Full Report:

LAX00GA073 HISTORY OF FLIGHT On January 11, 2000, at 2100 hours Pacific standard time, an Arrow Falcon Exporters, Inc. OH-58A helicopter, N901SF, was destroyed when it impacted terrain in an agricultural field 3 miles southwest of Crows Landing, California. The airline transport certificated pilot and one passenger were fatally injured. The helicopter, operated as a public-use aircraft under 14 CFR Part 91 by the San Francisco (California) Police Department (SFPD), was conducting a night cross-country flight. The helicopter departed from Porterville, California, at 1915, and was destined for San Francisco. Instrument meteorological conditions prevailed at the nearest weather reporting locations and no flight plan was filed. According to police department personnel, the helicopter was flown to Porterville the prior day for routine (50-hour) maintenance. Porterville is the location of Arrow Falcon Exporters, Inc., holder of the helicopter's type certificate. Personnel at Arrow Falcon said that the helicopter arrived in Porterville about 1230 on Monday, January 10. The pilot had intended to arrive earlier (about 0730), but was delayed departing by fog in San Francisco. They departed from Porterville for the return flight at 1915 on January 11th. The pilot's wife said she spoke with him between 1400 and 1600 on the afternoon of the accident, before he left Porterville. She did not recall anything in the conversation that suggested he felt he was under pressure to return. He asked her about the weather, and she told him there was a storm approaching and to wait another night in Porterville if necessary. En route, in proximity of Los Banos, the pilot advised Stockton Approach Control that he intended to navigate via interstate highway 5 at 1,000 feet altitude (msl). Radar and radio contact with the helicopter were subsequently lost. The wreckage was located the following morning. Recorded radar track data (Continuous Data Recording, CDR) was obtained from the Federal Aviation Administration (FAA) Stockton Terminal Radar Approach Control (TRACON) facility (attached). At 2052:09, the helicopter was 3 miles west of the town of Gustine, and about 1 mile east of highway 5. In the next minute the aircraft track veered northeast, away from highway 5, and, in the following minute, the track made a wide left turn to a northwesterly heading, which placed the helicopter back on a converging heading with highway 5 from about 2.5 miles east of the highway. At 2056, the track was back near highway 5 and proceeded northbound along the east side of the highway. At 2056:59, the track was approximately 1/4 mile east of highway 5 and, in the next minute, the track made a right-hand turn to a south-southeasterly heading at which time radar contact was lost near the accident site at 2058:00. In the same 6-minute period, the helicopter's mode-C altitude readout varied from 500 feet (msl), down to 300 feet, then up to 800 feet, then down to 300 feet, up to 600 feet, down to 400 feet, and then up to 700 feet before radar contact was lost at 600 feet. PERSONNEL INFORMATION The pilot was employed by the SFPD in 1974 and performed various ground-based law enforcement duties until August 21, 1999, when he was transferred to the Special Operations Division (Air Support Unit) and became a police department helicopter pilot. Prior to 1999, and unrelated to his law enforcement duties, the pilot was issued fixed wing and rotary wing pilot certificates on his own initiative. He earned a private pilot certificate in 1978; a commercial pilot certificate in 1979; an instrument rating in 1979; a multiengine rating in 1985; and an airline transport pilot certificate in 1986. The pilot also earned a flight instructor rating for single engine airplanes in 1984, for multiengine airplanes in 1985, and for instrument instruction in 1990. Prior to March, 1996, the pilot accumulated approximately 4,700 hours in fixed wing aircraft performing flight instruction, charter flying (Part 135), and personal flying. Between September 1989 and March 1990, the pilot took a leave of absence from the police department and flew with a regional airline as a first officer in twin-engine turboprop aircraft. According to a flight instructor who flew with the pilot, the pilot began training in rotary wing aircraft in 1996, on his own initiative, in anticipation of the reactivation of the police department air support unit. He desired to become a pilot in the unit, which had been deactivated since 1977. He acquired a private pilot helicopter rating in September 1996; a commercial helicopter rating in November 1997; and an instrument helicopter rating in February 1998. At the time of the accident the