A comparison of the takeoff run used by the aircraft and that predicted by the manufacturer shows that the aircraft's performance during the takeoff was significantly less than expected. The fact that the aircraft could get airborne and slowly climb, but later could not maintain altitude with the landing gear retracted indicates that the aircraft's performance decreased during the flight. As the pilot and the witnesses agreed that no sudden loss of performance occurred, it was concluded that a slow deterioration in engine power output occurred over the period of the flight. As the takeoff run was longer than expected, this deterioration began at the start of, or during, the takeoff run.
The reduction in performance that was felt by the pilot after takeoff was consistent with the normal loss of performance produced by the gear retraction.
A loss of performance of an aircraft can be due to an increase in drag, a decrease in power from the engine, environmental factors such as down draughts, or a combination of these factors. An increase in drag can be the result of:
- damage to the airframe,
- a configuration change (flaps or gear extended), or
- poor technique by the pilot (for example, incorrect airspeed or unbalanced flight).
An increase in drag is not considered a likely explanation for the degraded performance for the following reasons:
- After the accident, there was no damage evident other than that which was consistent with the circumstances of a ditching and recovery.
- The aircraft was recovered with landing gear retracted and flaps at the 10 degree position.
- It is considered unlikely that an experienced pilot who was current on single-engined, light aircraft could mishandle the aircraft to an extent that would have produced the observed performance.
As no other aircraft flying at the time experienced difficulty in climbing after takeoff, environmental factors are not considered to have contributed to the degraded performance of the accident aircraft. This leaves a reduction in engine power as the most likely explanation for the degraded performance. This is also supported by the witness evidence that the engine note on the accident flight was unusual for the aircraft type.
As no fault could be found with the engine and propeller, other reasons for the power loss were considered. Taking off with the propeller control in the cruise position would have produced sufficient power for the aircraft to maintain altitude or to climb slowly. Selecting the fuel pump to the high position would have produced a total loss of power (as shown by previous accidents) or a sudden, substantial loss of power, not the gradual loss of power observed. Fuel starvation resulting from prolonged, unbalanced flight with less than one quarter full fuel tanks would have produced a sudden, total loss of power. Vapour lock in the fuel line would also have produced total loss of power. The ambient conditions were not conducive to induction system icing.
The observed performance of the aircraft was consistent with a restriction in the fuel supply line. If such a restriction existed, it would only have become apparent at high power settings; it would not have shown up during the engine runup. At medium and low power settings, sufficient fuel could flow past the restriction to produce normal engine power output and indications. However at high power settings, such as takeoff power, the engine would consume fuel at a faster rate than could flow past the restriction. A high power output would be produced for a short period as excess fuel in the engine fuel system was consumed. As the amount of excess fuel reduced, the engine power output would progressively decrease until the rate of fuel consumed by the engine equalled the rate of fuel passing through the restriction.
The partially open fuel selector valve would have restricted the fuel flow to some extent, but should have allowed the engine to produce sufficient power to maintain altitude or slowly climb. The fact that the aircraft was apparently unable to maintain altitude during the later stages of the flight indicates that the partially open fuel selector valve, on its own, does not explain this accident. The partially open fuel selector valve could have restricted the fuel flow to a greater extent and produced the observed aircraft performance if:
- dirt or other foreign matter further obstructed the partially open port in the fuel selector valve, or
- during the accident flight, the fuel selector valve was in a "more-closed" position than when the aircraft was recovered (only one or two millimetres of movement would have been needed).
Although the pilot had sufficient distance in which to reject the takeoff at any point up until liftoff, he was unfamiliar with this specific aircraft and its expected performance. Even though the takeoff run was longer than predicted, it was similar to the takeoffs in the Maule floatplane that the pilot was most familiar with and to him, may have appeared normal. That perception was reinforced by the lack of any obvious changes in engine indications or aircraft performance during the takeoff. The first indication the pilot perceived of the deteriorating performance was when the aircraft was slow to climb as the landing gear was retracting. Rejecting the takeoff at this point may have presented a serious risk to the aircraft and its occupants. It was concluded that the pilot did not perceive that the takeoff was abnormal until he was past the point at which the takeoff could be safely rejected.
The report by the pilot of seeing 22 inches of manifold air pressure after takeoff could not be reconciled. Such an indication could only be produced by a blockage of the engine induction system or by the throttle valve closing. Examination after the accident found no blockages and the throttle linkages were operating normally. However, in a high stress situation and in an aircraft type that he had not flown recently, the pilot may have misread the manifold air pressure gauge.
CONCLUSION
Although the observed performance of the aircraft was consistent with a restriction in the fuel supply system, the reason for the low power output from the engine could not be conclusively determined.