NTSB Factual Report

THIS CASE HAS BEEN MODIFIED OCTOBER 30, 2007.

HISTORY OF FLIGHT

On October 8, 2006, approximately 1600 Pacific daylight time, a Robinson R44 helicopter, N769RT, impacted the surface of Lake Tapps, near Sumner, Washington, and sunk in about 20 feet of water. The private pilot received fatal injuries, his passenger received serious injuries, and the aircraft, which was owned and operated by the pilot, sustained substantial damage. The local 14 CFR Part 91 personal pleasure flight, which departed the same area about 50 minutes prior to the accident, was being operated in visual meteorological conditions. No flight plan had been filed, and there was no report of an ELT activation.

According to the passenger, the flight was "kind of a spur of the moment event." She stated that the fact that the flight had not been a pre-scheduled event was not unusual, and that it "happened often." She said that she and the pilot had been talking, and jointly decided to go take a look at some property that the pilot was "developing" in the general area. After departure from an area near the pilot's home, the flight proceeded south about 10 miles to the Orting/Carbon River area, where the pilot took about five minutes to look around. After that, he headed northeast for about seven miles to the town of Buckley, where he over-flew some property he wanted to observe. He then headed west, paralleling Highway 410, until he got to the Meyers Road area. At that point, the pilot over-flew the house of a friend. He then told the passenger that he was going to land in an area near the house, but after the passenger told him that she was uncomfortable with the landing zone, he elected not to land. He then decided to proceed back to the original point of departure.

According to the passenger, when they came back to the general area from which they departed, the pilot maintained approximately the same altitude he had been at en route, until he reached an area about one-half mile north of the landing site. At that time he turned south, toward his home (and landing site), and started a descent that brought them in low over the trees that are on a ridge just to the south of 16th Street East (see map). According to the passenger, as they continued the descent toward the landing site, the pilot suddenly made a rapid movement of the stick (cyclic) toward the left. This was followed almost immediately by the helicopter "jerking violently" to the right (clockwise rotation). The helicopter then seemed to her to stabilize for one to two seconds, during which time she asked the pilot what was wrong, and he responded with, "I don't know." The helicopter then started spinning to the right and descending. The passenger believes that the helicopter spun around two full times, and then impacted the water "very hard." At the moment of impact, she felt something "snap or break" in her back, and then the aircraft almost immediately filled with water and sank rapidly to the bottom. After it came to rest on the bottom of the lake, the passenger looked and felt for the pilot, but could not determine where he was, so she unlatched her seatbelt, pushed up through an opening, and swam to the surface.

In a post-accident interview, the passenger said that just about the time the helicopter started to spin, she saw a number of "warning lights" come on, but she did not remember hearing any aural warnings. She further stated that she was not aware of any changes in engine rpm, nor did she remember hearing anything unusual from the engine. She did report during the interview that she was later told by her daughter and another person who were standing near the landing site watching them approach, that the engine had made a couple of loud "pops" just before the helicopter began spinning.

A number of witnesses on the ground observed the aircraft as it arrived in the area and proceeded toward the landing site. In addition to the passengers daughter, six of those witnesses stated that they heard the engine "backfire" or "pop" between one to three times just prior to, or during the aircraft's spiraling descent. Most of the witnesses that heard the noises said that there was a sudden decrease in the engine rpm just after the sequence of rapid popping or backfiring. One witness reported that the engine rpm had suddenly increased just prior to the popping, and then rapidly decreased immediately after the reported noises. One witness said that the noise was either a pop or a loud "clunk." This same witness said that the engine momentarily "revved higher" just after the pop, and then the rpm decreased, and that the engine did not come to a stop. None of the witnesses reported seeing anything separate from the aircraft during the sequence of events. One witness said that he clearly saw the main rotor blades slow dramatically or stop at the beginning of the descent, and that he heard the engine "stop running." One witness said that to him the sound was more like a "mechanical pop" than an engine backfire. A number of the witnesses that went out in the water to try to help the occupants reported a strong gasoline smell around the impact area. A number of witnesses estimated that the helicopter was about 100 feet above the water when the accident sequence began.