pilot had logged 392 total helicopter flight hours, of which 147 were in the same make and model helicopter. In the previous 6 months the pilot logged total flying time of 256 hours (165 in helicopters); 42 hours at night (31 in helicopters); and 18 hours of simulated or actual instrument time (3 in helicopters). He logged 34 instrument approaches (6 in helicopters). The pilot's training was received from a civilian school flying Schweizer 269 (piston engine) helicopters. After completing his helicopter instrument training in February 1998, with a total of 160 hours in the Schweizer 269, the pilot acquired 23 hours of Bell 206 (turbine engine) helicopter experience, flying with a traffic reporter in Los Angeles in August 1999. Following assignment to the Special Operations Division in August 1999, the pilot received an additional 17.2 hours of training during October and December 1999. The chief flight instructor at the school, who did not personally instruct the pilot, said that after the pilot obtained his helicopter ratings, he returned to the school and rented a Schweizer 269 to gain experience. It was during this time that the chief instructor had to give the pilot additional training regarding his inadequate radio procedures, and for not complying with the airport noise abatement procedure. He characterized the pilot as a "do it my way kind of guy" who was somewhat arrogant and did not accept advice from other, more experienced, helicopter pilots. The chief instructor reported that he had "some reservations" when he learned that the pilot was going to be the SFPD helicopter pilot because he was "very bold with little experience in helicopter operations." The flight instructor, who did instruct the pilot, characterized him as an aggressive, mission-oriented pilot/law enforcement officer who, although qualified in the OH-58, on several occasions flew missions requiring skills that were beyond his rotary wing experience level. The flight instructor felt that the pilot's personal commitment to the mission objective sometimes influenced his judgment. According to the instructor, the pilot was the sole pilot in the Air Support Unit and reported to police department managers who did not possess specific aviation expertise. The unit did not have, or operate in accordance with, an Operations Manual. AIRCRAFT INFORMATION The helicopter was manufactured by Bell Helicopter Company and delivered to the United States Army in 1972. The helicopter was surplused by the Army and ownership was transferred to the city of San Francisco. The helicopter was subsequently converted for civil use, and was issued a civil (Restricted Category) airworthiness certificate in accordance with the Arrow Falcon Exporters, Inc., Restricted Category type certificate. Under the type certificate, the helicopter was approved for agricultural application operations under visual flight rules only. In November 1999, because the helicopter was not being used for agricultural operations, the SFPD surrendered the Restricted Category airworthiness certificate to the FAA's Oakland (California) Flight Standards District Office, and, at the time of the accident, the helicopter was being operated as a public-use aircraft. The helicopter retained the Bell serial number but with an Arrow Falcon sequence number. N901SF's sequence number was -6, its production date was May 6, 1997, and delivery date was June 16, 1997. The helicopter's pitot-static, altimeter, and transponder systems were tested and certified in accordance with 14 CFR Part 91.411 and 91.413 requirements on November 9, 1999. The helicopter was equipped with a wire strike protection kit installed by Bell Helicopter Company during manufacture of the helicopter. METEOROLOGICAL CONDITIONS At 1628, the pilot contacted the Rancho Murrietta Automated Flight Service Station (AFSS) and received a (pilot) abbreviated weather briefing. The pilot was advised that instrument meteorological conditions prevailed along the route of flight and that visual flight rules were not recommended. The pilot initially asked the briefer for surface weather at Oakland and San Francisco. The briefer replied "there's a slew of weather advisories, did you want me to give you all those, AIRMETS and SIGMETS?" The pilot declined to receive the advisories. The San Francisco surface observation was for scattered clouds at 1,000 feet, broken clouds at 2,000 feet and overcast ceiling at 4,5000 feet with visibility of 2.5 miles in mist. The Oakland surface observation was for broken clouds at 1,000 feet, and overcast ceiling at 2,000 feet with visibility of 1.5 mile in light rain and mist. The briefer then volunteered the San Jose surface observation, which was a