NTSB Factual Report

HISTORY OF FLIGHT

On August 25, 2017, about 1850 Pacific daylight time, a Robinson R-44 Raven II helicopter, N61NM, was substantially damaged when it encountered a dynamic rollover during a forced landing in Eldorado Hills, California. The commercial pilot and his passenger received minor injuries. The helicopter was registered to A M Helicopter Services, LLC and operated by a private individual as a personal flight, which was conducted under the provision of Title 14 Code of Federal Regulations Part 91. Visual meteorological conditions prevailed and no flight plan was filed for the local flight that departed Auburn Municipal Airport (AUN), Auburn, CA at 1830.

According to the accident pilot and airframe and powerplant mechanic, he flew the helicopter once in the airport traffic pattern and then taxied to the fuel isle. After adding 20 gallons of 100 low lead (LL) aviation grade gasoline, the pilot and his passenger, also an aircraft mechanic, departed on the accident flight. The magnetos were checked for correct operation prior to departure, but the pilot did not recall if he allowed the rpm to stabilize before he captured the measurement. They flew around a soccer field, so the passenger could take photographs and then flew to a nearby lake where they completed a practice autorotation. During the end of the maneuver, as they approached the ground, both the pilot and passenger heard the low rotor rpm horn, which was accompanied by its corresponding annunciator light. Subsequently, the pilot increased the engine power, which in turn increased the rpm.

The pilot and passenger decided to meet a friend at his home, they changed course to the nearby destination and approached his property from the north. The pilot decelerated and transitioned into an in-ground effect hover over his friend's backyard and noticed that the downwash started to blow debris into the pool. Although his friend was signaling the pilot to land on a concrete block adjacent to the pool, the pilot elected to select an alternate landing site. He flew the helicopter backwards into and out of ground effect hover over a solar panel array, which was at a lower terrain elevation, but the main rotor rpm started to decay and the low rotor rpm horn engaged. The pilot applied full power, which increased the manifold pressure to 28 in Hg and placed the rotor rpm at approximately 96%; however, the helicopter started to descend. The pilot lowered the collective, which returned the rotor rpm to the normal range. He then completed a right pedal turn and rotated the helicopter about 360° to look for a more suitable landing site, but the helicopter started sinking again. The pilot informed his passenger that he was going to set the helicopter down on the solar panel array. As the left skid impacted a solar panel, the passenger exited the left side of the helicopter and ran to a grassy area east of their position. Almost simultaneously, the helicopter ascended about 20 ft prior to descending towards the grassy area. Subsequently, the right skid impacted the ground followed by the left, which resulted in a dynamic rollover and the helicopter came to rest on the left side.

According to the property owner, the passenger of the accident helicopter contacted him to ask if they could land on his property. About 4 minutes after the owner gave them permission to land, he heard the helicopter approaching his backyard from the north. During this time, he observed single puffs of smoke coming from the helicopter every 3-4 seconds. The property owner directed the helicopter to land on the pool deck, made up of a concrete slab, but there appeared to be some confusion as he then observed the helicopter enter a hover over some foliage behind his pool. The helicopter turned 90° to the east while sliding to the left while it discharged more smoke. After the helicopter ascended over the solar panels it entered a pedal turn as it started to descend. The helicopter then turned west and north before it disappeared from the witness' view. Approximately 15 seconds later the helicopter "popped" back up into the witness' view momentarily and turned southeast, now with a constant stream of smoke coming from the helicopter. The stream of smoke ceased for a second and then continued. The helicopter settled back down out of the witness' view over the solar array and seconds later he heard it impact branches and then the ground. He recounted that the smoke color was a very dark gray color that never changed, but the engine sounded continuous throughout the accident sequence.

In a subsequent interview, the passenger stated that he did not hear an engine or rotor harmonic during the dynamic rollover. The pilot and passenger did not observe any engine roughness.

A postaccident examination of the helicopter revealed substantial damage to main rotor blades and cockpit.

PERSONNEL INFORMATION

According to the pilot, he had accumulated a total of 3,985 total flight hours and 66.5 hours in the helicopter make and model. The pilot estimated that he flew about 3 total hours in the helicopter make and model within the 90 days that preceded the accident. He further estimated that he amassed about 500 total flight hours in Robinson R-22s, but did not accrue any flight experience in the R-22 in the last 90 days.

