Failed Exam

By David Jack Kenny

POWER UP Magazine

8 Minutes

Resource Hub

Inadequate inspections cost four helicopter occupants their lives.

Conducting flights under CFR Part 91 offers operators considerably greater flexibility than under the stringent regulations Parts 135 and 121 impose on charter flights and scheduled air carriers, respectively. That’s true even for certain limited classes of revenue flights, including instruction in rental aircraft and nonstop air tours that remain within 25 nm of their point of departure.

Maintenance requirements, in particular, for Part 91 operations are limited to the 100-hour or annual inspections required for any aircraft flown for compensation. Manufacturers’ service bulletins—even those labeled “mandatory”—and procedures recommended for unusual situations are considered merely advisory, left to the operator’s discretion. The fact that they’re not strictly required, however, doesn’t mean that following the guidance of the people who designed, built, and probably know the aircraft better than anyone else isn’t a very, very good idea.

The Flight

On the afternoon of Feb. 15, 2021, three passengers boarded a Bell 206B-3 for a planned 17-minute sight­seeing flight around St. Thomas, the most populous of the US Virgin Islands. The weather was fine, with clear skies and 12-kt. easterly winds.

A witness in his front yard said he saw the helicopter fly over his house and continue out over the ocean; as the aircraft made a 180-degree turn back toward shore, he began recording video of the flight with his mobile phone. After about six seconds, the recording captured a puff of dark smoke from the engine bay that quickly dispersed in the rotor wash. The ship suddenly yawed left and then right before descending in a right turn toward a wooded hillside and out of the witness’s sight.

The Aircraft

The helicopter’s engine was the original unit installed when the aircraft was manufactured in 1981. At the time of the accident, it had logged 11,519 hours in service. Between December 2008 and January 2009, at an engine total time (ETT) of 8,488 hours, the compressor section was removed for inspection and overhaul; a new stage 1 and the stages 2 and 3 compressor wheels were installed, while the remaining stages’ original wheels remained installed.

The helicopter was brought to St. Thomas on Nov. 26, 2019, and although the US National Transportation Safety Board (NTSB) didn’t report the airframe and engine times as of that date, two 100-hour inspections were performed later, the first on Mar. 1, 2020, at 11,406 hours ETT, and the second on Jan. 25, 2021, at 11,504 hours ETT.

In a commercial service letter (CSL), the engine manufacturer recommends giving engines subjected to saltwater exposure a compressor rinse after the last flight of the day, spraying fresh water into the compressor inlet while turning the engine over with the starter motor.

The main wreckage of the St. Thomas accident. (NTSB Photo)

The helicopter’s logbook had record of 13 such compressor rinses during the 74 hours of operation between Jul. 20, 2020, and Jan. 13, 2021.

An earlier CSL, first issued in 1991 and most recently revised in 2007, calls for adding inspection of the case, blades, and vanes to the 300-hour inspection checklist for aircraft “operating in a corrosive and/or erosive environment.” This CSL also sets a calendar time limit of 6 months for noncoated compressor wheels and 12 months for coated wheels such as those on the accident helicopter.

The most recent 300-hour inspection documented in the logbook was signed off on Feb. 16, 2017, at an ETT of 10,960 hours—four years and 559 hours earlier than the accident date.

The aircraft records also included a 300-hour inspection checklist from Jan. 11, 2018, that cited 11,197 hours of engine operation, but no corresponding entry was made in either the engine logbook or the airframe logbook. If that inspection in fact took place, the subsequent 300-hour engine inspection would still have been more than two years and 22 flight hours overdue.

An undated engine inspection checklist, possibly from the aircraft’s 2020 100-hour inspection, showed no initials or other markings in the 300-hour section. And in the records from the helicopter’s last 100-hour inspection, in 2021, the 300-hour section of the engine checklist was crossed out and marked “NA.”