few clouds at 1,900 feet, a broken ceiling at 3,500 feet, and an overcast ceiling at 4,700 feet with unrestricted visibility. As the briefer was reading the San Jose observation and said there were few clouds at 1,900 feet, the pilot interrupted him and said "1,900 feet at San Jose, oh good," and after the briefer finished the observation, the pilot said "so San Jose is basically staying VFR." The briefer then volunteered the San Francisco and Oakland terminal forecasts, which both forecast (until 2000) scattered clouds at 1,500 feet, broken clouds at 2,500 feet, and broken (Oakland) or overcast (San Francisco) clouds at 4,000 feet with visibilities of 5 miles or greater with occasional conditions of broken clouds at 1,000 feet, overcast clouds at 2,000 feet, and visibility of 2 miles (Oakland) or 4 miles (San Francisco) in light rain showers. The briefer then further volunteered that AIRMETS were in effect for mountain obscuration and IFR conditions and that "VFR would not be recommended." The pilot then inquired about weather reports along the western side of the San Joaquin valley, and the briefer replied that there were only a limited number of reporting points. The Merced surface observation, an automated (AWOS) report, was scattered clouds at 2,000 feet, broken clouds at 3,700 feet, and overcast clouds at 4,500 feet with visibility of 1.75 miles in light rain and mist. The pilot expressed concern about the visibility, and the briefer noted that several special observations from Merced in the last hour showed the visibility varying between 1.25 and 3 miles. Naval Air Station Lemoore was reporting scattered clouds at 8,000 feet, broken clouds at 15,000 feet, and visibility of 7 miles. The pilot then asked for the Livermore surface observation. The briefer replied that 20 minutes earlier a special observation (AWOS) was broken clouds at 2,800 feet, and overcast clouds at 3,700 feet with 6 miles visibility in light rain and mist. Over the previous hour the visibility had varied between 2.5 and 6 miles. The pilot then replied "sounds good," and stated that he intended to fly via the San Luis reservoir to San Martin and then to San Jose and San Francisco. The briefing ended at 1641. At Modesto, 18 miles north of the accident site, at 2053, the weather was: a few clouds at 200 feet (agl), broken clouds at 2,500 feet, and overcast clouds at 3,000 feet with visibility of 2 statute miles in fog; temperature 51 degrees Fahrenheit; dew point 50 degrees Fahrenheit; and the altimeter setting was 30.12 inHg. At 2113, the weather was: scattered clouds at 2,900 feet, overcast clouds at 3,400 feet; and the visibility, temperature, and dew point were unchanged. The altimeter setting was 30.13 inHg. At Merced, 30 miles east of the accident site, at 2141, the weather was: scattered clouds at 2,700 feet, broken clouds at 3,500 feet, and overcast clouds at 4,500 feet; visibility was 2 statute miles in fog; temperature 51 degrees Fahrenheit; dew point 50 degrees; and the altimeter setting was 30.14 inHg. An airman's meteorological notice, AIRMET SIERRA, was in effect for the area where the accident occurred. The AIRMET called for areas of instrument flight conditions and mountain obscuration until 0100 the following morning. Cloud ceilings were forecast to be below 1,000 feet agl, and visibilities less than 3 statute miles in precipitation, fog, and mist. A witness, who lives about 4 miles southeast of the accident location, reported that, near the time of the accident, there was a low overcast ceiling with local areas of fog, and that it was a very dark night. COMMUNICATIONS The Safety Board investigator listened to a re-recording of communications between the pilot and Stockton Approach Control (Castle Radar at Stockton Air Traffic Control Tower). The tape was provided by the FAA's Southwest Region Quality Assurance Office. The times are approximate based upon elapsed time from the start of the recording. At 2029, the pilot contacted Stockton Approach Control and reported his position was 20 miles north-northwest of [NAS] Lemoore, en route to Oakland, and requesting flight following at 1,500 feet msl. The controller issued a transponder code (5125) and subsequently radar identified the helicopter 4 miles south of Firebaugh and issued the Stockton altimeter (30.12). The pilot questioned the position south of Firebaugh and the controller verified the position relative to Firebaugh airport. At 2032, the controller asked the pilot to verify his altitude and the pilot replied "900 climbing back to 1,200". At 2034, the pilot requested a radio check and the controller replied loud and clear. The pilot said, "we just changed radios, that's all." At 2040, the pilot