WRECKAGE AND IMPACT INFORMATION

The helicopter came to rest on the muddy bottom of the lake in about 20 feet of water. It was recovered from the lake the day after the accident, and taken to AvTech Services, LLC, in Maple Valley, Washington, where it underwent a post-accident inspection. That inspection revealed that the airframe was basically intact, but severely distorted due to hydraulic impact damage. There were numerous failures of flight control torque tubes, all of which were consistent with impact sequence overload. The tail cone was still attached to the aircraft body, and the empennage was still attached to the aft tail cone bay. The lower surface of both the aircraft body and the most-forward six bays of the tail cone showed severe hydraulic crushing. The empennage was intact, but the stabilizer was bent slightly downward. All four of the rubber drive belts were intact in their respective pulley grooves, and none showed any signs of rolling or slipping. The belt tension actuator was fractured, but otherwise appeared normal. The sprag clutch looked and functioned normally, and the main rotor gearbox was undamaged and rotated freely. The tail rotor driveshaft had been bent during the impact sequence, but it showed no indication of pre-impact rotational damage. Both the intermediate and aft flex couplings were intact, and the tail rotor gearbox was intact and free to rotate. The tail rotor gearbox was drained, and contained mostly water, with numerous small droplets of blue lubricating oil. The tail rotor output shaft, hub, blades, pitch change links, and pitch change slider were all undamaged and were free to function normally.

After the airframe and engine inspection at AvTech Services were completed, the engine was sent to Textron Lycoming production facility in Williamsport, Pennsylvania, where it underwent a Federal Aviation Administration monitored test run on the production automated engine test stand. The engine test run sequence lasted for a total of 31 minutes. During the test sequence, the engine was run at 1,500 rpm for five minutes, 1,800 rpm for five minutes, 2,200 rpm for five minutes, an additional 2,200 rpm for one minute (for magneto check), full throttle (2,783 rpm and 25.5 inches of manifold pressure) for ten minutes, and 600 rpm (idle) for five minutes. At each test rpm, data was recorded for allowable speed variation, and cylinder head temperature. At full power and at idle additional measurements were made for oil pressure, and oil gallery temperature. At full power, fuel pressure was also recorded. Except for one parameter, oil pressure at full power, all recorded data points were within acceptable limits, including rpm drop-off for single-magneto operation at 2,200 rpm. The lowest oil pressure measurement made during the 10 minute full power run (74.8 psi) was two-tenth of a pound per square inch below the low limit of 75 psi. The lowest measured oil pressure during the five minutes the engine was at idle (52.0 psi) was 17 psi above the low limit of 35 psi.

As part of the investigation, the carburetor was taken to Precision Airmotive in Marysville, Washington, where it underwent a series of NTSB-monitored flow tests and inspections. The first test was performed with the carburetor in the as-recovered condition. No attempt was made to remove any of the contamination that may have entered the carburetor bowl/system through the discharge nozzle or idle circuit during the accident sequence or while it was at rest on the bottom of the lake. Additionally, the mixture control arm, which had been bent during the accident sequence, and which kept the mixture control from being moved to the full rich position, was left on its shaft during this initial test. This kept the mixture control from being moved to the normal full-rich testing position. Also, during this test the throttle arm could only be moved to about 80 percent of the full-power position. This was due to the fact that impact damage had resulted in a slightly bent throttle valve shaft and damage to the accelerator pump, thus restricting the movement of both. Therefore, the full-power position could not be evaluated during the initial test series. All "fuel flow" measurements taken during this test were leaner than the lean limit at all tested airflow settings.

After the fist test, the carburetor was disassembled, inspected, cleaned, and reassembled. During this process, small particles of both organic and non-organic material were removed from the discharge nozzle, float valve seat, and carburetor bowl (no contamination was found in an earlier examination of the carburetor bowl fuel inlet finger-screen filter) The carburetor was then flow-tested a second time, but still with the bent mixture control arm in place. During this test, the throttle lever was moved to the full-power position, although with some resistance. All "fuel flow" measurements taken during this test were leaner than the lean limit at all tested airflow settings.

After the completion of the second test, the bent mixture control arm was removed, an undamaged replacement arm was installed (allowing movement to the full rich position), and a third test was conducted. During the third test, all flows were within both rich and lean limits.

