N9159FHughes Helicopters HUGHES 369D

On February 9, 2022, at 1628 eastern standard time, a McDonald Douglas MD-369D helicopter, N9159F, was substantially damaged when it was involved in an accident near Bel Air, Maryland. The commercial pilot was not injured. The helicopter was operated as a Title 14 Code of Federal Regulations Part 91 positioning flight.

The pilot stated that while conducting powerline inspection work earlier in the day, line personnel reported that the helicopter was making a strange "whistle" noise. The pilot inspected the helicopter and no anomalies were noted or observed. The pilot continued with normal operations, but the noise continued and one of the operator’s superintendents took a video, where a “whistle”-like sound could be heard. The pilot landed and ceased all human external cargo operations. He then reviewed the video, re-examined the helicopter, and spoke with company maintenance personnel. Though no obvious mechanical issues were observed, the pilot “parked” the helicopter for the remainder of the workday.

At the end of the workday, the pilot again inspected the aircraft and found no mechanical reason not to reposition the helicopter back to its normal base of operations. He and another company helicopter departed as a flight of two. Several minutes into the flight, the pilot said the ENGINE CHIP light illuminated. He told the other pilot that even though the engine seemed to be operating normally, he would need to land as soon as practicable. Shortly after, the engine began to make a “grinding” noise along with an odor of engine oil, which eventually became smoke in the aft section of the passenger compartment. With the presence of smoke and the potential for an inflight fire, the pilot initiated an emergency descent-to-land to a suitable landing area. During the descent the engine noise and smoke in the aft section of the cabin intensified and began moving to the forward section of the cockpit. Descending through the landing flare, as the pilot leveled the helicopter to land, the engine stopped producing power and smoke in the cockpit reduced his visual reference to the ground. The pilot attempted to slow the rate of descent and impacted the ground in a near-level attitude. During the ground run the front portion of the skids dug into the ground, causing the helicopter to pitch forward. The pilot applied aft cyclic to keep the helicopter level. During the landing sequence, the main rotor blades struck the tail boom, which resulted in the horizontal and vertical stabilizers and the tail rotor assembly separating from the helicopter. Residual oil was observed on the interior and exterior surfaces of the engine access doors and on the interior of the engine compartment.

The helicopter was recovered by the operator and taken to their facility in Gettysburg, Pennsylvania. Before the engine was removed, an external examination of the lubrication system between the airframe and the engine was conducted. Neither the aircraft-mounted oil reservoir nor the oil cooler were damaged, and no residual oil was noted within the cabin area (where the cooler and reservoir were located). The engine was then removed from the airframe and shipped to Keystone Turbine Services, Coatesville, Pennsylvania, where a full engine examination was conducted under the supervision of the NTSB.

Examination of the engine revealed that each of the external oil lines were secure except for the oil line which supplies pressure oil to the turbine sumps. The line connected to a horizontal fire shield and a T-fitting near the Nos. 6 and 7 bearing sump. The line was fractured at the horizontal fire shield and misaligned. Also, the clamp that secured the line to the turbine module was fractured.

The scavenge oil filter was clean and full of oil. A small amount of ferrous debris was observed on the filter. The pending bypass button was not extended. Both the upper and lower engine magnetic chip detector plugs were removed, and ferrous debris was observed on both.

The gas producer turbine rotor (N1) was seized but the power turbine (N2) turned and was connected to the powertrain. No damage was noted on the first stage compressor blades or compressor inlet. The fourth stage turbine wheel was normal in appearance when viewed from the exhaust collector.

The accessory gearbox was intact except for one fractured compressor mount pad. As the compressor was removed, the mount was found liberated from the gearbox housing. The gearbox housing and cover were split apart, and all the internal gears and bearings were intact, except for the No. 2 ½ bearing. The No. 2 ½ bearing was missing 6 rollers, consistent with having fallen out during the compressor removal, and were recovered from the bottom of the engine stand. The rollers were undamaged. The compressor bore on the gearbox cover displayed a wear step between the 1:00 to 12:00 position.

The combustion section was not damaged. The inner surface of the outer combustion case barrel was missing material and was cracked. Several pieces of barrel material were recovered from the turbine inlet. The combustion liner was covered in carbon, but no mechanical damage was observed.

Several areas within the turbine module displayed evidence of oil starvation:

• No. 8 bearing cavity

• No. 8 bearing sump oil scavenge line (was also clogged with debris/dry residue)

• No. 8 bearing sump oil supply tubes

• No. 6/7 bearing cavity

• No. 6/7 bearing scavenge orifice

• No. 6/7 bearing oil supply tube

• No. 6/7 bearing external sump can

The No. 6 bearing rollers were disintegrated, which precluded inspection of the No. 7 bearing.

The No. 8 bearing balls were also disintegrated and the remaining components (inner race, outer race, and separator) were removed. The bearing components displayed thermal signatures consistent with overtemperature operation.

The gas producer (GP) rotor tie bolt nut was loose but remained in position within the locking feature; the GP rotor was intact. The trailing edge blade tips of the first stage turbine wheel displayed mechanical damage and the upstream face of the wheel was circumferentially gouged and smeared. The energy absorbing ring location tabs and corresponding slots in the GP support displayed heavy wear.

The No. 5 bearing turned freely and smoothly and displayed some brown discoloration. The N2 rotor was removed and all airfoils were intact. The upstream face of the third stage turbine wheel was coated in a gray carbon-like substance. The N2 shaft was intact with some discoloration and carbon noted on the exterior surface. The N1 shaft was intact and exhibited some slight bulging at the end of the shaft.

The engine compressor turbine assembly, gearbox housing, oil pump, oil supply line with fittings and clamp assembly, N1 coupling, GP turbine assembly, GP turbine support (which included a sump nut, retaining ring and plate, and the No. 8 oil supply jet), No. 8 bearing, No. 8 rotating and stationary seal, No. 8 bearing spanner nut, and the outer combustion chamber, were sent to the NTSB Materials Laboratory for examination.

Examination of these components by the NTSB Materials Laboratory determined numerous instances of high heat damage and fracture due to high-cycle fatigue.

The gearbox housing support lug located on the upper left side of the gearbox exhibited fracture surfaces consistent with high-cycle fatigue. The oil pump, which was installed on the interior of the gearbox, was intact. There were no fretting contact marks noted on the pump mating to blended area on gearbox housing. Pitting consistent with cavitation damage were noted on gear teeth in the pressure body.

The oil supply line that was found fractured during the engine exam, and its associated support clamp, also exhibited fracture surfaces consistent with high-cycle fatigue. This oil line supplied oil to the Nos. 6 and 7 bearings in the turbine section. The fracture of this line most likely led to the rapid deterioration of the bearings from oil starvation.

The GP support and its related components revealed the retaining plate shear pin was fractured. The fracture surfaces showed curving crack arrest lines and dark tinting consistent with fatigue. The retaining plate was installed on the aft side of the GP support hub and had multiple recesses machined into the outer diameter, including those for accommodating the shear pin and the No. 8 oil supply jet. A mark was observed on the clockwise side of the recess for the shear pin corresponding to contact with the aft piece of the shear pin that was not recovered. A separate mark was observed near the clockwise end of the recess for the No. 8 oil jet, and a corresponding contact mark was observed on the lower inboard side of the No. 8 oil supply jet body. The marks on the retaining plate recesses corresponded to contact with the shear pin and the oil jet, respectively, as the retaining plate rotated counterclockwise relative to the GP support hub.

On the forward side of the retaining plate, damage from fretting contact with the No. 8 bearing outer race was observed near the retaining lug on the forward face next to the inside diameter. The surfaces of the retaining lug on both the clockwise and counterclockwise sides showed damage from fretting contact with the No. 8 bearing outer race. The contact damage was more extensive on the lower (clockwise) side of the lug, consistent with the bearing outer race rotating counterclockwise relative to the retaining plate. (The GP turbine and No. 8 bearing inner race normally rotate clockwise).

Fretting contact marks were observed on the sump nut face corresponding to contact with the outer diameter of the retaining plate. A wear contact mark was observed on one of the castellation surfaces on locking flange on the aft side of the sump nut. The wear mark corresponded to contact with the retaining ring with the sump nut flange approximately flush to the aft side of the GP hub.

The retaining ring for the sump nut is installed in a groove at the aft side of the castellated GP support hub. The aft side of the GP support hub was deformed radially outward at five of the castellations. As a result of the deformation, the retaining ring did not fully seat in the groove around the diameter of the GP support hub. However, the deformation did not appear to affect full engagement of the retaining ring with the sump nut lock flange.

The e-ring seal was not located, and there was no documentation confirming the presence or absence of the e-ring seal in any photographic documentation or notes taken at the time of the engine teardown.

The No. 8 bearing inner race, cage, and outer race were intact, and all rolling elements were missing. The No. 8 bearing outer race and cage were held within the fractured aft piece of the stationary seal. The notch on the aft side of the No. 8 bearing outer race and the corresponding lug on the forward face of the retaining plate were intact but marks consistent with heavy contact were observed. The shear pin restraining rotation of the forward end of the stationary seal was intact in the GP support. The components of the No. 8 bearing were substantially darkened consistent with high heat and loss of lubrication.

The stationary seal for the No. 8 bearing was tinted consistent with high heat exposure and was fractured at the forward end of the bearing cup adjacent to the shoulder for the bearing outer race. Flat oxidized areas with curving boundaries were observed consistent with fatigue. The fatigue initiated from multiple origins at the inner diameter of the cup wall around the circumference. The planes associated with each origin were slightly angled relative to the circumferential plane consistent with torsional loading combined with flexure or tension loads. The relative orientation of the angle was consistent with torsion loading associated with the aft end of the seal loaded counterclockwise relative to the forward end, which is consistent with the rotation observed associated with the retaining plate. Ratchet marks were rubbed from contact with the mating side of the fracture, consistent with counterclockwise rotation of the aft side of the fracture relative to the forward side following fracture.

The fracture surface on the No. 8 stationary seal was further examined using the scanning electron microscope (SEM). The fracture surface showed substantial damage from post-fracture rubbing contact, and fatigue striations were also observed.

The shoulder on the No. 8 stationary seal adjacent to the fracture surface showed heavy fretting contact damage on the aft face adjacent to the inside diameter of the shoulder, consistent with contact with the No. 8 bearing outer race. Fretting damage was also observed on the outside diameter of the stationary seal where it contacted the GP support hub.

According to information provided by representatives for the engine manufacturer, the No. 8 stationary seal should have an interference fit in the GP support hub within approximately ¼-inch of the shoulder against which the bearing cup portion of the seal rests, and the stationary seal has two part-number options available to achieve the proper interference fit. The “-1” part has an outside diameter that is slightly larger than the “-2” part.

The inside diameter of the GP support hub was measured, and the measured value could be within the specified maximum inside diameter after accounting for the range of accuracy of the measuring device. The outside diameter of the No. 8 stationary seal was measured on the forward piece in the interference fit region using calipers. Several measurements were taken including areas that appeared to have original machining markings, and none would provide an interference fit with the as-measured or the as-specified inside diameter of the corresponding area on the GP support. It was undersized by 0.008 inch or more, depending on which part number had been installed. A part number for the No. 8 stationary seal could not be obtained due to heat and impact damage.

Examination of the No. 8 rotating seal revealed heavy circumferential rubbing damage and high heat damage. White deposits were observed in the rotating seal ridges. The flange in the middle of the rotating seal was fractured around the circumference, and the remaining portion was deformed aft.

The No. 8 spanner nut lock flange was deformed in one location and had a semicircular section removed from a second area. The size of the semicircular cutout was consistent with removal of a deformed flange segment from a prior installation.

Pieces of the interior wall of the outer combustion case were re-assembled and two of the pieces were darkened relative to the mating piece, consistent with post-fracture heat exposure. Examination of the fractured pieces showed substantial rub damage with lips on either side of the surface consistent with vibratory contact damage. After an initial optical examination using a stereo microscope, the fracture surfaces on the two pieces without heat damage were further examined using the SEM. Although most of the surfaces were obliterated by rub damage, striations consistent with fatigue fracture were observed.

The compressor assembly was intact and rotated freely. Some impact marks were observed on the aft end face of the impeller stub shaft.

According to overhaul records, the turbine section was last overhauled on July 2, 2020, when the GP turbine wheels were replaced due to service time limits. A review of the engine manufacturer’s overhaul maintenance manual (OHM), section 72-50-00 L, revealed the following inspections were required when the turbine module was removed to address life expired 1st or 2nd stage turbine wheels:

(1) Visually inspect the No. 1 and No. 2 turbine nozzle assembly for any damage or discrepancies.

(2) Inspect the gas producer tie bolt, 2nd-stage splined adapter and compressor to turbine coupling shaft. Replace if necessary.

(3) Inspect all oil nozzles and passages for carbon formation and/or obstructions. Clean as necessary. During assembly, make sure oil will pass through all designated nozzles and that they are targeted appropriately.

(4) Inspect the power turbine rotating labyrinth 9--12 and 13--18 seals.

(5) Inspect 3rd-- and 4th--stage turbine wheels for obvious cracks or damage.

(6) Visually inspect the power turbine inner shaft I.D. for excessive carbon buildup. Clean any excessive carbon from the power turbine inner and outer shaft.

The required inspections necessitate disassembly, inspection, and reassembly for the turbine module components, including the GP support bore diameter and the condition of the No. 8 stationary seal; however, there was no indication in the overhaul records that the No. 8 stationary seal had been inspected, removed, or replaced.

A few months after this overhaul, in September 2020, the turbine module was removed from the engine by the operator and sent to a repair facility due to a N2 lockup. The repair facility ended up removing the GP support and sending the unit to another facility where the 4th stage wheel was replaced. According to the engine manufacturer, the repair facility that received the GP support should have been following the same OHM inspection criteria as stated above, which included inspection of the No. 8 stationary seal at the time the 4th stage wheel was replaced. There was no entry in the maintenance record that the No. 8 stationary was repaired/replaced at the time the GP support/4th stage wheel was replaced.

On April 25, 2020, at 1240 eastern daylight time, a Hughes 369D helicopter, N9159F, was substantially damaged when it was involved in an accident near Pylesville, Maryland. The pilot was not injured. The helicopter was operated under the provisions of Title 14 Code of Federal Regulations (CFR) Part 133 as a rotorcraft external load operation.

The pilot reported that while he was performing human external cargo (HEC) long line operations, he heard on the radio that ground personnel were having difficulty moving a conductor power line (wire) nearby. He proceeded to the landing zone, which was about 300 to 400 ft from the area requiring assistance, and dropped off the HEC. Then, while hovering, he picked up a conductor hook via the long line (with assistance from ground personnel) and continued to the area that needed support.

He reported that after the hook was attached to the conductor wire, he began maneuvering for about 10 to 15 seconds to move the wire a short distance laterally, as a crane was supporting the weight of the wire. According to the pilot, while maneuvering, he applied "slight aft and up pressure" to move the conductor wire and there was no lateral banking. He believed the pitch attitude during the maneuvering was about 5° to 10° nose up. After the conductor wire was moved to the desired area, the pilot maneuvered to remove the hook from the wire, but before the hook was free, the helicopter entered a left yaw and the engine began "spooling down."

The pilot reported that he subsequently heard the "engine out alarm" and entered an autorotation by "slamming the collective down." The pilot reported that the loss of engine power occurred about 150 ft above ground level (agl) and that he immediately pulled the belly band release lever—one of two levers needed to release the long line (the belly band was a secondary cable support system the operator used for HEC operations to provide redundancy in the event of an inadvertent release of the cargo hook; see figure). The pilot stated that he did not have sufficient time to pull the second (mechanical release) lever on the cyclic control to release the long line.

As the helicopter entered the flare, the pilot pulled the collective up to complete the autorotative landing, but the long line, which remained attached to the helicopter and conductor wire, became taut and caused the helicopter to roll onto its left side. The main rotor blades impacted the ground.

Multiple witnesses on the ground reported that they heard the helicopter's engine lose power while the pilot was maneuvering, and they subsequently observed the helicopter begin a rapid descent. One witness stated that when the helicopter was about 3 ft from the ground, “the long line got tight and started to tip the aircraft over.”

The following figure shows the belly band around the fuselage, the main hook, and long line.

Figure. View of the helicopter at the accident site

The accident helicopter’s fuel system was composed of two interconnected fuel tanks installed beneath the passenger seats. Fuel was delivered to the engine from a fuel pick-up port on the left side of the left tank. The rotorcraft flight manual stated that the total usable fuel was 421.9 lbs.

Manufacturer Guidance

In November 2015, MD Helicopters, the type certificate holder at the time, published Operational Safety Notice OSN2015-002, “Fuel Starvation Due to Unporting of Fuel Supply Pick-Up.”

The notice warned operators that when the helicopters are used to conduct operations with a “long line” attached to pull or tow objects on the ground, a significant side load can be placed on the helicopter. These side loads can create high fuselage pitch and roll angles as well as uncoordinated flight, which in turn can increase the amount of unusable fuel and result in fuel starvation due to unporting of the fuel supply pick-up.

The notice further stated in part: MDHI Helicopters are not specifically certified for operations with the potential for sustained high fuselage pitch and roll angles in uncoordinated flight, such as powerline stringing operations. To help mitigate the possibility of fuel starvation and the potential safety risk, consider modifying fuel management procedures for such operations. Instead of allowing such operations with minimum fuel safety margins associated with normal flight attitudes during coordinated flight, consider increasing minimum fuel level requirements when operations will involve high deck angles in pitch and roll during uncoordinated flight.

The accident helicopter’s fuel system was composed of two interconnected fuel tanks installed beneath the passenger seats. Fuel was delivered to the engine from a fuel pick-up port on the left side of the left tank. The rotorcraft flight manual stated that the total usable fuel was 421.9 lbs.

Manufacturer Guidance

In November 2015, MD Helicopters, the type certificate holder at the time, published Operational Safety Notice OSN2015-002, “Fuel Starvation Due to Unporting of Fuel Supply Pick-Up.”

The notice warned operators that when the helicopters are used to conduct operations with a “long line” attached to pull or tow objects on the ground, a significant side load can be placed on the helicopter. These side loads can create high fuselage pitch and roll angles as well as uncoordinated flight, which in turn can increase the amount of unusable fuel and result in fuel starvation due to unporting of the fuel supply pick-up.

The notice further stated in part: MDHI Helicopters are not specifically certified for operations with the potential for sustained high fuselage pitch and roll angles in uncoordinated flight, such as powerline stringing operations. To help mitigate the possibility of fuel starvation and the potential safety risk, consider modifying fuel management procedures for such operations. Instead of allowing such operations with minimum fuel safety margins associated with normal flight attitudes during coordinated flight, consider increasing minimum fuel level requirements when operations will involve high deck angles in pitch and roll during uncoordinated flight.

Photographs provided by a Federal Aviation Administration (FAA) inspector who examined the helicopter at the accident site found that the helicopter had rolled over and come to rest on its left side, and the long line remained attached from the main hook on the helicopter to the power line. The tail boom and main/tail rotors sustained substantial damage. There was no evidence of fuel spillage at the accident site.

Additional examination of the helicopter supervised by the NTSB investigator-in-charge found that the cyclic, collective, and throttle each had continuity through the full range of motion. The main hook release lever opened the hook normally when activated.

There were no obstructions observed in the turbine air inlet. The oil filter and fuel filters were clear of any remarkable debris. Pressure and leak tests were performed on the engine’s pneumatic and fuel system; no leaks were observed on either system. The electrical fuel pump (start pump) would not activate when electrical power was supplied to the helicopter. A replacement electrical fuel pump was installed on the helicopter and functioned normally. With the new electrical fuel pump installed, a total of 146 lbs (21.5 gallons) of fuel was pumped from the helicopter. This volume was consistent with the fuel gauge, which displayed about 150 lbs.

The engine was subsequently removed and test run under the supervision of the NTSB investigator-in-charge. The engine produced idle through takeoff power, with no anomalies observed, and all engine parameters remained within tolerances throughout the test run.

For a portion of the test run, the positive pressure fuel supply was eliminated to simulate conditions similar to an electrical fuel pump failure. The engine continued to produce takeoff thrust consistent with the previous data when positive fuel pressure was available.

FAA Regulations and Guidance

Advisory Circular (AC) 133-1B, Rotorcraft External-Load Operations, provided the following two definitions for Class B and C Rotorcraft-load combinations (RLC):

Class B RLC. The external load is jettisonable, carried above or below the skids, and lifted free of land or water during the rotorcraft operation. An air conditioner unit being lifted onto the roof of a tall building is an example of a Class B load (§ 1.1).

Class C RLC. The external load is jettisonable and remains in contact with land or water during the rotorcraft operation. Wire stringing, dragging a long pole, and boat towing are some examples of Class C loads (§ 1.1).

AC 133-1B does not contain minimum fuel standards based on the specific type of RLC class to be flown.

Part 133 requires no additional fuel minimums beyond that required in 14 CFR 91.151, Fuel Requirements for Flight in VFR Conditions.

As part of this investigation, MD Helicopter provided the NTSB a computer model that outlined a combination of static pitch and roll angles and corresponding fuel levels at which the fuel pick-up point may become unported. According to this information, with about 21.5 gallons of fuel onboard, at 0° lateral banking, the pitch up attitude required to unport the fuel pick up was 28.5°. The computer model could not account for dynamic flight operations that may affect the movement of fuel in the tanks (for example, maneuvering, turbulence, or uncoordinated flight, which would allow for fuel to move freely within the fuel tank).

A search of the NTSB’s aviation accident database for 14 CFR Part 133 fuel starvation events involving any rotorcraft type performing Class C RLC operations found three reports relevant to this investigation.

In 1990, the NTSB investigated a helicopter accident involving a MD369D that was conducting external long line operations (LAX91LA054). The report stated that by duplicating the helicopter's pitch attitude and fuel load of 115 lbs postaccident, the fuel pick-up point became unported at fuel quantity levels at or below 115 lbs. The exact pitch attitude was not specified in the report.

In 2012, the NTSB investigated a helicopter accident involving an MD369E that was conducting external long line operations (WPR12LA328). The report stated that 117 lbs of fuel remained on board and previous investigations of similar accidents determined that the fuel tank supply pickup can become unported with a fuel load of less than 151 lbs when pitch-and-roll attitudes approach 20º.

In 2017, the NTSB investigated a helicopter accident involving an MD369E that was conducting external long line operations (ERA17LA209). The report stated that 14 gallons (93 lbs) of fuel remained on board and that the low fuel level light illuminated when pitch up attitudes similar to those during the accident were duplicated.

Additionally, in 2008, the Australian Transport Safety Bureau investigated a helicopter accident involving a MD369ER (Aviation Occurrence Investigation AO-2008-025) that was conducting power line stringing operations. The investigation found through testing that it was possible to introduce air into the fuel system through the fuel tank pick-up point when fuel quantity was less than 85L (151 lbs) and subjected to a 20° nose up and 20° right roll attitude.

The accident helicopter operator’s FAA-approved RLC flight manual (RLCFM) and corporate policy manual required that for all flight operations, the MD 500D helicopter land with no less than 100 lbs of fuel. For Class C RLC operations, the RLCFM specifically required the following:

Always start any Class C external load with a full tank of fuel. As the aircraft leans over in a steep bank to the right, this may easily uncover the fuel sump. There should be a maximum of one hour of flight time while performing any Class C external load.

The operator required that its Job Hazard Analysis form be completed before each flight. The “Fuel Check Off and Limitations” section of the form contained two options for the MD 500: 100 lbs Landing Minimum” and

Wire/ Rope Pull and Wreck Out Operations: Maximum 1 hr (45 min F/FF_ flight time with max fuel load. The form completed before the accident flight indicated that the option for 100 lbs landing fuel minimum had been selected.

The form contained no references to RLC classes.

During postaccident interviews, the pilot reported the following concerning his understanding of when the more restrictive fuel minimums (maximum 1 hour of flight time with a maximum fuel load) would be required:

…when we are doing Side Pull Operations. This is for when the hook is relocated from the bottom of the aircraft and installed on the side. This would be for pulling of rope or a small steel cable for powerline construction. There is a lot of right lateral banking when pulling the rope and steel cable during this flight profile.

This operation that we were conducting on the accident day was more in line with Class B operations. I understand when you have a load attached to a fixed object it becomes a C Load, but this situation did not fit that flight profile (hook on side and a high right lateral bank). The flight profile was more along the lines of a slight nose up attitude, no lateral or banking took place.

After this accident, the operator updated the Job Hazard Analysis form and the RLCFM to specifically associate fuel minimums with RLC classes (for example, Class B or C). In addition, specific examples of long line operation (for example, water bucket, lifting/moving wire, rope pull) are provided in the fuel minimum policies.

The operator also reported that safety briefings were held with relevant operational staff to ensure their understanding with the revised fuel minimum policies.