asked "could you give us a steer for Los Banos," and the controller replied that the heading was 300 degrees and the distance 6 miles. The pilot acknowledged "6 miles at our 12 o'clock, 300 degrees, thank you." At 2043, the pilot requested the current altimeter setting for Stockton and the controller issued the setting (30.12). At 2044, the pilot asked the controller "you show Los Banos at one o'clock, one mile," and the controller replied "No, actually I show it just off your right, abeam, 3 o'clock, a mile and a half". The pilot replied "good call, we're follow highway 5. We did an update on the weather and it looks like we're gonna have to go all the way to Concord and come back to Oakland." The controller replied that the Livermore weather was 1,400 foot broken ceiling, 4,700-foot overcast, and visibility 10 miles. The pilot replied, "OK, Livermore is 1,400 broken, 1,700 overcast and visibility 10. How's Oakland?" At 2045, the controller replied that the Oakland weather at 1953 was wind 240 degrees at 7 knots; visibility 12 miles; 2,500 foot scattered clouds, 5,000 foot broken clouds; and the altimeter setting was 30.19 inHg. The pilot replied, "OK, very good, if I can get through the Livermore pass, the Altamont, I'll proceed on into Oakland." At 2051, the pilot told the controller that he was going to parallel highway 5 at 1,000 feet. This was the last communication from the helicopter. At 2100 hours, the controller transmitted three times "01 Sierra Fox, Stockton Approach" without receiving a reply. WRECKAGE AND IMPACT INFORMATION The accident location was in a level, unplanted, recently plowed, dry, agricultural field on the western side of the San Joaquin Valley in central California. The latitude was 37 degrees 21.57 minutes north, and the longitude 121 degrees 05.54 minutes west (GPS). The elevation was approximately 170 feet msl. To the west and northwest, the terrain slopes gently upward to the coastal mountain range. Interstate highway 5 was visible about 2 miles west and approximately 75 feet higher along the mountain foothills. All of the helicopter was present at the site and there was no fire. The wreckage was distributed along a linear path approximately 350 feet long oriented approximately 210 degrees (magnetic) with several rotor blade leading edge sections as far as 600 feet from the initial impact point. At the northeast end of the wreckage path was a hole and an area of freshly disturbed dirt approximately 8 feet wide, 20 feet long, and 8 inches deep. Imbedded in the ground at the southern perimeter of the hole were the right landing gear skid, the front portion of the left skid forward of the cross tube attachment, and the front cross tube. When viewed from the side the front cross tube was deformed in the right-skid-aft direction. The nose end of the left skid was embedded approximately 1 foot into the ground and that of the right skid approximately 2 feet into the ground in the center of the area of disturbed dirt. When viewed lengthwise, the skids exhibited approximately 10 degrees of upward bending at the forward (nose) end. Adjacent to the area of disturbed dirt, approximately 5 feet west, was another hole resembling a trough about 2 feet wide, 1 foot deep, and approximately the length of one main rotor blade. The trough was oriented approximately north-south and contained small pieces identifiable with the tip of the main rotor blades. The fuselage was destroyed and the flight control system was separated at many locations. All of the flight control connecting hardware was in place, and the fracture surfaces in the flight control linkage exhibited a homogenous shiny gray appearance and were accompanied by bending. Approximately 50 feet further along the wreckage path was the tail boom assembly with the tail rotor, tail rotor gearbox, and drive shaft. The tail rotor blades were intact, however, one blade was bent approximately 20 degrees inboard at midspan. The tail rotor, gearbox, and drive shaft rotated freely by hand, and the pitch change controls were continuous up to the point of separation from the forward fuselage. The individual disks of the tail rotor drive shaft flexible couplings were splayed apart between bolt attachment locations and the drive shaft exhibited a 45-degree twisting deformation at the separation from the forward drive shaft segment. The right-hand horizontal stabilizer was separated from the tail boom and was lying on the ground nearby, as were pieces of the upper portion of the vertical fin. The left horizontal stabilizer and the lower portion of the vertical fin were intact. The cockpit instrument pedestal was located at 90 feet along the wreckage path. Most of the instruments, switches, and