NYC93LA129

After takeoff, the engine experienced a partial power loss. The pilot set up for a forced landing on a lake. Prior to landing, the airplane struck a tree, and then came to rest in the lake. Post accident investigation found the two magnetos were driven by a single drive which had a broken impulse coupling spring. The broken spring allowed the engine timing to retard aproximately 20 degrees. When the engine was run, it would not exceed 1900 RPM and was it ran rough. When the impulse coupling was replaced, the engine produced 2800 RPM. Metallurgical examination of the failed spring found the fracture originated with fatigue and then failed in overload. According to the NTSB metallurgical report, the cracks found at the origin of the failure were created when the spring was manufactured.

On Sunday, June 27, 1993, at 1606 eastern daylight time, a Piper PA-28-236, N8220J, operated by Million Air of Teterboro, New Jersey, lost power, and struck trees prior to ditching in Pine Cliff Lake, West Milford, New Jersey. The airplane received substantial damage, and the two pilots received serious injuries. Visual meteorological conditions prevailed. No flight plan had been filed for the flight operating under 14 CFR 91.

The dual instructional flight originated at Teterboro, with a stop at West Milford. After departure from West Milford, the pilots reported, the engine lost power at 500 to 700 feet. Due to terrain, they elected to land in the nearby lake and struck a tree while on final approach.

The engine was test run. According to a written report from FAA Inspector Mr. John D. Cheris, of the Windsor Locks, Connecticut, Flight Standards District Office (FSDO):

After several...attempts, the engine fired and ran very poorly. Backfiring and sporadic operation was noticed. The engine speed was no higher than approximately 1900 revolutions per minute.(RPM) No power could be obtained or sustained...It was found that the impulse coupling appeared to have a failed spring...a serviceable used magneto impulse coupling...[was] installed as a test...the engine fired off readily. An RPM of (2800) was attained...It appeared to reach full power.

A telecon with Mr. Tim Davis of the Teledyne Corporation, current owners of the Bendix Magneto Division, confirmed that the failure, as described would result in a considerable power loss. It would not be unusual to suffer a loss of 1000 RPM as experienced during the test of the subject engine. A total loss of timing would be in the area of 20 degrees thereby changing the operating timing from 23 degrees to 3 degrees. This would cause a considerable loss of engine performance.

The dual magnetos were driven by a single drive, with a single impulse coupling, P/N 10-382952, S/N 29031. According to the NTSB Metallurgical report, 94-03:

Examination [of the spring]...revealed indication of dual fracture mode. A small area of the fracture contained radial markings emanating from the corner between the rounded side and the flat outside surface of the coiled spring. The radial markings are usually associated with a fatigue fracture mechanism. The rest of the fracture had irregular features typical of ductile overstress separation...A longitudinal section, tangential to the rounded side of spring, was cut from a piece of the spring...Examination of the metallographic specimen in the as-polished condition revealed small parallel cracks...The 45 degree orientation of the cracks was consistent with shear cracks developing during spring forming.

FAA Airworthiness Directive, 78-09-07-R3, calls for a 500 hour interval on impulse coupling inspections. According to the airplane maintenance log book, the inspection was last completed on October 31, 1991, at an aircraft hour meter reading of 879.2, 360.9 hours prior to the accident.

The manufacturer, originally Bendix, and currently Teledyne Continental Motors have issued multiple Service Bulletins, starting with SB586 (June 1978), and currently SB632B, dated September 17, 1993, which call for magneto overhaul every four years or engine overhaul, whichever comes first. Examination of the engine log book failed to find any reference to the magneto being overhauled. Service Bulletins compliance is not a requirement for 14 CFR 91 operators.

According to Mr. Jim Brown, an Accident Investigator with Lycoming, the manufacturer of the engine, the serial number on the magneto matches the serial number of the magneto on the engine when it was delivered to Piper Aircraft Corp, on January 21, 1980.