Airlink J41 Accident 24 Sep 2009 CAA Prelim Report Released
Thread Starter
Join Date: Apr 2007
Location: Western Cape
Posts: 73
Likes: 0
Received 0 Likes
on
0 Posts
Airlink J41 Accident 24 Sep 2009 CAA Prelim Report Released
INTERIM REPORT NO.1 IN RESPECT OF THE INVESTIGATION INTO THE CAUSE OF AN ACCIDENT
INVOLVING A JETSTREAM 4100 AIRCRAFT SHORTLY AFTER TAKE OFF FROM DURBAN INTERNATIONAL
AIRPORT ON 24 SEPTEMBER 2009
Foreword
The objective of an accident investigation is to establish the cause (s) of the accident and to take steps to prevent
a further occurrence. As such the objective is not to apportion blame or liability.
The purpose of this investigation is therefore to ensure that the investigation is conducted in the most effective
and comprehensive way to establish the cause(s). The investigation team is committed to adhering to the
International Provisions defined in Annex 13 to the Convention on International Civil Aviation, of which South
Africa is a signatory.
The purpose of this Interim Report is to provide the travelling public with information in respect of the progress
with the accident investigation into the cause(s) of the accident.
Issued by the Accident and Incident Investigation Division
SACAA
22 December 2009
2
TABLE OF CONTENTS
FOREWORD 1
Table of Contents 2
SYNOPSIS 3
ORGANISATION OF THE INVESTIGATION 2
1.0 FACTUAL INFORMATION 4
1.1 History of Flight 4
2.0 Wreckage and Impact Information 4
2.1 Damage to Aircraft 4
2.2 Damage to propellers 6
2.3 Damage to Engines 8
3.0 Flight Recorders 10
4.0 Tests and Research 10
4.1 Description of Audio Events and Flight Data events. 10
5.0 Initial Findings 12
6.0 PROPOSED SAFETY RECOMMENDATION 14
7.0 Activities to follow: 15
Attachment A –FDR graph 16
3
SYNOPSIS
The aircraft was on a repositioning flight from Durban International Airport to Pietermaritzburg
Airport with only the three crew members aboard. During the take-off roll smoke was seen
coming from the aircraft’s right engine. The aircraft became airborne and climbed to
approximately 500 ft above mean sea level (AMSL) before losing altitude and making a forced
landing on a small field in a built up area, about 1.4km from the end of the departure runway.
The three crew members were seriously injured in the impact, as well as a member of the public
who was struck by the wing of the aircraft as it landed. One crew member subsequently died
from his injuries a few days after the accident.
Figure 1‐ Reconstructed flight path of the aircraft
Wreckage site
4
Organization of the Investigation
The process followed was in compliance with internationally accepted practices. A team of
accident investigators were dispatched to carry out the onsite investigation which included the
photographing and video-taping of the accident site and wreckage.
Following completion of the initial assessment and documenting of the accident site, the
wreckage was moved to a hangar at Johannesburg, where a more detailed investigation and
strip down of selected components was conducted. This had been done with the support and
cooperation of the operator.
The South African investigation team has been supported by an Accredited Representative (AR)
from the Air Accident Investigation Branch (AAIB) of the United Kingdom and the National
Transportation Board (NTSB) of the United States of America (USA). Both are being assisted by
Advisors.
5
1.0 FACTUAL INFORMATION
1.1 History of Flight
On the day prior to the accident the aircraft had diverted into Durban Airport due to weather.
The crew rested overnight at a local hotel before reporting for duty at 0445Z the next day to
position the aircraft, without passengers, to Pietermaritzburg. The crew consisted of a
commander, who occupied the left-hand seat, a co-pilot, who occupied the right-hand seat,
and a cabin crew member who sat on the flight deck jump seat for the flight.
There were no reported technical problems during the pre flight preparation and the co-pilot
requested start from ATC at 0530Z. As the aircraft was not scheduled to be at the airport
there was no GPU immediately available on their departure and the crew started the
engines using the internal battery. After start ATC cleared the aircraft to taxi to the holding
point for Runway 06 and at 05:56:16Z cleared it to take off. The aircraft commenced its take
off roll from Runway 06 and shortly before it became airborne smoke was seen trailing from
the aircraft. It continued to climb to an altitude of about 500 feet AMSL and yawed to the
right, the commander reporting to ATC that they had “lost an engine”. The aircraft then
descended and made a forced landing in a small field within a residential area, 1.4
kilometers from the end of Runway 06.
The three crew members were badly injured in the impact and had to be released by the
emergency services. A member of the public was also injured when he was struck by the
wing of the aircraft as it landed. The Pilot-in-Command subsequently died in hospital 13
days later from his injuries.
1. Wreckage and Impact Information
2.1 Damage to aircraft
2.1.1 The aircraft struck the ground with the tail cone first in a wings-level attitude with
the landing gear retracted. It skidded for approximately 25 meters before impacting an
electrical pole and crossing a tarmac road. The aircraft then collided with a concrete
palisade fence and came to rest with the fuselage broken in two and detached from the
wings (see photograph). The distance for the first impact point to the main wreckage was 62
meters and the aircraft had come to rest on a heading of the 060ºM.
6
Figure 2: View of the wreckage site
Examination at the accident site found both propellers and spinners to have been badly
damaged but with little or no evidence of rotation. Superficially, both engines were intact
with no evidence of non-containment but, whilst the left engine bore no visual evidence
of pre-impact unserviceability, it was obvious that the right engine had severe pre-impact
distress in the region of the 3rd stage turbine and rear bearing.