circuit breakers were destroyed. The attitude gyro instrument was destroyed and the case was found empty except for one gimbal bearing. The attitude gyro rotor was located loose on the ground, and, when examined under a magnifying glass, did not exhibit circumferential scrape marks. The directional gyro indicator was jammed reading 210 degrees. The master caution and low rotor rpm annunciator lights were destroyed. The annunciator light panel was sent to the Safety Board laboratory for filament analysis and the findings are presented in the section entitled "Tests and Research." The altimeter (Kollsman) setting on the aircraft's altimeter was 30.10. The main rotor hub assembly, with the white dot rotor blade and the root end of the red dot blade attached, was located 190 feet along the wreckage path. The main rotor drive shaft was separated immediately below the rotor hub. The drive shaft separated at a right angle to the shaft without crippling or appreciable bending, and the circumference of the shaft exhibited a 45-degree sheer lip. Inside the hub area, above the separation, paint on the drive shaft was visibly separating from the shaft in a helical spiral pattern. When viewed from the tip, the white dot blade was bent smoothly aft along its span about 12 inches and was bent upward about 3 feet over the outer half-span. Approximately 2 feet of the root end of the red dot blade remained attached to the hub. The remainder of the red dot blade leading edge was located, in 1- to 4-foot sections, over a (roughly) 600-foot linear path extending approximately 180 degrees from the trough-like soil disturbance at the northeast end of the wreckage field. The leading edge sections exhibited torsion and bending deformation with chordwise abrasions extending into the painted area on the blade surface aft of the leading edge. The honeycomb portion of the blade aft of the leading edge was scattered over a large area, typically in 6-inch size pieces. At 210 feet along the wreckage path was the engine and main rotor transmission, which remained attached to a section of the fuselage ceiling structure above the baggage compartment. The transmission output shaft and the accessory drives turned freely and smoothly when rotated by hand. The main rotor output shaft was bent approximately 5 degrees. The engine exhibited impact damage and would not rotate by hand. The blades of the first stage compressor, viewed through the inlet, exhibited modest foreign object damage, and the turbine exhaust area, viewed through the exhaust nozzle, contained brown soil resembling that of the field. The engine fuel nozzle was clear. The fuel filter was clear and the filter and fuel lines to the engine contained a clear amber fluid, which smelled like jet fuel. The airframe structure between the engine and transmission was deformed and the engine-to-transmission drive shaft was separated at the K-flex couplings. Gouge marks were present in the airframe structure adjacent to the K-flex coupling locations. The fracture surfaces of the K-flex fingers exhibited a homogenous, dull gray appearance. The flight control servo actuators were on the ground near the transmission. The pilot valves operated freely by hand and the output shafts of all three servos were bent. The surface of the shafts exhibited a chrome-like, shiny appearance. The hydraulic pump was broken at the mounting base and was separated from the transmission. A clear red liquid resembling hydraulic fluid came from the hydraulic pump when the shaft was rotated by hand, and the hydraulic reservoir was approximately 2/3 full. The hydraulic filter bypass buttons were set (not popped). Approximately 10 feet from the engine and transmission was the fuel bladder, which remained intact but was punctured near the refueling port. The bladder was about 1/3 full of a clear amber fluid resembling jet fuel. Abeam the engine and transmission, but 30 feet to the right, was the aft portion of the left landing gear skid and the aft cross tube assembly. At 350 feet along the wreckage path was the aircraft's battery. According to Arrow Falcon personnel, the occupants of the helicopter were not wearing helmets when they departed from Porterville. No helmets were observed in the wreckage. The engine from the helicopter was disassembled and inspected on February 24, 2000, under supervision of the Safety Board investigator, at the facilities of Rolls-Royce Allison in Oakland. When viewed from the rear, the engine aft of the accessory gear case (the hot section) was deformed to the right approximately 20 degrees and the compressor discharge tube on the left side of the engine was separated at the forward coupling while the tube on the right