Catastrophic engine failure on Flight UA328

[DaveE]

Wow, that sounds like a long time between inspections! What is the standard interval in flight hours between engine inspections? Does routine engine inspection include the fan blades? Or is the inspection interval for them longer? Do they use x-rays to inspect the interior of the blades?

It's certainly not a routine inspection, It's very difficult (expensive) to do. They are inspected with thermal acoustic imaging. Which basically involves launching an acoustic wave into the blade and inspecting for hot spots which are created at fractures. It's only done at engine manufacturers with only a few technicians in the whole country trained to do it. This special inspection is required because they need to find internal cracks. For smaller/simpler blades they can use fluorescent dye inspection, which is much more common and easier to do.

It's not about this style of these turbofan blades, but I thought this video about blade inspection was interesting if you want to geek out a bit.

 

[Astronuc]

Wow, that sounds like a long time between inspections! What is the standard interval in flight hours between engine inspections? Does routine engine inspection include the fan blades? Or is the inspection interval for them longer? Do they use x-rays to inspect the interior of the blades?

We'll have to wait for the NTSB to finalize a report. I imagine that they will look at service life (hours), but more importantly would be cycles of operation.

NTSB docket - United Airlines Flight 328 Boeing 777 Engine Incident
https://www.ntsb.gov/investigations/Pages/DCA21FA085.aspx

An NTSB structures engineer and two investigators from the NTSB's Denver office collected fallen debris with local law enforcement and safety agencies over the next several days. Most of the structure from the inlet cowl and fan cowl doors was recovered and identified. Recovered portions of the inlet cowl, fan cowl door structure, and inlet cowl attach ring were laid out in a hangar, as shown in figure 1. The inlet cowl, fan cowl doors, and thrust reversers will be examined further to map damage and cowl failure patterns after the fan blade failure, and to examine the subsequent progression of fire in the thrust reversers.

Initial examination of the right engine fire damage, as shown in figures 2 and 3, found it was primarily contained to the engine's accessory components, thrust reverser skin, and composite honeycomb structure of the inboard and outboard thrust reversers. Both halves of the aft cowl appeared to be intact and undamaged, and all four pressure relief doors were found in the open position. The spar valve, which stops fuel flow to the engine when the fire switch is pulled in the cockpit, was found closed; there was no evidence of a fuel-fed fire. Examination of the engine accessories showed multiple broken fuel, oil, and hydraulic lines and the gearbox was fractured.

Initial examination of the right engine fan revealed that the spinner and spinner cap were in place and appeared to be undamaged (see figure 5). The fan hub was intact but could not be rotated by hand. All fan blade roots were in place in the fan hub, and two blades were fractured. One fan blade was fractured transversely across the airfoil about 5 inches above the base of the blade at the leading edge and about 7.5 inches above the base of the blade at the trailing edge. The blade's fracture surface was consistent with fatigue. A second fan blade was fractured transversely across the airfoil about 26 inches above the base of the blade at the leading edge and about 24 inches above the base of the blade at the trailing edge (see figure 6). The second blades fracture surfaces had shear lips consistent with an overload failure. The remaining fan blades were full length but all had varying degrees of impact damage to the airfoils.

As a result of this incident, on February 22, 2021, Pratt & Whitney issued Special Instruction 29F-21 providing revised thermal acoustic image (TAI) inspection threshold intervals to 1000 cycles for the first stage low pressure compressor (LPC) blades on the affected engines. On February 23, 2021, the FAA issued Emergency Airworthiness Directive 2021-05-51, which instructs owners and operators of Pratt & Whitney PW4077 and similar type engines to, before further flight, perform a TAI inspection of the first stage LPC blades for cracks and to remove the blade from service if it does not pass the inspection and replace the blade before further flight.

An initial review of maintenance and inspection data for the blade that exhibited fractures consistent with fatigue revealed that it had experienced 2,979 cycles since its last inspection. The subject blade underwent TAI inspections in 2014 and 2016. The TAI inspection data collected in 2016 was examined again in 2018 because of a February 13, 2018, incident involving a Boeing 777 with Pratt & Whitney PW4077 engines (DCA18IA092). The maintenance records group and the powerplants group are reviewing these inspection records to examine the presence and disposition of any anomalies in this TAI data near the fracture initiation point.

The last paragraph refers to 2,979 cycles vs 2,979 flights as reported in the AP article. I don't know if NTSB refers to a flight as a cycle, but when I think if cycles I think of periodic loading and unloading. It could be based on revolutions per unit time integrated over time. The repetitive loading cycle would be complicated. One would have to consider flutter in addition to revolutions.

NTSB News Release
https://www.ntsb.gov/news/press-releases/Pages/mr20210305.aspx

The previous 2018 event (United Airlines N773UA, Flight UA1175, 13 February 2018)
http://aerossurance.com/safety-management/ndi-failures-b777-pw4077-fbo/

 

[Vanadium 50]

The NTSB report was far more illuminating than the newswire reports.

