The ATSB does not have the resources to carry out a full
cost-benefit analysis of every recommendation. The cost of any
recommendation must always be balanced against its benefits to
safety, and aviation safety involves the whole community. Such
analysis is a matter for the body to which the recommendation is
addressed (for example the Civil Aviation Safety Authority in
consultation with the industry).
FACTUAL INFORMATION
Background
On 23 March 2001, a New Zealand registered Hughes 500D helicopter,
ZK-HMN, sustained an in-flight engine flameout that necessitated a
forced landing near Milford Sound, New Zealand. The pilot and
observer, the only occupants, escaped uninjured, however the
helicopter was substantially damaged. The New Zealand Transport
Accident Investigation Commission (TAIC) conducted an investigation
of the accident (TAIC Investigation Report 01-003 refers). The
helicopter's engine, an Allison model 250-C20B, serial number
821211F, with 8,066 hours time since new (TSN) was removed to an
Australian test facility for engine test runs, with a
representative of TAIC attending. The results of the test run
indicated an anomaly in the fuel control unit (FCU).
FCU initial test rig testing
FCU part number 2924644-29, serial number 336660, with 1,400.25
hours time since overhaul (TSO), was removed and tested at an
authorised Australian overhaul facility. Upon testing, the FCU
could not be operated successfully and failed the maximum governor
spring setting test point (TP) number 9.000. After consultation
with the FCU manufacturer, the FCU was partially disassembled to
allow internal examination. During the examination of the FCU, it
was noted that the flyweight clip adjustment appeared to be out of
tolerance. The clip was adjusted to the specified clearances and
the test continued. The FCU was then successfully operated and
tested.
FCU background
The FCU was overhauled on 28 December 1996 by an Australian
overhaul facility. Following overhaul, it was certified as passing
all tests on the test rig as required by the FCU manufacturer. It
was then installed on the engine of ZK-HMN on 17 March 1997. On 31
December 1997 it was removed after completing 238.5 hours TSO, for
rectification of a defect which was entered in the logbook as
'Constant operation at 110% N2 (power turbine speed)'. The FCU was
then shipped back to the same overhaul facility for repair. Upon
induction testing, the FCU could not be successfully operated and
failed the maximum governor spring setting TP, among other test
parameters. Further examination revealed two 50mm cracks at the FCU
drive governor body housing, and a resulting air leak. The overhaul
facility recommended a complete overhaul of the component.
Subsequently, on 3 July 1998, the FCU was partially stripped to
replace the drive governor housing, with all seals that were
disturbed replaced. A review of the FCU documentation indicated
that the FCU was then successfully tested and shipped back to the
operator.
On 31 July 1998, the FCU was reinstalled on the engine of ZK-HMN,
replacing a loan component. The engine had operated 162.0 flight
hours with the loan FCU installed. There were no discrepancies
annotated in the helicopter or engine logbooks referring to any N2
anomalies while the loan component was installed. During the next
1,000.3 hours of engine operation, the operator removed and
replaced the Power Turbine Governor (PTG) three times attempting to
resolve a discrepancy of engine N2 drooping and hunting.
FCU flyweight clip
Examination of the flyweight clip of the subject FCU during
technical disassembly revealed that the clip may have never been
physically adjusted. The flyweight clip adjustment of the component
set the travel limits of the FCU flyweights in order to adjust fuel
flow according to engine RPM and power demands. The nominal
adjustment of the flyweight stop clips and the flyweight levers
during overhaul or repair, according to the manufacturer's manual
were '0.100-0.105 inches' clearance. The measured dimensional
clearance between the flyweight stop clips and the flyweight levers
of the accident FCU when checked were .042 and .044 inches
(measuring two sides). That value correlated to the measurements of
a new, unadjusted, flyweight clip.