side exhibited compression buckling. When the stand was rotated, approximately 2 cups of clean amber oil ran from the engine. When viewed from the rear, the tail rotor output shaft was twisted off about 1 foot aft of the accessory case. The twist deformation was in the shaft rotation clockwise direction. The external lines (plumbing) and most accessories were damaged. The power turbine governor was broken at the attachment to the power lever; however, the input shaft from the accessory section turned freely. The fuel control unit case was cracked at the base and the input shaft was bent and did not rotate. The case of the compressor bleed valve was fractured. The engine driven fuel pump was intact and the filter was clean. The engine sump oil collection container contained clean oil and no metal. The two magnetic particle plugs were clear. When the hot section of the engine was separated from the forward (cold) section, the turbine to compressor coupling tube was broken about 1 inch from the forward end. The fracture exhibited modest compression buckling and a bent lip outboard on one side and a straight tensile tear on the opposite side. When the hot section was disassembled there were bright, shiny abrasion surfaces on the forward face of the number 3 turbine wheel and the aft face of the number 3 stator; between the forward face of the number 3 stator and the aft face of the number 4 turbine wheel; and between the forward face of the number 4 turbine wheel and the aft face of the number 4 stator. The abrasions were accompanied by local areas of heat discoloration and areas of metal smearing on the number 3 turbine and number 3 stator. Additionally, there were circumferential scrapes on the inner radius of the number 3 stator shroud over approximately 180 degrees. When the compressor assembly was disassembled there were bright, shiny abrasions over the entire circumference of the compressor shaft between the first and second compressor disks and between the second and third compressor disks. The tips of the stator vanes at these compressor stages exhibited a similar bright, shiny appearance with sharp edges. When the impeller shroud was removed, there were circumferential score marks, areas of rough abrasion, and local areas of metal smearing on the inner surface over approximately 200 degrees of rotation. The tips of the impeller blades opposite the markings were bright, shiny, and worn flat with sharp edges. After removal of the turbine and compressor assemblies, the N1 and N2 gear trains rotated freely in the accessory case, and the free wheeling (Spragg) clutch locked and released when rotated by hand in opposite directions. The three flight control actuators (left and right cyclic and one collective), the hydraulic pump assembly, and the hydraulic filters were examined at Bell Helicopter Company in Fort Worth, Texas on April 17, 2000, under the surveillance of a Safety Board investigator. Damage to the three actuators precluded functional testing; however, disassembly inspection revealed no visible anomalies. The hydraulic pump and associated on/off switch functioned in accordance with the manufacturer's specifications for serviceable pumps. Both the pressure and return fluid filters were clear and their respective bypass buttons were not popped. MEDICAL AND PATHOLOGICAL INFORMATION Autopsies of the two occupants were performed by the Stanislaus County, Office of the Coroner, cases 00-01-20 and 00-01-21. A toxicological analysis of the pilot was performed by the FAA's Civil Aeromedical Institute (attached). TESTS AND RESEARCH The helicopter's annunciator light panel assembly was examined for stretched light bulb filaments at the Safety Board laboratory in Washington, D.C. The engineer's report, attached, states that both light bulbs in the sector labeled "HYD PRESS" were intact but stretched. Both light bulb filaments in the sector labeled "INST INVERTOR" were a little stretched and broken in many parts. No other light bulb with a stretched filament was found. The light bulbs in the sector labeled "FUEL BOOST," the two "SPARE" sectors, and the previously discussed sector labeled "HYD PRESS" contained intact filaments, while all the other filaments were broken. According to the Arrow Falcon party representative, it is not uncommon for the hydraulic pressure light to flicker on momentarily if the pilot makes an abrupt, large, flight control input. ADDITIONAL INFORMATION Other additional parties were: Mr. Barry Lloyd, San Francisco Police Department; and Mr. Victor Symonds and Mr. Donald Brooks of Arrow Falcon Exporters, Inc., Porterville, California. The aircraft wreckage was released to the San Francisco Police Department on July 7, 2000.