The right engine had lost its rear bearing cover and tail cone and it was possible to see
that the rear bearing was missing its inner race, rollers and cage as well as the bearing
retention nut and locking cup. As a result of operation with the bearing degraded, the 3rd
stage turbine blades could be seen to have lost a considerable amount of material span
wise through rubbing. No lubricating oil was visible on the sight glass of the reservoir.
Although a search was conducted for the missing parts at the accident site, airport and
surrounding areas, none were recovered.
Figure 3: View of right hand engine with rear bearing cover and tail cone missing (inset
picture to illustrate these parts which were not recovered).
7
2.1.2 The wreckage was transported to a hangar in Johannesburg, South Africa for further
investigation. At the hangar facility the propellers were removed and sent to an approved facility
for a teardown investigation assisted by a representative from the propeller manufacturer.
2.2 Damage to propellers
The propeller was dismantled at an independent, approved facility in the RSA during the period
12 – 16 October 2009. The Investigator was assisted by a specialist from the manufacturer.
2.2.1 Observations as to damage to the No. 1 Propeller (L/H)
I. All the blades were on the latches with a low blade angle.
II. According to witness marks left on the link pins, the blades were not in the
feather position, but were forced into the feather position as a result of the
impact.
III. The blades do not show any significant signs of power during the impact
sequence.
IV. According to available information the propeller was correctly maintained as
required by the manufacturer.
Figure 4: Damage to No. 1 propeller
2.2.2 Observations as to damage to the No. 2 propeller (R/H)
I. All the link pins were broken and it could not be established whether the propeller
blades were on the latches at the time of impact.
II. However, according to witness marks left on the link pins, it would appear that the
blades were in the feather position, but were forced into the reverse position as a
result of the impact.
8
III. This is further enhanced by the damage sustained to the trailing edges of the all the
blades as well as the twisting of the blades. According to the FDR it would also
appear that the propeller RPM at the time of impact was approximately 50%.
IV. According to available information the propeller was correctly maintained as required
by the manufacturer.
Figure 5: Damage to No. 2 propeller
2.3 Damage to Engines
Both engines were packaged and sent to the manufacturer to conduct further teardown
investigation under SACAA, AAIB and FAA supervision. The results are summarized below.
2.3.1 Damage to Engine 1 (left hand)
The teardown of this engine found no evidence of any pre-impact failures or anomalies.
Its condition was consistent with it being stationary, or near-stationary at impact. The
latching fuel shut-off solenoid within the Integrated Engine Control (IEC) unit was found
in the shut off condition, indicating that an electrical signal had been received to change
its state from the ‘Run’ condition.
2.3.2 Damage to Engine 2 (right hand)
Whilst confirming the damage and missing parts noted on site, the teardown also found
severe loss of turbine blade material on all three stages. In addition, two blades from the
second stage rotor had fractured and a piece of the rim of the rotating air seal was
missing. See photographs below and description of second stage rotating air seal
failures. As the teardown continued towards the compressor and reduction gearbox
9
rubbing damage became less severe and the SACAA and the AAIB are satisfied that the
root cause of the right engine failure was the fractured second stage seal rim.
Figure 5: Views of right hand engine 2nd stage turbine rotating air seal plate (note also
broken turbine blades)
2.3.2.1 Second Stage Turbine Rotating Air Seal failures
The rotating seal is effectively a machined plate which is an interference fit on a
boss on the turbine disc. It serves as a mounting for the air seal between the
rotor and the second stage nozzle guide vanes but also has the function of
retaining the second stage turbine blades and prevents the blades from migrating
forwards in their fir-tree roots: this is the purpose of the rim. Information from
Honeywell has been that there should be a small clearance between the rim and
the blade roots, which do not experience any significant axial forces for the rim to
react.
In 1999, when the first case of seal rim failure was recognized, metallurgical
examination showed that a fatigue mechanism was responsible and that the
fatigue had originated at an area of sharp fretting/wear caused by contact
between the rim and the blade roots. Blade movement was discounted as the
cause of the contact and instead it is thought that the seal plate itself was flexing,
probably under some form of resonant condition.
10
The most damaging effect of the seal rim failure appears to be a considerable
imbalance of the second stage turbine rotor. Typically this vibration has led to
fatigue failures of the rear bearing oil feed and scavenge pipes and consequent
oil starvation and deterioration of the bearing. Possibly with the exception of one
of the other cases of seal failure in 2009 (see below) it is understood that
detachment of the bearing cover, tail cone and retention nut has not been seen
before.
Two further similar cases of seal rim failure led to the issue of Service Bulletin
72-7204 in 2001 which required examination of the seal rim for wear at each Hot
Section Inspection (HSI) of the engine, currently every 4,500 engine hours.
However, at least a further seven cases of rim failure have occurred since then
on parts which have not achieved 4,500 hours Time Since New (TSN) and some
had failed after only some 1,500 Cycles Since New (CSN). This number of
occurrences does not include the ZS-NRM failure and nor does it include a
recent case in which a Jetstream 4100 aircraft reportedly suffered two seal rim
failure events within a 12-day period in 2009
On the right engine of ZS-NRM, the seal had failed after a total of 1,314 CSN and
the left engine was found with the rim worn beyond limits after only 570 cycles.
3.0 Flight Recorders
3.1 The aircraft was equipped with a CVR and FDR as required by the South African Civil
Aviation Regulations, 1997.
