N645JCRobinson Helicopter R44 II · 2008

On July 24, 2009, about 1829 eastern daylight time, a Robinson R44 II, N645JC, registered to CFMC, Inc., sustained main rotor blade damage during an autorotative landing into the Gulf of Mexico near Dunedin, Florida. Visual meteorological conditions prevailed at the time and no flight plan was filed for the 14 Code of Federal Regulations (CFR) Part 91 personal, local flight from Peter O Knight Airport (TPF), Tampa, Florida. The helicopter was substantially damaged due to extended salt water submersion and the certificated private pilot and two passengers were not injured. The flight originated from TPF about 1730.

The pilot stated that earlier that day both fuel tanks were filled and he then flew the helicopter on an uneventful 15 minute flight. The helicopter was then placed in a hangar and about 1730, he and two passengers arrived at the airport for an intended local flight. He performed an inspection of the helicopter while simultaneously informing the passengers about the helicopter, then moved it to an area for departure. He informed the passengers of the use of life vests, and entry doors, and started the engine and hover taxied for a midfield departure. He flew north 4 to 5 miles to remain clear of nearby Class B airspace, then flew northwest, west, then southwest flying towards Clearwater. He then proceeded over the Gulf of Mexico and turned right, flying in a northerly direction. He thought he saw something in the water near the shore and performed two 180 degree turns or a "figure 8," and at the completion of the second 180 degree turn, while flying at 600 feet, the helicopter felt sluggish. He did not notice any yaw motion or unusual vibrations, and he did not notice the engine rpm reading.

The pilot further stated that he applied forward cyclic control to increase indicated airspeed and banked to the left. He saw the low main rotor warning light, and banked to the right, flying towards shore. He then heard the main rotor low warning horn and intended on descending to get into ground effect. He sensed other lights come on, one of them being the fuel light. When the helicopter was between 30 and 40 feet, he flared (applied aft cyclic control), and while descending pulled collective control, which was full up at or just before water contact. He then applied left cyclic control to stop the main rotor blades and all occupants exited the helicopter. The two passengers donned their life vests and all swam to shore.

The helicopter remained submerged for more than 3 days before being recovered. Examination of the helicopter and engine following recovery was performed by a Federal Aviation Administration (FAA) airworthiness inspector. The inspection of the helicopter revealed the four primary engine driven/rotor drive belts were in good condition. The belts were cut then removed. The main fuel tanks were full of salt water and the auxiliary tank had a mixture of salt water and approximately 6 inches of aviation fuel. The gascolator was also full of salt water. All noted fuel injection system components had salt water inside each respective unit. Rotation of the engine by hand revealed valve train continuity and compression at all cylinders. No defects were noted to the air induction or exhaust systems. The magnetos were rotated by hand and no spark was noted at any of the ignition towers. Inspection of the servo fuel injector (fuel servo) revealed the throttle and mixture controls remained attached. The fuel inlet screen was examined and found to have salt water contamination. The magnetos and fuel servo were retained for further examination.

Examination of the fuel servo was performed at the manufacturer’s facility with FAA oversight. Visual inspection of the fuel servo revealed extensive external corrosion. The unit was flushed with preservative oil in order not to contaminate the test bench fluid. The fuel servo was then placed on the test bench and did not flow to specifications. Disassembly inspection of the fuel servo revealed corrosion consistent with salt water submersion. No evidence of preimpact failure or malfunction was reported.

Examination of both magnetos was performed at a FAA certified repair station with NTSB oversight. The right magneto was placed on a test bench as received and did not produce spark. The points, which were corroded, were lightly sanded and the magneto was returned to the test bench. The magneto produced spark at all ignition tower when operated between 643 and 4,704 magneto rpm. Further inspection of the magneto revealed the carbon brush was stuck in the down position; no evidence of arching was noted on the coil tab. No carbon tracking was noted on the distributor block. The capacitor tested .3 microfarads (new specification is .375 +/- 10 percent). The points of the left magneto were lightly sanded and the magneto was then placed on the test bench. The magneto produced spark at all ignition tower when operated between 643 and 4,704 magneto rpm. Further inspection of the magneto revealed the carbon brush was stuck in the down position; no evidence of arching was noted on the coil tab. No carbon tracking was noted on the distributor block. The capacitor tested .25 microfarads (new specification is .375 +/- 10 percent).