NTSB Factual Report

HISTORY OF FLIGHT

On December 18, 2000, at 1530 eastern standard time, an Aerospatiale (Eurocopter) SA365-N1, N89SM, was destroyed during an attempted go-around at Allegheny County Airport (AGC), West Mifflin, Pennsylvania. The certificated airline transport pilot and one onboard maintenance technician were seriously injured, while another onboard maintenance technician received minor injuries. There was no flight plan filed for the flight, which was conducted under visual meteorological conditions. The post-maintenance test flight was conducted under 14 CFR Part 91.

According to the pilot, he was conducting a track and balance flight. He completed a pre-flight inspection, checked the fluids, and ensured the cowlings were closed and secured. He started the helicopter, lifted off, and completed a hover check. After departing the airport, the onboard maintenance technicians requested that the pilot put the helicopter in an 80-knot, 1,000 foot-per-minute climb for about 30 seconds. He then flew the helicopter straight and level at 80 knots for an additional 30 seconds. He was next requested to fly the helicopter straight and level at 120 knots, and as it accelerated, he noticed that his right pedal input was not compensating for the additional power. He then applied full right and full left pedal with no response from the helicopter.

The pilot then told the maintenance technicians that he had "no rudder control," and that he was going to see what helicopter control he had while making a 70-knot approach.

The pilot called the air traffic control tower, apprised the controller of the situation, declared an emergency, and asked that fire trucks be dispatched. The tower controller advised him that he could use whatever runway he wanted, and the pilot opted for runway 28 since "the winds were not a factor with only 4 knots and 28 was the widest runway."

During the first approach, the pilot maintained airspeeds between 70 and 80 knots, and approaching the runway, the helicopter's nose was out to the right. When the helicopter was 2 to 3 feet above the runway and still descending, the pilot eased the collective down to bring the nose around for a run-on landing; however, "the nose came from right to left so fast that a safe run-on landing could not be made."

Then pilot attempted "numerous" approaches to runway 28, and then "tried runway 23 for more favorable winds," but found that the helicopter would not respond "as per tail rotor loss procedures of the flight manual." During his last approach, the helicopter "became uncontrollable and the only good thing about the final landing was that we all lived to tell the story."

One of the onboard maintenance technicians noted that, during the return to the airport, the pilot stated that he thought he had regained some control. The pilot attempted to do a run-on landing; however, when the airspeed dropped below 100 knots, the helicopter began to rotate to the left. At that point, the pilot was sure he had no control of the tail rotor.

The pilot attempted 8 or 10 approaches to runway 28, and after talking to two of the company's other pilots on the radio, he decided to try to land on runway 23. After the first approach, the pilot advised the assisting pilots that he "thought it would better on the new runway." During the next approach, the maintenance technician looked out the front windshield, when the helicopter "was almost level with the runway." When the helicopter was over the runway threshold, it was about 6 to10 feet off the ground. The maintenance technician thought the pilot was going to land, "but [the pilot] turned it hard left and leaned into it pulling collective to kill airspeed. It worked, but we started to spin 360, climbing and pitching up and down all at the same time." The helicopter then pitched "hard nose up and the ground was coming up fast." The maintenance technician then heard one of the assisting pilots on the headset yelling, "get the nose down" several times, then the helicopter impacted the ground.

According to one of the assisting pilots, the accident pilot confirmed over the radio that he had no tail rotor control. Also, after the accident pilot switched runways, he found that "the aircraft seemed to behave better during the approach with the right crosswind but the wind was less than 10 knots and did not help much." He also stated that every time he slowed below 100 knots, the nose of the helicopter would swing to the left.

The assisting pilot thought the accident pilot made three approaches to runway 23. They discussed the possibility of an autorotation, "but the recommended procedure in the RFM is the running landing." The accident pilot also stated several times that the nose of the aircraft would just not stay to the right at airspeeds below 100 knots.

During the last approach, the accident pilot appeared to slow the helicopter, and attempted a go-around. However, as he added power, the nose pitched up and the aircraft began a climbing left spin. Another assisting pilot told him to lower the collective and get the nose down "which [he] appeared to do prior to impact."

Another assisting pilot stated that the accident pilot made some low approaches to determine what airspeed the nose of the helicopter would begin to turn. "The [accident pilot]...stated that the slowest he could get before the aircraft started to turn was about 80 knots." The assisting pilot asked the accident pilot to turn off the autopilot yaw channels to eliminate possibility of an AFCS yaw actuator problem. He also advised the accident pilot of the procedure for loss of tail rotor drive/control failure in flight., which included making a shallow approach at 90-100 knots and at 15-20 feet agl. "Start a collective reduction and keep the descent going to 1-2 feet above the runway. As the nose of the aircraft aligns, lower collective and allow main and nose landing wheels to touch down, and subsequently use differential braking for directional control."

