LAX01FA277

the manufacturer's pilot's failure to maintain yaw control and main rotor speed while recovering from an unusual attitude induced by the customer pilot's inadequate control inputs while maneuvering at low altitude. The manufacturer's pilot's inadequate supervision of the flight is also causal. A factor in the accident was the customer pilot's lack a familiarity with the NOTAR yaw control system.

A staff pilot for the manufacturer was providing a sales demonstration flight to a potential customer's pilot when yaw control of the helicopter was lost, it entered a spin, collided with the ground, and rolled over. During the flight the customer's typical aerial application pest control flight profile was to be flown. The customer pilot flew Bell 206's currently in the mission and had no experience in the NOTAR anti-torque system. The maneuver was a simulated aerial application pass followed by a turn around. The manufacturer's pilot said the maneuver was very docile and consisted of a pass down a creek bed at 60 knots and 50 feet. The customer pilot initiated a gentle cyclic pull-up to a 10- to 15-degree nose up attitude while entering a right turn as the helicopter decelerated. At the 90 degree point in the turn, the helicopter was at 40 to 50 knots with a 30-degree or less bank angle to the right. As the nose came around in the turn, the nose tucked down to about 20 degrees below the horizon. The manufacturer's pilot took over the controls and added near full aft cyclic to level the nose; at this point, he received a low rotor warning horn (indicating a rotor droop at 95 percent or lower). Coincident with the low rotor warning, the helicopter began a right yaw rate. He added full left pedal, but the yaw rate continued at what he described as a "slow pedal turn rate," eventually completing 4 to 6 complete revolutions. At this point the helicopter began descending and he added collective. He immediately got a "power" audio warning, indicating that he was exceeding the upper power limit of the engine. The right yaw rate also increased with collective input. The pilot then modulated the collective between the low rotor warning and the excessive power warning in an attempt to both control the yaw and stop the descent. As he lowered the collective and the rotor speed began to build, the yaw would slow, but the helicopter then began descending faster. As he added collective to slow the descent, the yaw rate would increase. As the helicopter neared the ground, the yaw finally stopped; however, the helicopter was translating sideways toward a berm. The helicopter touched down on the right skid against the berm and it rolled over. The manufacturer's pilot said his hands were following on the controls as the customer pilot flew and he did not perceive any unusual control inputs. He further stated that his feet were about 1 inch from the anti-torque pedals and he did not feel any inputs from the customer pilot on the pedals. Based on his extensive flight test experience in this helicopter, the manufacturer's pilot believes that the aft cyclic input to correct the nose down pitching moment induced a rotor droop, and that the droop was the initiating event in the yaw rate onset. He does not know how low the rotor speed went, but the warning is triggered at 95 percent Nr. The lower limit of the Nr green arc is at 90 percent. The NOTAR anti-torque control system uses air from a pedal controlled jet thruster nozzle on the end of the tail boom to provide anti-torque control. Additional yaw control is provided by vertical stabilizers, which are largely effective only above 20 to 30 knots. Air is supplied to the thruster by a fan driven by a power takeoff shaft from the main transmission. There is a direct relationship between the speed of the main rotor and the speed of the fan. The pilot said he believes that the initial yaw onset was because the rotor drooped low enough to slow the fan below the speed which could supply the necessary air volume to the thruster to control the yaw and they were not fast enough for the vertical stabilizers to have any effect. After the occupants had extricated themselves from the wreckage, the customer pilot asked the pilot what had happened. He replied to them that he didn't know what happened and added, "Maybe we should have turned to the left instead." The helicopter's FAA approved Rotorcraft Flight Manual (RFM) contains several warnings about low speed maneuvering. Under the heading "Unanticipated Right Yaw", the RFM section notes that an unanticipated right yaw can occur when operating at low altitude and airspeeds below 60 knots when a pilot's attention is distracted by events outside the helicopter and sufficient control inputs are not made to adjust for changing aerodynamic conditions. The paragraph states, "If no directional or cyclic control inputs are made, a nose down pitch and a right roll may follow the right yaw." The section notes that this is most likely to occur at speeds below 60 knots in uncoordinated right turns, while flying out of trim with too much right pedal, or right turns to a downwind condition. Examination of the helicopter revealed no evidence of a preimpact malfunction or failure of the control system.