AIRCRAFT INFORMATION

According to FAA records, the helicopter was manufactured in 2006 and registered to A M Helicopter Services on May 20, 2008. The helicopter was powered by a Lycoming IO-540-AE1A5, a direct drive, air cooled, 260 hp engine. A review of the logbooks revealed that the helicopter's most recent 100 hour airframe inspection was completed on May 21, 2017 at an accumulated time of 1,991.3 flight hours. At the time of the accident, the helicopter had accumulated a total of 2,095 flight hours.

The engine's most recent 100-hour inspection was performed by IA Maintenance, a service facility in Woods Cross, Utah. In this entry, the mechanic recorded that he performed a differential compression check, cleaned, gapped, and tested the spark plugs, inspected the magneto and tachometer points and checked engine to magneto timing. Additionally, the mechanic noted that he complied with Lycoming Service Bulletin 301B. With the exception of the hobbs times, the previous 100-hour inspection record, dated January 1, 2017, was identical to the May 2017 entry including the cylinder compression values. The helicopter owner reported that a top overhaul had never been completed.

The owner leased the accident helicopter to Utah Helicopters, a flight school, for 2 years beginning in 2015 for flight instruction to generate income before the tail and main rotor blades reached their maximum calendar life limit. According to a representative of the flight school, they accumulated about 500 hours of flight time in the accident helicopter. It was subsequently returned to the owner on August 21, 2017 at which time the pilot reported to the owner that the helicopter was underpowered. According to the accident pilot, the ferry pilot's description of the helicopter's performance was not communicated to him when the owner invited him to fly the helicopter after it arrived.

The two pilots that ferried the accident helicopter from Utah to California had accumulated about 300 hours of flight time in the helicopter when it was operated by the flight school. Both pilots commented that the helicopter had been underpowered since the school received it, but the helicopter passed inspections, so they "managed" the performance deficiency.

METEOROLOGICAL INFORMATION

The recorded weather was captured from a weather station at AUN, located approximately 13.5 nm north of the accident site.

The 1855 recorded weather observation at AUN included wind from 250° magnetic, wind calm, visibility 10 statute miles, clear skies, temperature 32° C, dew point 26° C, and an altimeter setting of 29.86 inches of mercury.

WRECKAGE AND IMPACT INFORMATION

Airframe and Tailcone Damage

Scuff marks were observed on both the nose and belly of the helicopter and longitudinal scoring was observed on both the chin and helicopter belly. The left side of the roof and left door frames sustained impact damage along with the forward edge of the main fuel tank and the left windshield was shattered.

A small dent was observed in tailcone bay no. 4 on the left side and another small dent on the upper surface of tailcone bay no. 5 forward of the strobe light. The strobe light lens was shattered and the left side navigation antenna was damaged. The mid-span section of the upper vertical stabilizer displayed leading edge damage and trailing edge damage was observed on the horizontal stabilizer. The tail skid was bent upward with core scoring on the lower surface and the tail rotor guard was broken at the stabilizer, which occurred during the recovery process as photographs taken at the accident site show the tail skid in its original position and the tail rotor guard unbroken.

Flight Controls

Cyclic control movement was confirmed from the cockpit to the main rotor blades and collective control was verified from the collective to the main rotor blades through the swashplate. A main rotor blade pitch change link was broken, but the fracture surface was consistent with overload separation. All three main rotor servos operated smoothly when the cyclic and collective were actuated by hand. The anti-torque pedals were smooth and continuous from the cockpit to the tail rotor.

Drive System

Both the main rotor and tail rotor drive/output shafts rotated smoothly and oil was present in the gearboxes. The V belts were in good condition and in place on both sheaves. Continuity of the main rotor and tail rotor drive systems were confirmed from the lower sheave to the main rotor blades through the main rotor drive shaft and from the upper sheave to the tail rotor blades through its respective drive shaft.

Main Rotor and Tail Rotor Blades

One main rotor blade displayed upward bowing near the tip, chordwise creasing throughout the blade span, and was bent opposite the direction of rotation. Additionally, tree residue was observed along the leading edge.

The other main rotor blade exhibited upward bowing and was bent opposite the direction of rotation. Chordwise creasing was observed throughout the span of the blade. A 5 foot section of afterbody was separated from the blade spar about mid-span and tree residue was observed on leading edge and lower surface.

Both tail rotor blades remained attached to the hub. One tail rotor blade displayed multiple dents near the leading edge on the outboard skin and coarse scuff marks along the trailing edge. Tree residue was observed on both the inboard and outboard skins on both tail rotor blades.

Fuel System

The main fuel tank and auxiliary fuel tank did not display any breaches or show any leaks. Both fuel caps were secure and the vent tubes were clear. The gascolator screen displayed trace amounts of debris and the fuel exhibited an odor and color consistent with 100 LL aviation grade gasoline.