In addition:

"On the [accident] pilot's last approach, he appeared to almost touch down, but then started a go-around. He climbed and started to turn left. His speed was too slow to have the airflow over the side fins and vertical fins to be effective in counteracting the torque of the main rotor. The aircraft began spinning to the left. I transmitted to lower his nose and lower the collective. He did so and descended, and the spinning appeared to slow. The aircraft then nosed up. I did not have a good view of the actual contact with the ground. I did get permission to assist the FAA inspectors later. I observed that both AFCS yaw channel buttons were in the raised or off position, and that the yaw auto trim switch was off."

Another, non-company helicopter pilot also observed the accident. During the accident approach:

"...the aircraft slowed more than before, and it appeared the pilot was attempting to land.... When the aircraft was approximately 5 feet above the surface, with an estimated airspeed of 15-20 knots, the pilot pulled pitch in an apparent attempt to go around. The nose began to yaw left, and the aircraft entered a rapid climbing turn through approximately 150 degrees of arc and 30-40 feet of altitude. It began a rapid left spin while climbing another 50-70 feet. The aircraft made 4 or 5 complete rotations accompanied by pitch oscillations of approximately 20-30 degrees above and below a level attitude and roll oscillations of approximately 5-10 degrees. The aircraft then descended to the ground, impacting in a tail-low attitude."

According to a Federal Aviation Administration (FAA) inspector, the helicopter impacted the ground while still in a left spin, and with the nose tucked, but came to rest in an upright position.

PILOT INFORMATION

The pilot held an airline transport pilot certificate, with single engine land airplane, multi-engine land airplane and rotorcraft-helicopter ratings. His latest FAA second class medical certificate was dated July 21, 2000.

The pilot had acquired over 9,000 ours of flight time, with more than 5,000 hours in helicopters.

AIRCRAFT INFORMATION

According to company maintenance records, on December 11, 2000, the tail rotor servo was removed and replaced.

On December 15, 2000, one of the assisting pilots conducted two post-maintenance check flights on the helicopter, which included track and balance, and AFCS checks. During the second flight, there was a lateral oscillation during high speed cruise (above 140 knots), but it seemed to disappear at low airspeeds.

METEOROLOGICAL INFORMATION

Weather, recorded at the airport about the time of the accident, included variable winds at 4 knots, a few clouds at 2,500 feet agl, and a visibility of 10 statute miles. The temperature was 22 degrees Fahrenheit and the dewpoint was 10 degrees Fahrenheit.

WRECKAGE AND IMPACT INFORMATION

The post-accident on-scene investigation was conducted by FAA personnel. According to one inspector, the helicopter came to rest upright, the tail boom and main rotor system were broken away from the main fuselage, and the main fuselage was broken in half, just forward of the engine inlets. The tail rotor gearbox came to rest face-down.

In addition, the tail servo input was broken, and there was "very little" oil in the tail rotor gearbox. The maximum amount of oil that could be taken from the tail rotor gearbox "amounted to no more than a couple of cc's." The tail rotor components were retained for further examination.

TEST AND RESEARCH

On January 3, 2001, Safety Board investigators from the South Central Region, along with personnel from American Eurocopter and Corporate Jets, examined the tail rotor components at American Eurocopter's facilities in Grand Prairie, Texas.

During the examination, there was no evidence of oil leakage at the seals or the filler cap for the tail rotor gearbox. A small amount of oil was recovered, which was black in appearance. Normal coloration was brown. The oil level sight gage was marked with a black line, 6 to 7 mm below the minimum fill level.

The output drive shaft duplex bearing displayed no evidence of lubrication. The output drive shaft teflon lining was melted. The output drive shaft duplex bearing retaining ring was rotated beyond the lock tabs.

The gearbox cover oil level indication lens contained a "stain" lower than the minimum "red" line and the maximum "yellow" line. The gearbox was resting face down at the accident site.

The retention nut for the duplex bearing was "jammed", and the bearing had to be cut out of the output drive shaft for removal.

The output drive shaft contained metal shavings.

The duplex bearing cage was shifted and did not display evidence of lubrication. One of the inner races of the duplex bearing displayed signs of wear and heat damage. The outer races contained metal transfers similar to the bearing cage material. The drive shaft gear and bevel pinion did not display signs of wear and no anomalies were present.

