed.
Call "AFTER TAKEOFF Do the AFTER TAKEOFF checklist.
CHECKLIST"
Copyright © The Boeing Company. See title page for details.

April 01, 2007 D6-27370-804-BRI NP.21.43
Operator Procedures
B-1
Appendix B
B737-300/-500
Maneuvers Chapter MAN

Non-Normal Maneuvers Section 1
MAN.1 Maneuvers-Non-Normal Maneuvers
Approach to Stall Recovery
The following is immediately accomplished at the first indication of stall buffet or
stick shaker.
• Advance thrust levers to • Verify maximum thrust.
maximum thrust. • Monitor altitude and airspeed.
• Smoothly adjusting pitch • Call out any trend toward
attitude* to avoid ground terrain contact.
contact or obstacles.
• Level the wings (do not change
flaps or landing gear
configuration).
• Retract the speedbrakes.
When ground contact is no longer a
factor:
• Adjust pitch attitude to
accelerate while minimizing
altitude loss.
• Return to speed appropriate for
the configuration.
Note: *At high altitudes it may be necessary to decrease pitch attitude below the
horizon to achieve acceleration.
April 01, 2007 MAN.1.1
B-2
Appendix B
Maneuvers -
Non-Normal Maneuvers
B737-300/-500
Continued from previous page
If maneuvering is required, disengage
the autopilot and autothrottle.
Smoothly adjust pitch and thrust to
satisfy the RA command. Follow the
planned lateral flight path unless visual
contact with the conflicting traffic
requires other action.
Attempt to establish visual contact. Call out any conflicting traffic.
For a climb RA in landing configuration:

Disengage the autopilot and Verify maximum thrust set. Position flap
autothrottle. Advance thrust levers lever to 15 detent.
forward to ensure maximum thrust is
attained and call for FLAPS 15.
Smoothly adjust pitch to satisfy the RA
command. Follow the planned lateral
flight path unless visual contact with
the conflicting traffic requires other
action.
After positive rate of climb established, Position gear lever up.
call for GEAR UP.
Attempt to establish visual contact. Call out any conflicting traffic.
Upset Recovery
An upset can generally be defined as unintentionally exceeding the following
conditions:
• Pitch attitude greater than 25 degrees nose up, or
• Pitch attitude greater than 10 degrees nose down, or
• Bank angle greater than 45 degrees, or
• Within above parameters but flying at airspeeds inappropriate for the
conditions.
April 01, 2007 MAN.1.7
B-3
Appendix B
Maneuvers
737 Flight Crew Training Manual
Stall Recovery
The objective of the approach to stall recovery maneuver is to familiarize the pilot
with the stall warning and correct recovery techniques. Recovery from a fully
developed stall is discussed later in this section.
The following discussion and maneuvers are for an approach to a stall as opposed
to a fully developed stall. An approach to a stall is a controlled flight maneuver; a
stall is an out-of-control, but recoverable, condition.
Initial Conditions Approach Recovery

If ground contact is not a factor
Flaps Gear Bank Target% N1* At buffet or stick shaker:
• Apply maximum thrust
Up Up 0° 35 - 45% • Smoothly decrease pitch
15 Dn 25° 60 - 70% attitude to approximately 5°
above the horizon
30 Dn 0° 60 - 70% • Level wings
• Accelerate to maneuvering
*Approximate
speed for flap position
power settings
to achieve a • Stop descent and return to
1 kt/sec target altitude
deceleration • At altitudes above 20,000 feet,
pitch attitudes less than 5°
may be necessary to achieve
acceptable acceleration.
Maintain airplane in Note pitch Maneuver complete

trim until stick attitude at
shaker or buffet trim speed If ground contact is a factor
At buffet or stick shaker:
• Apply maximum thrust
• Smoothly adjust attitude as
needed to avoid terrain
• Level wings
• Accelerate to maneuvering
speed for flap position
• Level off at target altitude.

7.8 FCT 737 (TM) October 31, 2006
B-4
Appendix B
Maneuvers

Flaps
Do not retract flaps during the recovery. Retracting the flaps from the landing
position, especially when near the ground, causes an altitude loss during the
recovery.
Recovery
Recover from approach to a stall with one of the following recommended recovery
techniques.
Ground Contact Not a Factor
At the first indication of stall (buffet or stick shaker) smoothly apply maximum
thrust, smoothly decrease the pitch attitude to approximately 5 degrees above the
horizon and level the wings. As the engines accelerate, counteract the nose up
pitch tendency with positive forward control column pressure and nose down trim.
(At altitudes above 20,000 feet, pitch attitudes of less than 5 degrees may be
necessary to achieve acceptable acceleration.)
Accelerate to maneuvering speed and stop the rate of descent. Correct back to the
target altitude.
Ground Contact a Factor
At the first indication of stall (buffet or stick-shaker) smoothly advance the thrust
levers to maximum thrust and adjust the pitch attitude as needed to avoid the
ground. Simultaneously level the wings. Control pitch as smoothly as possible. As
the engines accelerate, the airplane nose pitches up. To assist in pitch control, add
more nose down trim as the thrust increases. Avoid abrupt control inputs that may
induce a secondary stall. Use intermittent stick shaker as the upper limit for pitch
attitude for recovery when ground contact is a factor.
When ground contact is no longer a factor, continue to adjust pitch as required to
maintain level flight or a slight climb while accelerating to maneuvering speed for
the existing flap position.
Autopilot Engaged
If an approach to a stall is encountered with the autopilot engaged, apply limit
thrust and allow the airplane to return to the normal speed. At high altitude, it may
be necessary to initiate a descent to regain maneuvering speed. If autopilot
response is not acceptable, it should be disengaged.