3.2 Both the CVR and FDR were recovered from the wreckage and successfully replayed at
the UK Air Accidents Investigation Branch (AAIB) laboratories. The CVR was of 30
minutes duration, and the FDR provided just over 60 hours of data. The accident flight
was recorded by both recorders, with the CVR record commencing from just before first
engine start and ending as the aircraft struck the ground. The CVR operates when
either aircraft battery or external electrical power is applied to the aircraft, whereas
the FDR operates from electrical power provided by the engine driven generators. The
FDR record ended approximately two seconds before the CVR.
3.3 The aircraft was configured for a flap nine takeoff. The takeoff roll appeared normal until
a few seconds before rotation, when the right engine torque indication started to
progressively reduce. The right engine RPM remained normal at this time, as did
indications for the left engine. The right engine torque continued to reduce and as
the aircraft climbed through about 400 ft AMSL, it reached zero. Shortly after this, the
flaps were retracted and at about 500 ft AMSL the aircraft momentarily leveled before
11
both the left engine torque and RPM indications rapidly reduced to zero. The aircraft
then descended. Approximately two seconds before impact, the right engine RPM
was 55 %, with the left engine torque and RPM indications remaining at zero.
4.0 Tests and Research
4.1 Description of Audio Events and Flight Data events.
4.1.1 The last 1 minute and 6 seconds of the CVR were analyzed and combined with the
information obtained from the FDR. A related FDR Graph is included as Attachment A.
The First Officer was the pilot flying (PF) with the Captain as pilot not flying (PNF).
From the FDR data it is evident that the right-hand engine failed after 70 knots but prior
to V1 as indicated by the right hand engine torque which started spooling down. This
occurred as the Tower advised the aircraft of a smoke trail behind the aircraft. All
hydraulic and oil pressures were still normal at this stage.
During rotation at 05:57:01Z, an unknown aircraft transmitted the words “Severe smoke”.
A warning sound [ping] is then heard and the First Officer (FO) stated “right oil
contamination”. The FDR data indicated that the aircraft was at a pressure altitude of
approximately 100 feet and at an indicated airspeed of 140 knots.
The Captain stated at 05:57:10Z that “we have lost an engine” and then “we are losing
an engine”. The FO responded at 05:57:14Z that “I have it, I have it” – “Keeping runway
track six thousand feet”- “Flap is zero””and confirms “we have lost an engine”. The FDR
now indicated a pressure altitude of approximately 350 feet with an indicated airspeed of
140 knots, the right-hand and left hand engine rpm was at 100%. The left hand engine
torque was still at approximately 100% with the right hand engine torque reducing
rapidly. The hydraulic and oil pressures were normal.
At 05:57:25Z the Captain notified the Tower “Okay we’ve lost an engine”. The
associated pressure altitude was approximately 480 feet with an indicated airspeed of
120 knots, The right-hand and left-hand engine rpm was at 100% with the left-hand
engine torque at 100% and hydraulic and oil pressure normal. The FO comments “We’re
not maintaining
At 05:57:28Z three audible warning sounds [pings] are heard and the Master Warning
switch is activated. The right-hand engine beta goes to zero. The left-hand engine torque
is at 100%. The pressure altitude is at approximately 490 feet and indicated airspeed
120 knots. An unidentified transmission advises” Your gear is still down”.
At 05:57:30Z the FO stated “gear up”, followed by the Captain saying “ok gear up”.
However, the left engine now spools down from 100% to zero within 7 seconds. The
pressure altitude was approximately 450 feet with an indicated airspeed of 125 knots.
12
At 05:57:33 three audible warning sounds [pings] are heard together with stick shaker
activation. The indicated airspeed reduces to approximately 117 knots with the angle of
attack of 14 degrees. A low hydraulic pressure is as well as low oil pressure recording is
activated on the right-hand engine.
A clicking sound like a switch or handle moving is then heard on the area microphone
with an associated sound of an engine running down.
The FO stated “wait, wait pitch forward Allister” with the pressure altitude at
approximately 450 feet and indicated airspeed of 130 knots. The pitch attitude was 7.5
degrees nose up.
A GPWS warning of “Don’t sink” was followed immediately by three pings. At 05:57:39Z
the stick shaker can be heard again followed by three audible warning sounds [pings].
The associated pressure altitude was approximately 350 feet with an indicated airspeed
of 110 knots. The pitch attitude is -2.5 degrees nose down angle of attack 14 degrees
and flap setting zero.
The FO comments “fly it out of here” followed by the GPWS stating “Too low”. Another
three audible warning sounds [pings] are heard together with a stick shaker sound in the
background and the FO states “gear is up flaps is***” whereby the Captain confirms
“gear is up flaps***”. The FDR now indicated a pressure altitude of approximately 150
feet and an indicated airspeed of 70 knots. The pitch attitude is 2.5 degrees nose up with
a flap setting of zero and a further three audible warning sounds [pings] are heard.
At 05:57:52 the CVR recording stops.
4.2 FDR loss of power
4.2.1 FDR power was lost approximately 2 seconds before impact due to both generators
dropping off line. The no 1 engine generator due to the engine being shut down and the
no 2 generator due to the spooling down of the failed engine. Recorded data of the last 2
seconds of the fight is therefore not available.
4.2.3 It should be noted that the throttle levers and condition levers settings are not
parameters to be recorded on the DFDR.
5.0 Initial Findings
On the basis of the first factual elements gathered in the course of the investigation, the
following facts have been established:
• Aircraft design and certification requirements call for the capability of a multipleengine
aircraft to continue take-off, climb, en route flight and a landing should one
13
engine become inoperative. This is demonstrated to the relevant certificating
authority by the designer and manufacturer during the aircraft’s certification process.