On February 15, 2001, a joint examination by Safety Board, Corporate Jets, and Eurocopter personnel was made at Materials Analysis, Incorporated, Dallas, Texas. Materials Analysis then produced a report of the examination, which was subsequently reviewed by a metallurgist from the Safety Board's Materials Laboratory.

According to the report, the duplex ball bearing, which served to transmit thrust loads from the servo actuator shaft to the tail rotor pitch change shaft, had accumulated 1,317 hours of service in the tail rotor gearbox, which itself had accumulated 3,190 hours of service. The gearbox was splash lubricated with MIL-23699 synthetic turbine oil, and the gearbox oil change interval was 1,200 hours. The gearbox had wide fleet use, and did not exhibit a history of bearing failures.

During the initial teardown, the duplex bearing was found to be seized, "which apparently caused torsional loads from the pitch change shaft to be transferred to the stationary servo actuator shaft. These loads resulted in failure of the actuator shaft, most likely by torsional overload."

In addition,

"Smearing was evident on the end faces of the outer bearing ring, indicating that it had spun within the aluminum pitch change shaft hub. Corresponding smear marks were visible on the internal shoulder and bearing retaining nut of the pitch change shaft hub. Thermal discoloration was evident on the outer bearing ring, the inner rings, and bearing balls. The thermal discoloration appeared to be related to frictional heating when the bearing spun in the shaft hub, which also apparently caused the Teflon ring installed in the groove of the pitch change shaft O.D. to melt as indicated by the blackened remains of the Teflon ring. There was no evidence of galling or smearing on the I.D. bore of the inner bearing rings or on the mating surface of the bearing journal of the titanium servo actuator shaft. The absence of damage indicates that the inner rings of the duplex bearing had not spun about the actuator shaft.

The bearing cage appeared to manufactured from a copper alloy and exhibited considerable smearing and galling along the mid-section of the entire outer circumference. A number of the ball pockets also showed metal smearing, and two pockets were significantly deformed from ball contact. Further examination of the ball pockets in one ball revealed that one side exhibited polishing wear from rubbing against the rotating balls, while the other side of the ball pockets exhibited metal smearing. Nearly all of the ball pockets exhibited smearing and galling damage on both sides of the pocket from contact with the rotating balls. The I.D. surface of the ball cage exhibited no unusual wear or damage.

Examination of the bearing cage revealed small but significant machining burrs at the outboard corner between some of the ball pockets and the O.D. of the cage. These burrs were in non-wear locations and were obviously remnants of a machining operation during manufacture and not from wear that occurred in service. Small fractures at the corner of some pockets were noted, indicating that a metal burr had been liberated."

The report also included a discussion and conclusions. Quoted conclusions included:

1. The seizure of the duplex ball bearing from the involved tail rotor gearbox apparently occurred as a result of metal transfer from the copper alloy bearing cage to the rolling elements of the bearing. The buildup of transferred cage metal and subsequent frictional heating and thermal expansion likely eliminated the operating clearances, causing the balls to jam between the inner and outer rings.

2. The metal transfer from the cage resulted from smearing and galling between the cage pockets and the balls and the O.D. surface of the cage on the I.D. land of the outer ring. The smearing wear damage most likely resulted from ineffective lubrication. No evidence of complete oil starvation was found.

3. Small metal machining burrs broken away from the corners of the cage ball pockets during operation may have contributed to initiation of metal transfer from the bearing cage and failure of the bearing.

4. Other than the small machining burrs observed on the ball cage, there was no evidence of significant manufacturing defects nor metallurgical defects which could be associated with the bearing failure.

5. Based on the information available, the deficient lubrication of the bearing was most likely from a low oil level in the gearbox or inadequate oil flow to the duplex bearing.

6. There was no evidence to indicate that an abnormal pitch change mechanism or gearbox operation contributed to the bearing failure.

ADDITIONAL INFORMATION

There were no emergency procedures for a stuck (fixed pitch) tail rotor scenario in the helicopter flight manual.

The closest emergency procedure in effect, was for a "tail rotor failure" (a complete loss of thrust.) According to the helicopter flight manual:

"A tail rotor failure in power-on flight is indicated by a yawing motion; the rate of turn depends on the aircraft power and airspeed at the time of fail

NTSB Probable Cause

Insufficient lubrication within the tail rotor gearbox, which resulted in the seizure of the duplex ball bearing and a fixed pitch setting of the tail rotor. Also causal, was the pilot's loss of control, which was the result of a relative low speed/high power combination during a go-around attempt. A factor was the manufacturer's lack of procedures for fixed tail rotor pitch settings.