7.10 FCT 737 (TM) October 31, 2006
B-5
Appendix B
Maneuvers
Recovery from a Fully Developed Stall

An airplane may be stalled in any attitude (nose high, nose low, high angle of
bank) or any airspeed (turning, accelerated stall). It is not always intuitively
obvious that the airplane is stalled.
An airplane stall is characterized by any one (or a combination) of the following
conditions:
• buffeting, which could be heavy
• lack of pitch authority
• lack of roll control
• inability to arrest descent rate.
These conditions are usually accompanied by a continuous stall warning. A stall
must not be confused with the stall warning that alerts the pilot to an approaching
stall. Recovery from an approach to a stall is not the same as recovery from an
actual stall. An approach to a stall is a controlled flight maneuver; a stall is an
out-of-control, but recoverable, condition.
Note: Anytime the airplane enters a fully developed stall, the autopilot and
autothrottle should be disconnected.
To recover from a stall, angle of attack must be reduced below the stalling angle.
Nose down pitch control must be applied and maintained until the wings are
unstalled. Application of forward control column (as much as full forward may be
required) and the use of some nose-down stabilizer trim should provide sufficient
elevator control to produce a nose-down pitch rate. 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.
Under certain conditions, on airplanes with underwing-mounted engines, it may
be necessary to reduce thrust in order to prevent the angle of attack from
continuing to increase. Once the wing is unstalled, upset recovery actions may be
taken and thrust reapplied as needed.
If normal pitch control inputs do not stop an increasing pitch rate in a nose high
situation, rolling the airplane to a bank angle that starts the nose down may be
effective. Bank angles of about 45°, up to a maximum of 60°, could be needed.
Normal roll controls - up to full deflection of ailerons and spoilers - may be used.
Unloading the wing by maintaining continuous nose-down elevator pressure
keeps the wing angle of attack as low as possible, making the normal roll controls
as effective as possible.

7.28 FCT 737 (TM) October 31, 2006
B-6
Appendix B
Maneuvers

Pilots should maintain situational awareness since TCAS may issue RAs in
conflict with terrain considerations, such as during approaches into rising terrain
or during an obstacle limited climb. Continue to follow the planned lateral flight
path unless visual contact with the conflicting traffic requires other action.
Windshear, GPWS, and stall warnings take precedence over TCAS advisories.
Stick shaker must be respected at all times. Complying with RAs may result in
brief exceedance of altitude and/or placard limits. However, even at the limits of
the operating envelope, in most cases sufficient performance is available to safely
maneuver the airplane. Smoothly and expeditiously return to appropriate altitudes
and speeds when clear of conflict. Maneuvering opposite to an RA command is
not recommended since TCAS may be coordinating maneuvers with other
airplanes.
Upset Recovery
For detailed information regarding the nature of upsets, aerodynamic principles,
recommended training and other related information, refer to the Airplane Upset
Recovery Training Aid available through your operator.
An upset can generally be defined as unintentionally exceeding the following
conditions:
• pitch attitude greater than 25 degrees nose up, or
• pitch attitude greater than 10 degrees nose down, or
• bank angle greater than 45 degrees, or
• within above parameters but flying at airspeeds inappropriate for the
conditions.
General
Though flight crews in line operation rarely, if ever, encounter an upset situation,
understanding how to apply aerodynamic fundamentals in such a situation helps
them control the airplane. Several techniques are available for recovering from an
upset. In most situations, if a technique is effective, it is not recommended that
pilots use additional techniques. Several of these techniques are discussed in the
example scenarios below:
• stall recovery
• nose high, wings level
• nose low, wings level
• high bank angles

October 31, 2006 FCT 737 (TM) 7.33
B-7
Appendix C
APPENDIX 3-B
Airplane Upset Recovery Briefing
Nose-High, Wings-Level Recovery Techniques

Situation: Pitch attitude unintentionally more than 25 deg, nose high, and increasing; airspeed decreasing rapidly; ability
to maneuver decreasing.
• Recognize and confirm the situation.
• Disengage autopilot and autothrottle.
• Apply as much as full nosedown elevator.
• Use appropriate techniques:
– Roll to obtain a nosedown pitch rate.
– Reduce thrust (underwing-mounted engines).
• Complete the recovery:
– Approaching horizon, roll to wings level.
– Check airspeed and adjust thrust.
– Establish pitch attitude.
Page 3-B.65