This aircraft was factory-equipped with a carburetor heat assist device system. This system automatically changes the amount of carburetor heat application after the pilot has made the initial application, and acts inversely to the amount of collective input applied. Once the pilot applies carburetor heat to keep the carburetor air temperature (CAT) in the carburetor throat out of the yellow arc on the carburetor air temperature gage, the system provides for an automatic increase in carburetor heat when the collective is lowered, and automatically decreases the amount of carburetor heat as the collective is raised. The objective of the system is to decrease pilot workload, and to make more power available (carburetor heat reduction) when the collective is raised.

At the time of the recovery from the lake, the collective was found in the full-up position, and the carburetor heat control knob was found to be in an intermediate position between the OFF and FULL ON positions. An on-scene measurement determined that seven-eights of an inch of the control knob shaft was protruding from the panel.

As part of the investigation, the NTSB IIC asked Robinson Helicopter to estimate what position the carburetor heat knob would have been in during the cruise portion of the accident flight if the full-up collective at impact were responsible for repositioning the carburetor heat control knob to the seven-eights inch extended position. This determination was made using an estimated 1,850 pound gross weight for the aircraft, and a speed of approximately 90 knots, resulting in a collective position of about 55 percent application. According to Robinson, these parameters would result in the knob being extended about two inches during cruise flight. Two inches of extension is approximately 60 percent of the knob's allowable travel.

MEDICAL AND PATHOLOGICAL INFORMATION

On December 7, 2005, the pilot underwent an airman medical examination, and was notified by letter dated January 3, 2005, that he did not meet the medical standards of various subsections of Title 14 CFR Section 67. The letter advised him that he had a disqualifying general medical condition (depression), and that the medication he was taking (Zoloft) was unacceptable for medical certification purposes. The pilot's application for issuance of an airman medical was denied, and he was further advised that it was unlawful for him to exercise airman privileges without holding the appropriate medical certificate. The subject letter stated that the pilot could submit an application for reconsideration, but that in the event that he did not submit such a request within 30 days of the date of the letter, he would be deemed to have acquiesced in the denial, and would then be considered to have withdrawn his application for medical certification. The pilot did not submit an application for reconsideration.

ADDITIONAL DATA AND INFORMATION

The investigation was not able to positively determine the fuel load at the time the flight was initiated. According to the passenger, just prior to takeoff the pilot stated that he wanted to refuel sometime during the flight, and he asked if she had her credit card with her. She told him she did not have her card, and he replied that that was alright because they would have plenty of fuel.

About ten minutes into the flight, the passenger thought she heard a very faint noise coming from the top center of the fuselage. She described this noise as sounding like a small jewelry neck chain slapping very lightly against the exterior roof of the helicopter. She said the "slaps" occurred at about one-second intervals, and she thought she could hear it for about three or four minutes. She pointed this out to the pilot, who listened for the noise, but he did not hear any unusual sounds. Except for that three to four minute period, the passenger did not hear this noise, or any other noises that she felt were unusual. During the entire flight, including the period when the two occupants discussed the possible existence of an unusual noise, both the pilot and his passenger were wearing Bose noise reduction headsets. An inspection of the helicopter exterior roof found no indication of any object making repeated contact with its surface.

The FAA's forensic toxicology laboratory performed a forensic toxicology examination on specimens collected from the pilot, and the results were negative for carbon monoxide and cyanide in the blood, and negative for ethanol in the urine. The same examination confirmed positive results for the following:

2.028 (ug/ml, ug/g) Venlafaxine in the blood.

Detectable Venlafaxine in the urine.

Detectable Desmethylvenlafaxine in the urine.

Detectable Bupropion in the blood and the urine.

Detectable Bupropion metabolite in the blood and the urine.

0.03 (ug/ml, ug/g) Diphenhydramine in the blood.

Detectable Diphenhydramine in the urine.

Detectable Ibuprofen in the urine.

According to the FAA's Civil Aero Medical Institute (CAMI), Venlafaxine, Bupropion, and Diphenhydramine are disqualifying medications for airmen because they all have possible adverse side effects, and because of their potential to negatively effect pilot performance.

An autopsy performed by the office of the Pierce County Medical Examiner, classified the manner of death as accidental, and the cause of death was determined to be multiple blunt force injuries.

The primary wreckage was released to AvTech Service LLC, in Maple Valley, Washington, on November 6, 2006.

NTSB Probable Cause

The complete loss of engine power, for undetermined reasons, while on final approach, and the pilot's loss of control of the helicopter after the power loss. Factors include a lake below the final approach to the private landing pad.