• The Jetstream 4100 aircraft is certificated to require an operating crew of two pilots
and is able to takeoff and continue to climb should one engine fail, even at its
maximum certificated mass;
• It is a requirement that pilots should be trained and be competent to take off, fly, and
land such aircraft with one engine inoperative. Pilots are required to maintain
competency and are regularly assessed to ensure that such competency is
maintained;
• Operators define standard operating procedures (SOPs) which pilots are to follow
during the various emergencies that may occur. Such competency is again verified
in an actual flight test or in a simulator;
• A smoke trail was observed during the take-off roll and ATC advised the crew
accordingly. This could only be transmitted to the pilot during and on rotation, thus
excluding the possibility of the initiation of an aborted take-off. Duration of the
takeoff roll was about 18 seconds;
• The no.2 or right-hand engine failed on rotation and a power reduction occurred on
the no. 1 engine as the aircraft climbed to about 480 ft. above sea level;
• The aircraft was seen to climb and thereafter descend and impact with terrain. (Total
time from start of the take-off roll until impact was about 110 seconds);
• It has been verified that the aircraft’s mass was such that it should have been able
to have continued to climb and return to land on one engine;
• Rescue and fire-fighting services responded appropriately;
• The three crew members were seriously injured and a member of the public was
injured in the vicinity of the accident site. The PIC died from his injuries before the
he could be interviewed by Investigators;
• As the member of the public was an unsuspecting participant being hit by the
aircraft, any information provided by him would not be able add any further
information or understanding of what was happening in the cockpit, proceeding or
during the impact;
• Investigators have interviewed the surviving crew members. The co-pilot, who was
the designated pilot flying (PF) during the takeoff, unfortunately cannot recall any
detail of the takeoff and subsequent accident. Her husband was however able to
provide information that she initially shared with him during the first two days
following this accident, which she no longer can recall;
14
• It has been verified that the no.1 engine had in fact been serviceable throughout the
flight by means of a tear-down examination;
• Tear-down of the no. 2 Engine established that the engine had suffered a
catastrophic failure due to a fatigue failure of the second stage rotating air seal;
• This kind of failure appears to be the 13th known similar cause of failure of this type
of engine. One of these occurred on the operator’s aircraft where the crew was able
to abort the takeoff. What is of concern is that the number of cycles at which these
failures occur indicates a decreasing trend;
• It is trusted that the State of Design and Manufacture (USA) will initiate the
appropriate corrective action;
• The reason for the power reduction experienced on the no.1 engine, appeared to
have been a misidentification and shutdown of the serviceable engine. This has
been concluded from analysis of the DFDR and CVR information;
• No clear reason for this misidentification could be established other than a complete
deviation from the operator’s prescribed standard operating procedures (SOPs);
• Review of the training and assessment procedures of the operator did not identify
any deviation or known training issues.
6.0 INTERIM SAFETY RECOMMENDATION
6.1 Irrespective of whether the overall engine in-flight shut down and failure rate
caused by this second stage rotating air seal failures are acceptable in terms of
established airworthiness criteria, the SACAA is concerned that parts are failing
at service lives well below those expected. This has the potential to lead to the
further failures with this type of engine.
6.2 The SACAA is aware that some measures have already been taken by
Honeywell to understand the reason for seal failures but that the next stage,
necessary to formulate a solution to the problem, has not been progressed.
6.3 The SACAA therefore makes the following Interim Safety Recommendation to
the Federal Aviation Administration (FAA):
The FAA should require Honeywell Aerospace to
expedite efforts to produce an engineering solution to
the problem of second stage turbine rotating air seal
failures on Honeywell TPE331-14G/H engines.
15
6.4 As result of the deviation from the prescribed Standard Operating Procedures
(SOPs), it is recommended that:
7.0 Activities to follow:
Further action now includes submission of a draft Final Report to the participating States for
comment. These States are entitled to review the draft report for a period of at least 60 days
prior to responding with comments in terms of International obligations.
The investigation will be ongoing in participation with the accredited parties.
The Aircraft Accident and Incident Investigation Division (AIID) of the South African Civil
Aviation Authority (SACAA) wishes to acknowledge and thank the above participants and the
operator for their assistance and services rendered to date.
It is trusted that the investigation will lead to the introduction of corrective action, where any
deficiencies are identified, to ensure the continued safety of passengers transported in South
African airspace.
‐ENDThe
SACAA conducts a comprehensive audit in respect of the
compliance with all aspects of its Air Operator Certificate
requirements, inclusive of its training procedures and assessments of
the involved operator.
INVOLVING A JETSTREAM 4100 AIRCRAFT SHORTLY AFTER TAKE OFF FROM DURBAN INTERNATIONAL
AIRPORT ON 24 SEPTEMBER 2009
Foreword
The objective of an accident investigation is to establish the cause (s) of the accident and to take steps to prevent
a further occurrence. As such the objective is not to apportion blame or liability.
The purpose of this investigation is therefore to ensure that the investigation is conducted in the most effective
and comprehensive way to establish the cause(s). The investigation team is committed to adhering to the
International Provisions defined in Annex 13 to the Convention on International Civil Aviation, of which South
Africa is a signatory.
The purpose of this Interim Report is to provide the travelling public with information in respect of the progress
with the accident investigation into the cause(s) of the accident.