• Recognize and confirm the situation
• Disengage autopilot and autothrottle
Extracts from: Manufacturer’s Upset Recovery Training Aid

Figure 3-B.65
C-1
Appendix C
APPENDIX 3-B

• Apply as much as full nosedown elevator.
• Use nosedown stabilizer trim should stick forces be high.
Page 3-B.66

• Apply as much as full
nosedown elevator
Figure 3-B.66
C-2
Appendix C
APPENDIX 3-B

• Use appropriate techniques:
– Roll to obtain a nosedown pitch rate.
Page 3-B.67

• Roll to obtain a
nosedown pitch rate
Figure 3-B.67
C-3
Appendix C
APPENDIX 3-B

• Use appropriate techniques: reduce thrust (underwing-mounted engines).
Page 3-B.68

• Reduce thrust (underwing-
mounted engines)
Figure 3-B.68
C-4
Appendix C
APPENDIX 3-B

– Approaching horizon, roll to wings level.
– Check airspeed and adjust thrust.
– Establish pitch attitude.
Page 3-B.69

– Approaching horizon, roll to wings level
– Check airspeed and adjust thrust
– Establish pitch attitude
Figure 3-B.69
C-5
Appendix C
APPENDIX 3-B
Nose-Low, Wings-Level Recovery Techniques

Situation: Pitch attitude unintentionally more than 10 deg, nose low.
• Recognize and confirm the situation.
• Disengage autopilot and autothrottle.
• Recover from stall, if necessary.
• Recover to level flight.
– Apply noseup elevator.
– Apply stabilizer trim, if necessary.
– Adjust thrust and drag, as necessary.
Page 3-B.70

• Reduce thrust (underwing-
mounted engines)
Figure 3-B.68
C-6

3-2009_G-THOF

Uploaded by

Copyright:

Available Formats

3-2009_G-THOF

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

3-2009_G-THOF

Uploaded by

Copyright:

Available Formats

AIRCRAFT ACCIDENT REPORT 3/2009

Air Accidents Investigation Branch

Department for Transport

Report on the serious incident to

This investigation was carried out in accordance with

Published 21 May 2009

THE FOLLOWING REPORTS ARE AVAILABLE ON THE INTERNET AT

1/2008 Bombardier CL600-2B16 Challenger 604, VP-BJM January 2008

2/2008 Airbus A319-131, G-EUOB January 2008

3/2008 British Aerospace Jetstream 3202, G-BUVC February 2008

4/2008 Airbus A320-214, G-BXKD February 2008

5/2008 Boeing 737-300, OO-TND April 2008

6/2008 Hawker Siddeley HS 748 Series 2A, G-BVOV August 2008

7/2008 Aerospatiale SA365N, G-BLUN October 2008

1/2009 Boeing 737-81Q, G-XLAC January 2009

2/2009 Boeing 777-222, N786UA April 2009

The Right Honourable Geoff Hoon

Dear Secretary of State

Appendix A Fault History

AAIB Aircraft Accidents Investigation ft feet

Aircraft Accident Report No: 3/2009 (EW/C2007/09/12)

Registered Owner and Operator Thomsonfly Ltd

Aircraft Type Boeing 737-3Q8

Place of Incident On approach to Runway 26 at Bournemouth Airport,

Date and Time 23 September 2007 at 2250 hrs

The following Inspectors participated in the investigation:

G-THOF was not subjected to an engineering examination to ensure its continued

The investigation identified the following causal factors:

1. The aircraft decelerated during an instrument approach, to an airspeed

2. The trimmed position of the stabiliser, combined with the selection of

The investigation identified the following contributory factors:

1. The autothrottle warning system on the Boeing 737-300, although

Three Safety Recommendations have been made.

1.1 History of the flight

G-THOF was planned to operate from Faro in Portugal to Bournemouth,

The aircraft started to descend on the glideslope, and about 20 seconds

The commander initially levelled the aircraft at 3,000 ft before climbing to

1.2 Injuries to persons

Injuries Crew Passengers Others

1.4 Other damage

1.5 Personnel information

Male, Age 56 years

The commander had previously spent approximately 17 years on the operator’s

1.5.2 First Officer

Male, Age 30 years

1.6 Aircraft information

1.6.1 Leading Particulars

The aircraft held a current EASA Standard Certificate of Airworthiness, valid

The last maintenance carried out was a 1A check on 20 September 2007.

There were no relevant deferred defects or technical log entries.

The autothrottle computer (Boeing Part No 10-62017-30, Manufacturer’s P/N

1.6.2 Autothrottle system

The autothrottle system automatically positions the thrust levers to maintain

1.6.3 Autothrottle disengagement

Disengagement of the autothrottle will result from any of the following

1. Movement of the autothrottle ARM switch to OFF

2. Activation of one of the autothrottle disengage switches on the

3. Detection of an autothrottle fault within the computer by its

4. Two seconds having elapsed following touchdown

5. Thrust levers becoming separated by more than 10 degrees