Sequence of events
On 21 February 2003, at 1615 Western Standard Time, shortly
after commencing the take-off run on a charter flight from Lake
Johnston to Perth, Western Australia (WA), the crew of the Cessna
Aircraft Company 441 Conquest, registered VH-LBZ observed the right
engine exhaust gas temperature and power increasing without pilot
input. There were no other abnormal indications or illuminated
warning lights accompanying the unsolicited power increase from
that engine.
The aircraft was crewed by a pilot in command (PIC), as the
pilot flying, and a supervising pilot. The PIC rejected the takeoff
by attempting to reduce power from both engines with the power
levers. While the left engine reduced to idle power, the right
engine remained at maximum power. The aircraft veered to the left,
and both pilots applied the aircraft brakes and attempted to
maintain directional control. The PIC acted to move the fuel
control from AUTO to MANUAL mode by selecting the fuel computers
OFF, and selected both engine stop buttons to initiate engine shut
down. As the PIC was selecting the condition levers to EMER
SHUT-OFF, the aircraft entered the runway strip at an estimated
speed of about 65 kts.
Due to vibration of the aircraft over the rough terrain, the
crew could not confirm whether the right engine stop button had
been fully depressed, or whether the condition levers had been
fully selected into the SHUT-OFF detents prior to the aircraft
impacting a sand mound. The PIC secured the aircraft and the pilots
and passengers evacuated from the aircraft uninjured.
The wind was reported to be a `light' north-north-easterly, with
a crosswind component estimated at 3 kts. The crew of the aircraft
reported no wind gusts immediately prior to, or during the
occurrence.
The aircraft came to rest nose-down in soft sand approximately
45 m to the left of the runway strip, and 750 m from the brakes
release point. Inspection of the damaged aircraft by a salvage team
revealed that the nose landing gear had collapsed, the main landing
gears had sunk into the soft sand, and that both propellers struck
the ground. All engine control runs were confirmed to be intact,
and correctly rigged. On-site examination of the right engine
control and indicating systems was unable to explain the
uncontrolled increase in power from that engine during the
occurrence. The investigation was unable to confirm the
pre-occurrence continuity of the electrical harness that activated
the fuel control unit (FCU) manual mode solenoid valve or the
position of the computer mode switches (either AUTO or MANUAL)
prior to the pilot securing the aircraft.
The PIC had flown about 161 hours on the Conquest, all of which
were flown in the last 90 days. There was no indication that the
PIC had flown another type during that time. The supervising pilot
had a total of about 1,000 hours on type at the time of the
occurrence.
Temporary revision 10 to the aircraft Pilot's Operating Handbook
(POH) included the actions for an Uncommanded Power Increase
In-flight (takeoff, climb, cruise, approach or landing) and On
[the] Ground During Takeoff Roll or Landing Rollout. Actions in
response to an in-flight power increase included reducing the
appropriate condition lever and selecting the fuel computer switch
OFF. The memory actions in response to an uncommanded power
increase on the ground included:
Directional Control - MAINTAIN WITH DIFFERENTIAL
BRAKING AND RUDDER
Engine Stop Button (Engine with High Torque) -
Push
Power Levers - FLIGHT IDLE
The right engine FCU and a number of other engine components
were examined at the engine manufacturer's facilities in the United
States under the supervision of the National Transportation Safety
Board. It identified fine red dust deposits in, and located
downstream of the FCU in-line P-3 air filter, and a blocked flow
restrictor (Viscojet) within the FCU. Advice from the engine
manufacturer was that:
- When in the MANUAL mode, a blocked Viscojet under rare
conditions could act like a check valve, trapping P-3 air within
the FCU control bellows chamber;
- When operating in the AUTO mode, and with uninterrupted
electrical power to the FCU manual mode solenoid valve, the engine
should have been controllable, and decelerated when the PIC
retarded the condition levers to idle; and
- When in the AUTO mode, an interruption to the electrical power
supply to the manual mode solenoid valve could lead to P-3 air
being trapped in the control bellows.
The effect of P-3 air being trapped in the control bellows
chamber was that pilot control of the fuel supplied to the affected
right engine was not possible.
The operator reported that the occurrence aircraft history
included extensive operations in the hot, dusty areas of northern
and inland WA. At the time of the occurrence, the operator was
cleaning the P-3 air filter element every 100 hours of engine
operation. That action halved the aircraft maintenance manual
requirement to clean the element every 200 hours.
It appeared that, over time, operation of the aircraft in the
northern WA environment had resulted in the inability of the P-3
air filtration system, and element cleaning regime to prevent fine
red dust from infiltrating the internal, precision parts of the
FCU.
Two possible explanations exist for the loss of control of the
engine:
The trapped P-3 air in the control bellows meant that there
would have been no control in MANUAL mode and the PIC's reported
attempt to select MANUAL mode would, if successfully carried out,
not have returned fuel control authority to the pilot.
The PIC's stated action to lower the fuel computer guards would
have placed the computers in AUTO mode. For an uncontrolled fuel
supply to the right engine in the AUTO mode to be a possibility, it
necessitated an interruption to the electric power supply to the
FCU manual solenoid valve. The investigation could not confirm
whether such an interruption in power supply had occurred in this
case, nor could it be ruled out.
The PIC's initial reaction to reduce engine power and select
engine management to MANUAL mode, although understandable given the
circumstances, was not in accordance with the memory items
promulgated in the temporary revision to the POH. The partially
successful PIC's action to reduce engine power from both engines
resulted in an increased power difference between the left and
right engines, and contributed to the difficulty experienced by the
crew to maintain aircraft directional control. While the reasons
for the difference between the actions promulgated in the POH in
response to this emergency, and those taken by the PIC could not be
explained, it appeared that the PIC might have initially carried
out the actions for an uncommanded power increase when
in-flight.
It was unlikely that any action taken by the PIC, other than an
immediate and successful shut down of the affected right engine,
would have allowed him to maintain directional control, and prevent
the aircraft from departing the runway during the take-off roll.
Although not in accordance with the temporary revision to the POH,
ultimately, the PIC activated the engine STOP button and selected
the condition levers into the SHUT-OFF detents in order for that to
occur.