Issued by the Accident and Incident Investigation Division
SACAA
22 December 2009
2
TABLE OF CONTENTS
FOREWORD 1
Table of Contents 2
SYNOPSIS 3
ORGANISATION OF THE INVESTIGATION 2
1.0 FACTUAL INFORMATION 4
1.1 History of Flight 4
2.0 Wreckage and Impact Information 4
2.1 Damage to Aircraft 4
2.2 Damage to propellers 6
2.3 Damage to Engines 8
3.0 Flight Recorders 10
4.0 Tests and Research 10
4.1 Description of Audio Events and Flight Data events. 10
5.0 Initial Findings 12
6.0 PROPOSED SAFETY RECOMMENDATION 14
7.0 Activities to follow: 15
Attachment A –FDR graph 16
3
SYNOPSIS
The aircraft was on a repositioning flight from Durban International Airport to Pietermaritzburg
Airport with only the three crew members aboard. During the take-off roll smoke was seen
coming from the aircraft’s right engine. The aircraft became airborne and climbed to
approximately 500 ft above mean sea level (AMSL) before losing altitude and making a forced
landing on a small field in a built up area, about 1.4km from the end of the departure runway.
The three crew members were seriously injured in the impact, as well as a member of the public
who was struck by the wing of the aircraft as it landed. One crew member subsequently died
from his injuries a few days after the accident.
Figure 1‐ Reconstructed flight path of the aircraft
Wreckage site
4
Organization of the Investigation
The process followed was in compliance with internationally accepted practices. A team of
accident investigators were dispatched to carry out the onsite investigation which included the
photographing and video-taping of the accident site and wreckage.
Following completion of the initial assessment and documenting of the accident site, the
wreckage was moved to a hangar at Johannesburg, where a more detailed investigation and
strip down of selected components was conducted. This had been done with the support and
cooperation of the operator.
The South African investigation team has been supported by an Accredited Representative (AR)
from the Air Accident Investigation Branch (AAIB) of the United Kingdom and the National
Transportation Board (NTSB) of the United States of America (USA). Both are being assisted by
Advisors.
5
1.0 FACTUAL INFORMATION
1.1 History of Flight
On the day prior to the accident the aircraft had diverted into Durban Airport due to weather.
The crew rested overnight at a local hotel before reporting for duty at 0445Z the next day to
position the aircraft, without passengers, to Pietermaritzburg. The crew consisted of a
commander, who occupied the left-hand seat, a co-pilot, who occupied the right-hand seat,
and a cabin crew member who sat on the flight deck jump seat for the flight.
There were no reported technical problems during the pre flight preparation and the co-pilot
requested start from ATC at 0530Z. As the aircraft was not scheduled to be at the airport
there was no GPU immediately available on their departure and the crew started the
engines using the internal battery. After start ATC cleared the aircraft to taxi to the holding
point for Runway 06 and at 05:56:16Z cleared it to take off. The aircraft commenced its take
off roll from Runway 06 and shortly before it became airborne smoke was seen trailing from
the aircraft. It continued to climb to an altitude of about 500 feet AMSL and yawed to the
right, the commander reporting to ATC that they had “lost an engine”. The aircraft then
descended and made a forced landing in a small field within a residential area, 1.4
kilometers from the end of Runway 06.
The three crew members were badly injured in the impact and had to be released by the
emergency services. A member of the public was also injured when he was struck by the
wing of the aircraft as it landed. The Pilot-in-Command subsequently died in hospital 13
days later from his injuries.
1. Wreckage and Impact Information
2.1 Damage to aircraft
2.1.1 The aircraft struck the ground with the tail cone first in a wings-level attitude with
the landing gear retracted. It skidded for approximately 25 meters before impacting an
electrical pole and crossing a tarmac road. The aircraft then collided with a concrete
palisade fence and came to rest with the fuselage broken in two and detached from the
wings (see photograph). The distance for the first impact point to the main wreckage was 62
meters and the aircraft had come to rest on a heading of the 060ºM.
6
Figure 2: View of the wreckage site
Examination at the accident site found both propellers and spinners to have been badly
damaged but with little or no evidence of rotation. Superficially, both engines were intact
with no evidence of non-containment but, whilst the left engine bore no visual evidence
of pre-impact unserviceability, it was obvious that the right engine had severe pre-impact
distress in the region of the 3rd stage turbine and rear bearing.
The right engine had lost its rear bearing cover and tail cone and it was possible to see
that the rear bearing was missing its inner race, rollers and cage as well as the bearing
retention nut and locking cup. As a result of operation with the bearing degraded, the 3rd
stage turbine blades could be seen to have lost a considerable amount of material span
wise through rubbing. No lubricating oil was visible on the sight glass of the reservoir.
Although a search was conducted for the missing parts at the accident site, airport and
surrounding areas, none were recovered.
Figure 3: View of right hand engine with rear bearing cover and tail cone missing (inset
picture to illustrate these parts which were not recovered).
7
2.1.2 The wreckage was transported to a hangar in Johannesburg, South Africa for further
investigation. At the hangar facility the propellers were removed and sent to an approved facility
for a teardown investigation assisted by a representative from the propeller manufacturer.
2.2 Damage to propellers
The propeller was dismantled at an independent, approved facility in the RSA during the period
12 – 16 October 2009. The Investigator was assisted by a specialist from the manufacturer.
2.2.1 Observations as to damage to the No. 1 Propeller (L/H)
I. All the blades were on the latches with a low blade angle.
II. According to witness marks left on the link pins, the blades were not in the
feather position, but were forced into the feather position as a result of the
impact.
III. The blades do not show any significant signs of power during the impact
sequence.
IV. According to available information the propeller was correctly maintained as
required by the manufacturer.
Figure 4: Damage to No. 1 propeller
2.2.2 Observations as to damage to the No. 2 propeller (R/H)
I. All the link pins were broken and it could not be established whether the propeller
blades were on the latches at the time of impact.
II. However, according to witness marks left on the link pins, it would appear that the
blades were in the feather position, but were forced into the reverse position as a
result of the impact.
8
III. This is further enhanced by the damage sustained to the trailing edges of the all the
blades as well as the twisting of the blades. According to the FDR it would also
appear that the propeller RPM at the time of impact was approximately 50%.
IV. According to available information the propeller was correctly maintained as required
by the manufacturer.
Figure 5: Damage to No. 2 propeller
2.3 Damage to Engines
Both engines were packaged and sent to the manufacturer to conduct further teardown
investigation under SACAA, AAIB and FAA supervision. The results are summarized below.
2.3.1 Damage to Engine 1 (left hand)
The teardown of this engine found no evidence of any pre-impact failures or anomalies.
Its condition was consistent with it being stationary, or near-stationary at impact. The
latching fuel shut-off solenoid within the Integrated Engine Control (IEC) unit was found
in the shut off condition, indicating that an electrical signal had been received to change
its state from the ‘Run’ condition.
2.3.2 Damage to Engine 2 (right hand)
Whilst confirming the damage and missing parts noted on site, the teardown also found
severe loss of turbine blade material on all three stages. In addition, two blades from the
second stage rotor had fractured and a piece of the rim of the rotating air seal was
missing. See photographs below and description of second stage rotating air seal
failures. As the teardown continued towards the compressor and reduction gearbox
9
rubbing damage became less severe and the SACAA and the AAIB are satisfied that the
root cause of the right engine failure was the fractured second stage seal rim.
Figure 5: Views of right hand engine 2nd stage turbine rotating air seal plate (note also
broken turbine blades)
2.3.2.1 Second Stage Turbine Rotating Air Seal failures
The rotating seal is effectively a machined plate which is an interference fit on a
boss on the turbine disc. It serves as a mounting for the air seal between the
rotor and the second stage nozzle guide vanes but also has the function of
retaining the second stage turbine blades and prevents the blades from migrating
forwards in their fir-tree roots: this is the purpose of the rim. Information from
Honeywell has been that there should be a small clearance between the rim and
the blade roots, which do not experience any significant axial forces for the rim to
react.
In 1999, when the first case of seal rim failure was recognized, metallurgical
examination showed that a fatigue mechanism was responsible and that the
fatigue had originated at an area of sharp fretting/wear caused by contact
between the rim and the blade roots. Blade movement was discounted as the
cause of the contact and instead it is thought that the seal plate itself was flexing,
probably under some form of resonant condition.
10
The most damaging effect of the seal rim failure appears to be a considerable
imbalance of the second stage turbine rotor. Typically this vibration has led to
fatigue failures of the rear bearing oil feed and scavenge pipes and consequent
oil starvation and deterioration of the bearing. Possibly with the exception of one
of the other cases of seal failure in 2009 (see below) it is understood that
detachment of the bearing cover, tail cone and retention nut has not been seen
before.
Two further similar cases of seal rim failure led to the issue of Service Bulletin
72-7204 in 2001 which required examination of the seal rim for wear at each Hot
Section Inspection (HSI) of the engine, currently every 4,500 engine hours.
However, at least a further seven cases of rim failure have occurred since then
on parts which have not achieved 4,500 hours Time Since New (TSN) and some
had failed after only some 1,500 Cycles Since New (CSN). This number of
occurrences does not include the ZS-NRM failure and nor does it include a
recent case in which a Jetstream 4100 aircraft reportedly suffered two seal rim
failure events within a 12-day period in 2009
On the right engine of ZS-NRM, the seal had failed after a total of 1,314 CSN and
the left engine was found with the rim worn beyond limits after only 570 cycles.
3.0 Flight Recorders
3.1 The aircraft was equipped with a CVR and FDR as required by the South African Civil
Aviation Regulations, 1997.
3.2 Both the CVR and FDR were recovered from the wreckage and successfully replayed at
the UK Air Accidents Investigation Branch (AAIB) laboratories. The CVR was of 30
minutes duration, and the FDR provided just over 60 hours of data. The accident flight
was recorded by both recorders, with the CVR record commencing from just before first
engine start and ending as the aircraft struck the ground. The CVR operates when
either aircraft battery or external electrical power is applied to the aircraft, whereas
the FDR operates from electrical power provided by the engine driven generators. The
FDR record ended approximately two seconds before the CVR.
3.3 The aircraft was configured for a flap nine takeoff. The takeoff roll appeared normal until
a few seconds before rotation, when the right engine torque indication started to
progressively reduce. The right engine RPM remained normal at this time, as did
indications for the left engine. The right engine torque continued to reduce and as
the aircraft climbed through about 400 ft AMSL, it reached zero. Shortly after this, the
flaps were retracted and at about 500 ft AMSL the aircraft momentarily leveled before
11
both the left engine torque and RPM indications rapidly reduced to zero. The aircraft
then descended. Approximately two seconds before impact, the right engine RPM
was 55 %, with the left engine torque and RPM indications remaining at zero.
4.0 Tests and Research
4.1 Description of Audio Events and Flight Data events.
4.1.1 The last 1 minute and 6 seconds of the CVR were analyzed and combined with the
information obtained from the FDR. A related FDR Graph is included as Attachment A.
The First Officer was the pilot flying (PF) with the Captain as pilot not flying (PNF).
From the FDR data it is evident that the right-hand engine failed after 70 knots but prior
to V1 as indicated by the right hand engine torque which started spooling down. This
occurred as the Tower advised the aircraft of a smoke trail behind the aircraft. All
hydraulic and oil pressures were still normal at this stage.
During rotation at 05:57:01Z, an unknown aircraft transmitted the words “Severe smoke”.
A warning sound [ping] is then heard and the First Officer (FO) stated “right oil
contamination”. The FDR data indicated that the aircraft was at a pressure altitude of
approximately 100 feet and at an indicated airspeed of 140 knots.
The Captain stated at 05:57:10Z that “we have lost an engine” and then “we are losing
an engine”. The FO responded at 05:57:14Z that “I have it, I have it” – “Keeping runway
track six thousand feet”- “Flap is zero””and confirms “we have lost an engine”. The FDR
now indicated a pressure altitude of approximately 350 feet with an indicated airspeed of
140 knots, the right-hand and left hand engine rpm was at 100%. The left hand engine
torque was still at approximately 100% with the right hand engine torque reducing
rapidly. The hydraulic and oil pressures were normal.
At 05:57:25Z the Captain notified the Tower “Okay we’ve lost an engine”. The
associated pressure altitude was approximately 480 feet with an indicated airspeed of
120 knots, The right-hand and left-hand engine rpm was at 100% with the left-hand
engine torque at 100% and hydraulic and oil pressure normal. The FO comments “We’re
not maintaining
At 05:57:28Z three audible warning sounds [pings] are heard and the Master Warning
switch is activated. The right-hand engine beta goes to zero. The left-hand engine torque
is at 100%. The pressure altitude is at approximately 490 feet and indicated airspeed
120 knots. An unidentified transmission advises” Your gear is still down”.
At 05:57:30Z the FO stated “gear up”, followed by the Captain saying “ok gear up”.
However, the left engine now spools down from 100% to zero within 7 seconds. The
pressure altitude was approximately 450 feet with an indicated airspeed of 125 knots.
12
At 05:57:33 three audible warning sounds [pings] are heard together with stick shaker
activation. The indicated airspeed reduces to approximately 117 knots with the angle of
attack of 14 degrees. A low hydraulic pressure is as well as low oil pressure recording is
activated on the right-hand engine.
A clicking sound like a switch or handle moving is then heard on the area microphone
with an associated sound of an engine running down.
The FO stated “wait, wait pitch forward Allister” with the pressure altitude at
approximately 450 feet and indicated airspeed of 130 knots. The pitch attitude was 7.5
degrees nose up.
A GPWS warning of “Don’t sink” was followed immediately by three pings. At 05:57:39Z
the stick shaker can be heard again followed by three audible warning sounds [pings].
The associated pressure altitude was approximately 350 feet with an indicated airspeed
of 110 knots. The pitch attitude is -2.5 degrees nose down angle of attack 14 degrees
and flap setting zero.
The FO comments “fly it out of here” followed by the GPWS stating “Too low”. Another
three audible warning sounds [pings] are heard together with a stick shaker sound in the
background and the FO states “gear is up flaps is***” whereby the Captain confirms
“gear is up flaps***”. The FDR now indicated a pressure altitude of approximately 150
feet and an indicated airspeed of 70 knots. The pitch attitude is 2.5 degrees nose up with
a flap setting of zero and a further three audible warning sounds [pings] are heard.
At 05:57:52 the CVR recording stops.
4.2 FDR loss of power
4.2.1 FDR power was lost approximately 2 seconds before impact due to both generators
dropping off line. The no 1 engine generator due to the engine being shut down and the
no 2 generator due to the spooling down of the failed engine. Recorded data of the last 2
seconds of the fight is therefore not available.
4.2.3 It should be noted that the throttle levers and condition levers settings are not
parameters to be recorded on the DFDR.
5.0 Initial Findings
On the basis of the first factual elements gathered in the course of the investigation, the
following facts have been established:
• Aircraft design and certification requirements call for the capability of a multipleengine
aircraft to continue take-off, climb, en route flight and a landing should one
13
engine become inoperative. This is demonstrated to the relevant certificating
authority by the designer and manufacturer during the aircraft’s certification process.
• The Jetstream 4100 aircraft is certificated to require an operating crew of two pilots
and is able to takeoff and continue to climb should one engine fail, even at its
maximum certificated mass;
• It is a requirement that pilots should be trained and be competent to take off, fly, and
land such aircraft with one engine inoperative. Pilots are required to maintain
competency and are regularly assessed to ensure that such competency is
maintained;
• Operators define standard operating procedures (SOPs) which pilots are to follow
during the various emergencies that may occur. Such competency is again verified
in an actual flight test or in a simulator;
• A smoke trail was observed during the take-off roll and ATC advised the crew
accordingly. This could only be transmitted to the pilot during and on rotation, thus
excluding the possibility of the initiation of an aborted take-off. Duration of the
takeoff roll was about 18 seconds;
• The no.2 or right-hand engine failed on rotation and a power reduction occurred on
the no. 1 engine as the aircraft climbed to about 480 ft. above sea level;
• The aircraft was seen to climb and thereafter descend and impact with terrain. (Total
time from start of the take-off roll until impact was about 110 seconds);
• It has been verified that the aircraft’s mass was such that it should have been able
to have continued to climb and return to land on one engine;
• Rescue and fire-fighting services responded appropriately;
• The three crew members were seriously injured and a member of the public was
injured in the vicinity of the accident site. The PIC died from his injuries before the
he could be interviewed by Investigators;
• As the member of the public was an unsuspecting participant being hit by the
aircraft, any information provided by him would not be able add any further
information or understanding of what was happening in the cockpit, proceeding or
during the impact;
• Investigators have interviewed the surviving crew members. The co-pilot, who was
the designated pilot flying (PF) during the takeoff, unfortunately cannot recall any
detail of the takeoff and subsequent accident. Her husband was however able to
provide information that she initially shared with him during the first two days
following this accident, which she no longer can recall;
14
• It has been verified that the no.1 engine had in fact been serviceable throughout the
flight by means of a tear-down examination;
• Tear-down of the no. 2 Engine established that the engine had suffered a
catastrophic failure due to a fatigue failure of the second stage rotating air seal;
• This kind of failure appears to be the 13th known similar cause of failure of this type
of engine. One of these occurred on the operator’s aircraft where the crew was able
to abort the takeoff. What is of concern is that the number of cycles at which these
failures occur indicates a decreasing trend;
• It is trusted that the State of Design and Manufacture (USA) will initiate the
appropriate corrective action;
• The reason for the power reduction experienced on the no.1 engine, appeared to
have been a misidentification and shutdown of the serviceable engine. This has
been concluded from analysis of the DFDR and CVR information;
• No clear reason for this misidentification could be established other than a complete
deviation from the operator’s prescribed standard operating procedures (SOPs);
• Review of the training and assessment procedures of the operator did not identify
any deviation or known training issues.
6.0 INTERIM SAFETY RECOMMENDATION
6.1 Irrespective of whether the overall engine in-flight shut down and failure rate
caused by this second stage rotating air seal failures are acceptable in terms of
established airworthiness criteria, the SACAA is concerned that parts are failing
at service lives well below those expected. This has the potential to lead to the
further failures with this type of engine.
6.2 The SACAA is aware that some measures have already been taken by
Honeywell to understand the reason for seal failures but that the next stage,
necessary to formulate a solution to the problem, has not been progressed.
6.3 The SACAA therefore makes the following Interim Safety Recommendation to
the Federal Aviation Administration (FAA):
The FAA should require Honeywell Aerospace to
expedite efforts to produce an engineering solution to
the problem of second stage turbine rotating air seal
failures on Honeywell TPE331-14G/H engines.
15
6.4 As result of the deviation from the prescribed Standard Operating Procedures
(SOPs), it is recommended that:
7.0 Activities to follow:
Further action now includes submission of a draft Final Report to the participating States for
comment. These States are entitled to review the draft report for a period of at least 60 days
prior to responding with comments in terms of International obligations.
The investigation will be ongoing in participation with the accredited parties.
The Aircraft Accident and Incident Investigation Division (AIID) of the South African Civil
Aviation Authority (SACAA) wishes to acknowledge and thank the above participants and the
operator for their assistance and services rendered to date.
It is trusted that the investigation will lead to the introduction of corrective action, where any
deficiencies are identified, to ensure the continued safety of passengers transported in South
African airspace.
‐ENDThe
SACAA conducts a comprehensive audit in respect of the
compliance with all aspects of its Air Operator Certificate
requirements, inclusive of its training procedures and assessments of
the involved operator.
Join Date: Aug 2006
Location: On the move
Posts: 940
Likes: 0
Received 0 Likes
on
0 Posts
Ok that question has now been answered , they crashed themselves. Sorry to hear that outcome but these things happen under stress . It sounds like they were completely overcome by the accident, gear left down , trying to climb and then shut down the live engine.
Join Date: Oct 2009
Location: Surreal
Age: 54
Posts: 69
Likes: 0
Received 0 Likes
on
0 Posts
Yes, it appears that the wrong engine was shut down, but who did it ? Also the report clearly points out that they are sh1t engines. Pilots shouldn't have to live with this.
Join Date: Sep 2005
Location: jnb
Posts: 26
Likes: 0
Received 0 Likes
on
0 Posts
No surprise there
There is certainly truth in your last statement AVI8TOR and that is once again you are totally wrong infact most of your posts are self opinionated crap.
Must be so nice to be able to live life through the eyes of hindsight. The only thing thats truly sad is you are a part of our profession.
Must be so nice to be able to live life through the eyes of hindsight. The only thing thats truly sad is you are a part of our profession.
Join Date: Dec 2009
Location: gauteng
Posts: 16
Likes: 0
Received 0 Likes
on
0 Posts
I honestly don't believe that they will be able to survive this as their J41's make up more than half their operating fleet. They have been the flagships that have carried the company to where they are today. The media publicity will definetly have an impact on the rest of the fleet, remembering that they are already one embraer down as well.
Join Date: Jan 2005
Location: RSA
Posts: 287
Likes: 0
Received 0 Likes
on
0 Posts
Thanks for posting the report gyro.
From the CVR when Sonja says, "wait, wait pitch forward Allister"
sounds like he took control. Stress makes you do silly things, thats
why simulator training is so important, so hopefully when the proverbial
"sh1t hits the fan" our training kicks in and we sucessfully manage the situation.
Heard Link only send their pilots every 18 mths for sim recurrency.
From the CVR when Sonja says, "wait, wait pitch forward Allister"
sounds like he took control. Stress makes you do silly things, thats
why simulator training is so important, so hopefully when the proverbial
"sh1t hits the fan" our training kicks in and we sucessfully manage the situation.
Heard Link only send their pilots every 18 mths for sim recurrency.
Join Date: Apr 2006
Location: In the seat
Posts: 284
Likes: 0
Received 0 Likes
on
0 Posts
Right, that's the interim report lets wait and see what the contributing factors are - were they properly rested, was the training done correctly - then they'll need to find a "fall guy". Don't cut back work well?
Has JvJ packed his bags yet?
Has JvJ packed his bags yet?