FCU testing
FCU testing following overhaul or repair was completed on a
calibrated test rig to ensure acceptable levels of performance. The
test measured RPM, fuel flow, pressure, throttle speed, and airflow
against manufacturer provided performance values referred to as
Test Points (TP). According to the FCU manufacturer, with the
flyweight clip adjusted as found on initial technical disassembly
examination following the accident, the component would not have
successfully passed TP number 9.000. During testing of the
component prior to the technical disassembly examination, and
during an earlier induction for repair, it did not successfully
pass TP number 9.000.
FCU overhaul facility documentation examination
Overhaul facility documentation was examined for indications of
component testing results. According to the documentation, the FCU
passed all tests on the test rig following both overhaul and repair
of the component, including TP number 9.000.
The overhaul and repair documentation listed values to record the
clearance between the flyweight stop clips and the flyweight
levers. The repair documentation listed the recorded measurement as
0.101 inches. The overhaul documentation listed the recorded
measurement as 0.100 inches. According to the FCU manufacturer's
published requirements, both were considered as acceptable
measurements. Neither the overhaul, nor the repair parts list
documentation for the component, annotated a replacement flyweight
clip as being installed.
Foreign airworthiness regulator surveillance
The United States of America (USA) Federal Aviation Administration
(FAA) conducted regular surveillance of the overhaul facility
because of the use of overhauled and repaired components of that
organisation in aircraft registered in the USA. They informed the
Australian Transport Safety Bureau (ATSB) that in 1997 they had
revoked the facility's FAA Air Agency certificate because the
facility did not conform to FAA regulations. The FAA further
advised that, in December 2000, following several management and
personnel changes within the overhaul facility, a total
certification inspection was completed and the Air Agency
certificate was reinstated.
Australian Civil Aviation Safety Authority surveillance
The Australian Civil Aviation Safety Authority (CASA) provided the
ATSB with all surveillance files of the overhaul facility available
at the time of the request. According to the provided files, the
last surveillance of the organisation prior to the overhaul date of
the FCU was 13 February 1996. Following that audit, three
Non-compliance Notices (NCN) were issued outlining discrepancies.
On 17 December 1997, another audit was conducted with two NCNs
issued. On 16 June 1998, another audit was conducted with one
documentation discrepancy noted. On 30 September 1998, another
audit was completed with one documentation related NCN issued. On
16 April 1999, another NCN was issued citing no corrective action
following an internal audit. There was no documentation found on
the provided files annotating the revoking of the facility's FAA
Air Agency certificate in 1997. CASA also advised that there was no
formal mechanism through which CASA would be advised of the results
of any surveillance action taken by a foreign airworthiness agency
on an Australian Certificate of Approval holder.
ANALYSIS
Physical evidence following the technical examination of the fuel
control unit (FCU) revealed that the flyweight clip had not been
manually adjusted to the required value. However, assembly
documentation for the repair and overhaul of the component had
annotated that the flyweight clip had been adjusted and that the
dimensions were within acceptable limits.
The operator reported that the engine had operated with anomalies
in the N2 (power turbine speed) following both the overhaul and
repair of the FCU. Following the overhaul, upon removal for repair,
cracks were found in the governor (drive body) housing that
affected normal operation of the FCU. The reason for the FCU not
passing TP number 9.000 during the induction testing prior to
repair may have been due to the cracked governor (drive body)
housing.
CONCLUSION
The flameout of the engine of ZK-HMN was most likely a result of
internal anomalies of the engine FCU. The FCU was found to have a
replacement flyweight clip installed that was not adjusted to the
manufacturer's specifications. The investigation could not
conclusively determine when the flyweight clip was replaced with
the unadjusted clip.
Currently, there are no formal protocols for foreign regulatory
agencies to provide CASA with the results of an audit inspection of
an Australian Certificate of Approval (C of A) holder that is also
approved by that foreign agency. Information concerning the results
of those audits could provide valuable information for use in CASA
surveillance of Australian C of A holders.
As a result of this occurrence, the Australian Transport Safety
Bureau issues the following safety recommendation: