N9087
Enstrom F28A
Shortly before 1510, the pilot started the helicopter's engine, then departed on a sightseeing tour flight around the island with two fare-paying passengers. The operator anticipated the flight would last between 0.5 and 0.7 hrs. For weight considerations, the fuel load had been reduced before takeoff by an estimated 10 gallons bringing the usable quantity to between 16 and 19 gallons, as estimated using a dipstick. About 1558, while cruising off shore & nearly parallel to the shoreline, witnesses observed the helicopter fly past while making sounds like its engine was sputtering, stalling, or cutting in and out. The helicopter's course remained unchanged, & seconds later, other witnesses heard a loud bang sound. The helicopter then entered a rapid descent, & impacted in about 2.5-foot-deep ocean water in front of a hotel where people were swimming. An exam of the wreckage revealed a main rotor blade had impacted & severed the tail rotor drive shaft. No evidence was found indicating the pilot had practiced autorotations in the Enstrom helicopter during the 4.5 months since becoming rotorcraft rated. During the structural airframe and engine examinations, no evidence was found of precipitating mechanical malfunctions. Calculations based on maximum anticipated fuel consumption rates, and approximations for the quantity of fuel on board, revealed fuel exhaustion could occur within 0.1 hour of the flight's duration.
HISTORY OF FLIGHT On January 12, 1997, about 1558 hours Saipan standard time (0558 UTC), an Enstrom F28A, N9087, operated by the Micronesian Aviation Corporation, d.b.a. Macaw Helicopters, experienced a total loss of engine power while cruising offshore near a beach fronting the Saipan Grand Hotel, about 3 miles northwest of the Saipan International Airport, Saipan. The helicopter rapidly descended and was destroyed by impact forces and immersion into the salt ocean water. The accident occurred while the pilot was providing the two fare-paying passengers with a 14 CFR Part 91 sightseeing flight over the island. A company visual flight rules (VFR) flight plan was filed, and visual meteorological conditions prevailed. The commercial pilot and one passenger were fatally injured, and a second passenger was seriously injured. Two bystanders, who were in a designated swimming area in front of the hotel, were seriously injured by debris from the falling helicopter. The accident occurred while the pilot was returning to the Coral Ocean Point Proshop Heliport, where the flight had originated about 1510. The sole surviving occupant of the helicopter reported (through a translator) that the flight had proceeded normally when he heard a very loud boom sound which was dissimilar to engine noise. Immediately thereafter, an intense vibration was felt, and the helicopter rapidly descended. Eight statements were received from witnesses located on or near the beach in the vicinity of the Saipan Grand Hotel, and who observed the southbound flying helicopter seconds prior to the crash. The witnesses generally indicated that the helicopter had been cruising nearly parallel to the shoreline when the event occurred. Two statements were received from witnesses whose estimated position was between 2/10 and 5/10-mile north of the crash site, and they too indicated the helicopter had been flying nearly parallel with the shoreline. In summary, the eight witnesses closest to the crash site generally indicated hearing a noise, which they described as either being a loud bang or an explosion sound, followed by their observations of the helicopter descending at a steep angle with little forward airspeed. Three of these witnesses commented that when the helicopter descended its main rotor blade was rotating slowly. None reported hearing any engine noise. The two witnesses located north of the crash site both reported hearing engine noises. The witness located about 5/10-mile north of the crash site estimated that when the helicopter flew past his position it was between 300 and 400 meters off shore. The witness provided the following statement regarding his observations of the events which transpired thereafter: "Shortly after the helicopter had passed us, I heard the aircraft make a sound like the engine was stalling. I heard this sound a second time about two or three seconds later after which I observed the aircraft begin heading down at approximately a forty five degree angle. . . . At this time, I thought the aircraft was trying to land at the beach. About four to five seconds after I first observed the aircraft start heading down, I saw it crashed (sic) into the water." A witness located about 2/10-mile north of the crash site reported that he heard the engine make a sputtering sound like it was cutting out. Thereafter, he observed the pilot looking outside the helicopter just before it fell into the water. Neither of these witnesses reported observing the helicopter turn toward the shoreline during its descent. PERSONNEL INFORMATION A review of the pilot's personal flight record logbook indicated that he commenced rotorcraft flight training in July 1996, and on August 29, 1996, he passed the commercial pilot check ride. All flight training, including the flight check, was performed in a Bell 47. No evidence of additional dual instruction or autorotation practice was observed recorded in the pilot's logbook following his certification check ride. The operator reported that he had provided the pilot with a portion of the rotorcraft flight instruction. The pilot had been a company employee since June 2, 1996, and initially worked as a mechanic and fixed wing pilot. AIRCRAFT INFORMATION A review of the helicopter's maintenance records indicated the engine was overhauled in November 1995 by a contract facility. The operator's director of maintenance reported that he installed the engine into the accident helicopter on July 18, 1996. A mechanic, employed by the operator, reported that during the installation process, the engine was found timed to 19 degrees before top dead center (BTDC). The mechanic stated that he reset the timing to the appropriate 25-degree setting. The operator further reported that since July 18, no one had checked the engine timing or moved the magnetos on their engine mounting pads. All maintenance since the engine installation, including the last annual inspection, was accomplished by the operator's employees. The operator reported that he flew the accident helicopter on January 10, 1997, which was the last time it was operated prior to the accident flight. During the operator's pretakeoff engine check, the magnetos were functionally checked. The engine speed decreased between 50 and 75 revolutions per minute (RPM) when each magneto was momentarily switched off, and no roughness occurred. No anomalies to any helicopter system were noted during the flight. At its conclusion, the engine shut down procedure was accomplished during which the grounding of each magneto was checked. The engine briefly quit when each magneto was turned off. No anomalies were noted. The helicopter was not equipped with flotation devices. According to the Enstrom Helicopter Corporation participant, when the helicopter was manufactured shoulder harnesses were not available from the company, and none were installed in the helicopter. They have subsequently become available from Enstrom. METEOROLOGICAL INFORMATION The operator and persons who resided near the accident site area reported no unusual weather phenomena during the accident flight. The weather was generally described as being a typical "VFR" day with good visibility, calm sea, and a light easterly 5-knot breeze. COMMUNICATION According to air traffic controllers located at the Saipan Airport, all radio communications with the pilot throughout the flight were normal. WRECKAGE AND IMPACT INFORMATION The wreckage was recovered from the accident site and examined by Federal Aviation Administration (FAA) personnel prior to the National Transportation Safety Board's on-scene arrival. According to the Commonwealth of the Northern Mariana Islands, Department of Public Safety, and the helicopter operator, the main wreckage was primarily located in one main area. It was found partially submerged in estimated 2 1/2-foot-deep water. The tail rotor drive shaft (TRDS) was observed severed into two sections. The short section was found about 40 feet from the main wreckage, and the long section was found about 256 feet from the main wreckage. A TRDS bearing was found approximately 70 feet upwind (and up current) from the flight path. There was no evidence of fire. The FAA coordinator reported to the Safety Board that his examination of the helicopter commenced on January 14, 1997. The FAA coordinator's complete statement is included in the Safety Board's report. The following partially summarizes his observations and findings: Based upon the observed physical evidence, a main rotor blade struck the tail boom and severed the TRDS into two pieces. When the severance occurred, the drive shaft was rotating very slowly. The tail rotor blades were found intact. The control cables going to the tail rotor were found severed at the point of the main rotor blade strike. Three control rods are located within the mast structure, and during flight they constantly move up and down changing the pitch of the main rotor blades. The control rod, which had been connected to the sole main rotor blade which bore tail boom impact evidence, was found with signs of corrosion. No similar evidence of corrosion was observed with the other two control rods. The cooling fan assembly showed no evidence of rotational damage. The cooling blades were observed intact. The shroud around the cooling fan showed no evidence of rotational contact with the cooling blades. The drive belt was found engaged. No evidence of preimpact anomalies was found during the examination of the engine's exterior, accessory gears, fuel lines and inlet screen, fuel shutoff valve, oil sump, and the main gearbox. Several ounces of fuel were found in the gascolator bowl. The Enstrom Helicopter Corporation participant provided a report to the Safety Board further detailing the FAA's observations and providing additional comments. The participant noted evidence of the cockpit seats being deformed in a direction consistent with a severe downward impact force. Except for evidence of impact-related damage, the flight control system maintained its integrity. The participant's observations were consistent with the FAA's regarding the lack of fan and engine rotation evidence at impact. (See the Enstrom report for additional information.) The Safety Board's examination of the wreckage commenced on February 20, 1997. The aft portion of the tail boom's left side was observed crushed in an inward direction with an imprint mark matching the size and curvature of a main rotor blade's leading edge. When a main rotor blade was positioned on the mast and allowed to droop downward, its tip reached the area where the tail rotor drive shaft was observed to have been severed. The short, 3.3-foot-long portion of the drive shaft was found bent into a 45-degree arc, and it was observed severed from the 12.5-foot-long section of drive shaft. The only main rotor blade which was observed bent had its respective control rod disconnected from its fittings. Also, the rod's end plug was found pulled out. The air filter element was found devoid of foreign material. No obstructions were noted in the air hoses between the air cleaner and the fuel servo adapter connection. No evidence of foreign debris was found inside either of the fuel tanks. No blockages were found in any examined fuel lines. No evidence was of abrasions in the magneto wiring harness was observed. The right side passenger door was found with a crushed frame. The door latching/locking mechanism was observed with an impact mark consistent with the door being closed and locked at the time of ground impact. The engine's ignition wiring harness was examined and tested. No evidence of chafing, wear, or arcing was found. The starter gear teeth appeared undamaged. The engine fan shroud and the fan blades were examined. In concert with the FAA's previous observations, no scoring of the shroud surface and no abrasion signatures on the blade tips were noted. The blades appeared crushed against the shroud. The antitorque cables and their respective pulley assemblies were examined, and control continuity was confirmed between the pedals and the tail rotor assembly. MEDICAL AND PATHOLOGICAL INFORMATION On January 14, 1997, an autopsy on the pilot was under the direction of the Chief of General Support Division, Commonwealth Health Center, Vital Statistics, Saipan. Results of the FAA's toxicology tests on the pilot were negative for carbon monoxide, ethanol, and all screened drugs. TESTS AND RESEARCH Partial teardown examinations were performed of the engine and accessories. Also, follow-up laboratory examinations were performed on the helicopter's control rod and right magneto. Engine Examination. In summary, the partial teardown examination of the engine revealed no evidence of preimpact failures or malfunctions of any observed internal components. The crankshaft was rotated, during which procedure the continuity of the valve and gear train was observed, compression was obtained in all cylinders, all screens were found clear of metal particles, and evidence of lubricating oil was observed. (See the Lycoming Engine participant's report for complete details.) Accessory Examination. The operator reported that it had inadvertently failed to inspect the engine driven fuel pump pursuant to Lycoming's Mandatory Service Bulletin #525A which was adopted as compulsory by FAA airworthiness directive. The pump was partially disassembled and examined. The diaphragm was found with a nonpenetrating check in its surface, but its integrity was not compromised. The Bendix Servo Fuel Injector was examined under the FAA's supervision at the manufacturing facilities of Precision Airmotive Corporation. The unit was found impact damaged and environmentally altered due to salt water immersion. To the extent that it could be functionally tested, no discrepancies were reported and no bogus parts were found. Both of the magnetos were found bolted to the engine on their respective mounting pads. The attachment point fixture consisted of an engine stud/bolt which protruded through a slot in the magneto pad allowing for rotation of the magnetos in clockwise or counterclockwise directions to set their timing. Evidence of score marks was observed on the mounting pads in the area of the attachment bolt, and the timing of each magneto was found set to about 31 degrees BTDC. The correct magneto to engine timing is 25 degrees. No evidence was found indicating whether the magneto's observed attachment position had resulted from impact forces or occurred during the postcrash handling. Both magnetos were removed from the engine and taken to a repair station where they were examined and functionally tested. During the tests, no discrepancies were noted between the operation of the magnetos and the requisite performance requirements described by Teledyne's service manual. While examining the right magneto, a nonaircraft part was observed screwed finger tight into the breaker cover's orifice for the "P" lead. The part was identified by the repair station technician (and subsequently by the operator) as being a mechanic's timing jumper. The technician further reported that the timing jumper is customarily used to facilitate the timing operation because it eliminates the necessity of removing the breaker box cover. The jumper was dimensionally checked against a correct part and was found to be shorter in length. The operator reported that it routinely used the jumper in its engine maintenance operation. Magneto Operation. The operator's Director of Maintenance stated that if the helicopter had been operated with a broken "P" lead, the ignition would not shut off and the magneto would be "hot." If a "P" lead was physically disconnected, the magneto would ground out and would not function. If the "P" lead in the helicopter were to break, a timing adapter could be placed in the magneto and connected to the helicopter's broken "P" lead wire. Then the "P" lead would be functional and, according to the Director of Maintenance, it would work and you could fly the helicopter. Normally, when a timing jumper is inserted into the magneto it is only finger tight. Laboratory Magneto Examination. The right magneto was submitted for laboratory examination to ascertain if electrical arcing had occurred between the timing jumper assembly and the adjacent components. In summary, the laboratory reported finding a significant amount of arcing events did occur during magneto operation. The time at which these events occurred could not be established. The laboratory further noted that, absent deficiencies in the operation of the left magneto, the right magneto's arcing (grounding out) event would have caused only "some slight momentary reduction in the engine rpm." Laboratory Rotor Head Assembly and Push-Pull tube Examination. The rotor head assembly, the main rotor grips, and the push-pull tubes (control rods) were examined in the laboratory. Evidence of attachment bolt failure in shear was found in the rotor system. At the lower end of one of the push-pull rods,
loss of engine power due to fuel exhaustion, resulting from improper planning/decision; and the pilot's abrupt handling of the flight controls, while maneuvering to avoid swimmers, which resulted in main rotor blade contact with the tail boom. A related factor was: the presence of swimmers in the emergency landing area, which likely motivated the pilot to alter his approach for a forced/autorotative landing.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