The key was turned to the prime position to test the auxiliary fuel pump, which did not engage during the attempt.

Engine

Rotational continuity was established throughout each of the six cylinders, when the crankshaft was rotated by hand using the engine-cooling fan, which showed light compression on all six cylinders.

The ignition harness leads were removed from the spark plugs, most of the leads showed indications of arcing. The ignition lead wires to the cylinder no. 3 and 5 spark plugs were loose and the rim of the upper plug to cylinder no. 1 displayed corrosion. The spark plugs displayed excessive contamination, which consisted of pliable carbon deposits and other debris that caused electrode bridging. When compared to the Champion Aerospace Aviation Service Manual, the plugs were similar to the manual's classification of "severely fouled" plugs. The bottom spark plugs to cylinder nos. 1, 3, and 5 displayed oil fouling, consistent with the helicopter's resting position on its side at the accident site.

A magneto synchronizer was used to measure the magneto timing, which revealed a timing of 10° before top dead center. According to the engine manufacturer, the magnetos should be timed to 20° before top dead center.

Engine Test

The spark plugs were re-installed, a quart of piston engine oil was added and the helicopter was moved to an open area for an engine run. Initially, the engine was started normally and idled at 64% for about 2 minutes until the oil temperature was in the green range. During this time the engine ran rough and gray smoke streamed from the tailpipe. After the warm-up was complete, the rpm was increased to 100% and the engine continued to run rough. The rpm was then reduced to 75% and a magneto check was completed; the rpm level dropped to 61% when the ignition was turned to the left magneto and 55% when switched to the right magneto.

After the engine run, when comparing cylinder temperatures by hand, the temperature of cylinder no. 5 was cooler than the other cylinders. Additionally, the cylinder no. 3 intake pipe was found loose. The lower spark plugs were then removed and a compression test was performed on each individual cylinder at 80 psi. The results were as follows: no. 1; 28 psi, no. 2; 34 psi, no. 3; 52 psi, no. 4; 38 psi, no. 5; 36 psi, no. 6; 54 psi. During each cylinder compression test, air could be heard leaking from the exhaust.

A second engine run yielded similar results as the first test despite introducing clean spark plugs.

Valve Movement

The NTSB IIC examined the intake valve movement with the accident pilot at the owner's hangar. Initially, the pilot used the starter to turn the engine, while the NTSB IIC observed the intake valves, which confirmed that both intake valves for cylinder nos. 5 and 6 extended only about one-third the total travel of the other intake valves in the open position. A subsequent visual inspection of the camshaft revealed that the intake valve cam lobe to cylinder nos. 5 and 6 exhibited severe wear and spalling (see photograph 1).

Photograph 1: Intake Camlobe for Cylinder Nos. 5 & 6

ADDITIONAL INFORMATION

Utah Helicopters

According to two instructors from Utah Helicopters who consistently flew the accident helicopter, they reported the performance issues during company meetings and directly to the mechanics who serviced the helicopter. At the time, the company did not employ a full time mechanic and used contractors who were not consistently available. A maintenance flight was usually performed by one of the instructors following service who would experience the same performance issues. One of the instructors reported that a mechanic eventually told him that the engine was "old and tired" and was not likely to produce more power. The instructor eventually accepted this justification as the statement came from the certified airframe and powerplant mechanic, a credible source. As a result of the performance deficiency, the instructors would manage the problem by flying with a reduced load and in areas of lower terrain elevation.

According to the mechanic who performed the helicopter's most recent service, at the time the work was performed he was performing about 12 inspections a year to keep his inspection authorization current. Additionally, at the time his work as a mechanic was completed after working hours or on weekends as he was employed with a U.S. government agency in a full-time capacity, but not as a mechanic. The mechanic normally verified valve movement through a compression check rather than a visual inspection of the rocker arms while rotating the crankshaft. He further stated that he had not complied with service bulletin 301B during inspections in the past several years as it was not a requirement and the owner of Utah Helicopters had instructed him not to do any additional maintenance unless it was required.

Lycoming Service Bulletin 301B

The engine manufacturer issued a mandatory service bulletin in 1977 to serve as a standard for all valve maintenance procedures. According to section (1)(b) of the bulletin, the engine rocker box c

NTSB Probable Cause

Maintenance personnel's inadequate routine maintenance on the helicopter engine, which led to low cylinder compression, improper magneto timing, and camshaft lobe wear, which subsequently resulted in degraded engine performance.