The Investigation

NTSB analysis of the recorded cell-phone video concluded the helicopter’s ground speed was “about 39 kt.” before the smoke appeared, then decreased to “about 30” before increasing to “about 68 kt.” in descent. The high resolution and rapid capture rate (equivalent to 60 frames per second) of the mobile phone’s camera also enabled analysts to calculate the main-rotor speed, which decayed from 390 rpm (99% of nominal speed) to just 74 rpm (19%) in the last second of the recording. The helicopter’s rate of descent was estimated to have reached around 4,800 ft. per minute by the time the aircraft left the camera’s field of view.

The engine compressor following the accident. (NTSB Photo)

The wreckage was found on a steep, heavily forested hillside. None of the four occupants survived. The post-impact fire consumed almost the entire fuselage; two sections of the tail boom were found separated from the main wreckage. Both main-rotor blades remained attached to their hub, and while the fire had consumed the main-rotor gearbox housing, the gear train remained intact.

The manufacturer conducted an extensive examination of the engine core, which survived the fire. The first- and second-stage compressor blades were largely undamaged and showed no evidence of damage from foreign object debris (FOD). Two blades of the third-stage compressor wheel had fractured near their roots and weren’t located; the remaining third-stage compressor blades were damaged primarily along their trailing edges.

The fourth-, fifth-, and sixth-stage compressor blades were all missing, fractured at their roots. Fragments of some axial compressor blades were found in the axial compressor section and the impeller inducer, and “the impeller inducer exhibited evidence of hard body debris ingestion.”

The roots of the missing stage 3 compressor blades “exhibited signatures consistent with fatigue.” The fractures began near the pressure side of the blades’ trailing edges, but impact damage made it impossible to determine how the fatigue originated. Not surprisingly, the NTSB found the accident’s probable cause to be “a total loss of engine power due to fatigue failure of two of the stage 3 compressor blades.” But what caused the fatigue, and how did it escape detection?

Thermal damage to the compressor case halves prevented the investigative team from determining whether the stage 3 and 6 blades might have rubbed against the case and its plastic coating, which could have initiated the fatigue fractures, the NTSB noted in its report. (NTSB Photo)

The teardown inspection found “generalized corrosion … on the inner and outer diameters of the compressor wheels for stages 2–3 and stages 4 and 5, but no pitting corrosion.” Generalized corrosion was also present on both the inner and outer diameters of the stage 6 compressor wheel, and the fracture surfaces of all but one of the stage 6 blades showed “signatures of fatigue with multiple origins near the suction-side crown root.”

Sixteen of the stage 3 stator vanes in the compressor case were missing, along with all the stator vanes from stages 4–6. The remaining stage 3 vanes were flattened, and all surviving vanes, including the undamaged vanes in the first two stages, had generalized corrosion.

The Takeaway

The NTSB’s finding of probable cause states that “contributing to the failure of the compressor blades was the failure of maintenance personnel to inspect the compressor at the recommended interval for operation in corrosive environments.”

Not reported is whether this lapse represented a conscious decision or a lack of information. Providers of services in tourist areas often operate on thin margins; the combination of the added maintenance expense and aircraft downtime might have seemed prohibitive, at least during prime visitor season.

One can imagine the operator deciding to at least postpone the inspection until bookings slowed enough to accommodate the inevitable pause between flights. Without knowing the background of the technicians who worked on the aircraft or the completeness of the paperwork that arrived with it, however, one can’t exclude the possibility that they simply weren’t aware of the CSL.

Maintaining airworthiness requires researching all relevant airworthiness directives (ADs) and confirming or attaining compliance, but CSLs, service bulletins, and the like aren’t subject to the same mandate as ADs. While obtaining all the manufacturer’s advisories might seem like cheap insurance, in a busy, understaffed shop it could also be seen as a distraction from more urgent tasks.

Whatever the reason, not performing the 300-hour engine inspection proved to be a very costly false economy. The fact that one wasn’t technically required didn’t make skipping it a good idea.