1.0 HISTORY OF FLIGHT On August 14, 2001, at 0815 mountain standard time, a MD Helicopters, Inc., 600N, N70457, collided with the ground and rolled over while maneuvering during a demonstration flight, 6 miles east of Fountain Hills, Arizona. The commercial pilot was not injured, the airline transport pilot customer and two passengers sustained minor injuries, and the helicopter was substantially damaged. Visual meteorological conditions prevailed. The flight was operated by MD Helicopters, Inc., under 14 CFR Part 91 as a sales demonstration flight. A company VFR flight plan was filed for the local area flight that departed at 0745 from Falcon Field in Mesa, Arizona. 1.1 MD Helicopters Company Pilot's Statement In his written and oral statemens, the MD Helicopters company pilot stated that on the morning of the accident he was assigned to fly a marketing sales demonstration with the Brevard County, Florida, Mosquito Control District. The district's chief pilot was to fly the helicopter with the department director and the department mechanic riding along as passengers. Two specific demonstrations of the district's typical flight missions were to be conducted; personnel transport missions and the typical mosquito abatement aerial application control flight profile. The customer pilot flew Bell 206's currently in the department mission and had little experience with the MD products and none in the NOTAR anti-torque system. The pilot considered the customer pilot an experienced helicopter pilot. The pilot completed the normal preflight planning operations consisting of the weight and balance computations and the performance capability of the helicopter. The departure gross weight was 4,050 pounds and the anticipated density altitude in the practice area was about 4,000 feet. During preflight of the helicopter, the exceedance history of the helicopter was checked on the Quad indicator and the fault history verified on the engine instruments that record maintenance information. No exceedances were observed on the Master Caution Panel Lights. He had returned from a series of demonstrations in the helicopter in San Antonio, Texas, the day before, and reported that there were no write-ups on that flight and no open items prior to this departure. After starting, he hovered the helicopter over to the turf landing area and gave the controls to the customer pilot, who then performed a series of hovering turns and takeoffs and landings to get familiar with the control responses and the NOTAR system. When both of the pilots were comfortable with the customer pilot's performance, they departed for the Sycamore Creek training area. In the training area, the customer pilot conducted a series of approaches to normal landings and some takeoffs. Following this, they performed a pinnacle landing, then a confined area landing. The passengers then deplaned and a fly over was made so the passengers could evaluate the noise signature of the helicopter. After the passengers had reboarded the helicopter, the customer pilot wanted to evaluate the helicopter in the typical mosquito abatement aerial application flight profile he flies in the Bell 206. Principally the maneuver was a simulated aerial application pass followed by an aerial application type turn around. The pilot had the customer pilot conduct a thorough briefing on the maneuver until he understood exactly what the maneuver consisted of and what to expect. The helicopter gross weight was 3,900 pounds at this point, with the outside air temperature about 85 degrees, and the elevation of the creek bed approximately 1,500 feet msl. The maneuver was very docile and consisted of a pass down the creek bed at 60 knots and 50 feet. The customer pilot initiated a gentle cyclic pull-up to a 10- to 15-degree nose up attitude while entering a right turn as the helicopter decelerated. At the 90-degree point in the turn, the helicopter was at 40 to 50 knots with a 30-degree or less bank angle to the right. As the nose came around in the turn, the nose tucked down to about 20 degrees below the horizon. The customer pilot did not appear to put any corrective cyclic input into the controls and the pilot took over the controls. He added near full aft cyclic to level the nose; at this point, he received a low rotor warning horn (indicating a rotor droop at 95 percent or lower). Coincident with the low rotor warning, the helicopter began a right yaw rate. He added full left pedal, but the yaw rate continued at what he described as a "slow pedal turn rate," eventually completing 4 to 6 complete revolutions. The pilot said he believes that the left pedal application had an effect on slowing the yaw rate. At this point the helicopter began descending and he added collective. He immediately got a "power" audio warning, indicating that he was exceeding the upper power limit of the engine. The right yaw rate also increased with collective input. The pilot then modulated the collective between the low rotor warning and the excessive power warning in an attempt to both control the yaw (he still had full left pedal input in) and stop the descent. As he lowered the collective and the rotor speed began to build, the yaw would slow, but the helicopter then began descending faster. As he added collective to slow the descent, the yaw rate would increase. As the helicopter neared the ground, the yaw finally stopped; however, the helicopter was translating sideways toward a berm. The helicopter touched down on the right skid against the berm and it rolled over. The pilot said his hands were following on the controls as the customer pilot flew, and he did not perceive any unusual control inputs. He further stated that his feet were about 1 inch from the anti-torque pedals and he did not feel any inputs from the customer pilot on the pedals. Based on his extensive flight test experience in this helicopter, he believes that the aft cyclic input to correct the nose down pitching moment induced a rotor droop, and that the droop was the initiating event in the yaw rate onset. He was too occupied to look at the gages and does not know how low the rotor speed went. The warning is triggered at 95 percent Nr, and this is checked during each run-up before takeoff; it functioned correctly on the run-up at the factory pad before departure. The lower limit of the Nr green arc is at 90 percent. The NOTAR anti-torque control system uses air from a pedal controlled jet thruster nozzle on the end of the tail boom to provide anti-torque control. Additional yaw control is provided by vertical stabilizers, which are largely effective only above 20 to 30 knots. Air is supplied to the thruster by a fan driven by a power takeoff shaft from the main transmission. There is a direct relationship between the speed of the main rotor and the speed of the fan. The pilot said he believes that the initial yaw onset was because the rotor drooped low enough to slow the fan below the speed which could supply the necessary air volume to the thruster to control the yaw and they were not fast enough to have any aerodynamic effect from the vertical stabilizers. Once again he stated that the left pedal input did have an effect on slowing the yaw rate, but there was just not enough air volume to fully control the yaw. 1.2 Customer Pilot's Statement The customer pilot stated his concurrence with the MD Helicopters company pilot's statement up to the point where the helicopter began the aerial application turn. The customer pilot said his recollection was that at the 90-degree point, the nose tucked down, and the right yaw rate began immediately. He was certain that the tuck and the right yaw were simultaneous events. He said he put in nearly full aft cyclic and considerable left pedal to counteract both the tuck and the yaw rate onset, but without success in regaining control. The helicopter completed t