A "cycle" is a takeoff plus landing. There are some edge cases where a flight is 2 cycles. So 2979 cycles in 5 years is 15 hours between cycles. This is not a crazy number for a long-haul plane.

As @DaveE points out, inspections happen much more regularly. This particular kind of inspection is difficult and happens at a lower frequency. As far as I can tell, this (pre-accident) happens whenever the engine is removed for maintenance (20,000 flight hours) or 6500 cycles, whichever comes first. N772UA was probably getting close to the 20,000 hour mark.

 

[Astronuc]

Looking at 6500 cycles, as V50 indicated it could be near 20,000 hrs. If the aircraft (or engine), if it was captured in service for 6.5 to 7 hour flights, then it was approaching 20,000 hrs based on the following average flight durations:
6.5 hrs * 2979 = 19365 hrs
7 hrs * 2979 = 20853 hrs

The duration of the flight is typically 7 hours 10 minutes, but the apparent average time is about 6 hrs 45 min.
Some sample numbers https://flightaware.com/live/flight/UAL328

The return flight (HNL to DEN) has a shorter flight time, ~6.5 hrs.
https://flightaware.com/live/flight/UAL383

UA328, Damage on the ground - http://www.kathrynsreport.com/2021/02/incident-occurred-february-20-2021-in.htmlOn December 4, 2020, a 777-289 (JA8978) operated as JL904 experienced a similar fan blade out failure and partial loss of the fan cowl six minutes after takeoff at an altitude of 16,000 feet (5,000 m).
https://en.wikipedia.org/wiki/United_Airlines_Flight_328#Japan_Air_Lines_Flight_904
https://www.aerotime.aero/26620-japan-airlines-boeing-777-turns-back-after-engine-failure
https://www.aviation24.be/airlines/...suffers-in-flight-uncontained-engine-failure/
https://www3.nhk.or.jp/news/html/20201204/k10012745841000.html (in Japanese)

Japan Transport Safety Board will publish their 2020 annual report in July 2021, and that should have a summary of their investigation of JL904. The investigation report would come later.Metal-Fatigue Signs Link United 777 Fan Blade Failure With 2018 Incident (of course)
https://aviationweek.com/mro/safety...nk-united-777-fan-blade-failure-2018-incident

This matter gets more interesting considering - NTSB to address rare though deadly forward-blade engine failures (in 2019)
https://www.flightglobal.com/in-dep...-forward-blade-engine-failures/135460.article

The National Transportation Safety Board on 19 November will issue conclusions related to a dangerous though infrequent type of turbofan failure that has proved difficult to understand and tough to prevent.

The board is meeting to determine the probable cause of a 2018 in-flight failure of a CFM International CFM56 turbofan on a Southwest Airlines Boeing 737-700.

During that event one fan blade failed, and a portion of that blade flung outward and forward, toward the front of the engine, missing a containment ring designed to prevent failed engine components from exiting the engine.

The blade portion struck the engine’s cowl and engine inlet, sending shrapnel into the fuselage, which broke one window and killed one passenger.

That was a CFM engine rather than a PW engine, yet the failure event seems to be common. A fan blade is ejected forward damaging the engine cowling. Depending on when a fan blade fails, it could be defected inward to the fuselage, or otherwise upward, outward or downward without damaging the fuselage.

https://en.wikipedia.org/wiki/CFM_International_CFM56#Fan_blade_failure

 

[Astronuc]

The fan blade material has been reported as Ti-6Al-4V (UNS R56400), which is a very common Ti alloy.
https://en.wikipedia.org/wiki/Ti-6Al-4V

One interesting comment in the introduction of the article, "However, the poor shear strength and wear resistance of titanium alloys have nevertheless limited their biomedical use." I don't know the basis of the comment. Fatigue life is a concern for the designer and user.

Another interesting comment related to the BEA study cited below - "The alloy is vulnerable to cold dwell fatigue."

https://www.colorado.edu/faculty/ka...27_am_-_asen_5063_2015_final_report_okura.pdf (See section 2)
https://www.bea.aero/uploads/tx_elydbrapports/BEA2017-0568.en.pdfLarge Engine Uncontained Debris Analysis—High-Bypass Ratio Engine Update, DOT/FAA/TC-19/10, April 2019
http://www.tc.faa.gov/its/worldpac/techrpt/tc19-10.pdf

Regarding fatigue life - https://apps.dtic.mil/dtic/tr/fulltext/u2/a444511.pdf

 

[Tom.G]

On cold dwell facet fatigue in titanium alloy aero-engine components
https://www.sciencedirect.com/science/article/pii/S0142112316304054

That's a free download, originally published in:
International Journal of Fatigue
Volume 97, April 2017, Pages 177-189

 

[Vanadium 50]

International Journal of Fatigue

What a great title!

 

[Stormer]

Gearing solves one problem, but does add more mechanical stuff that can fail.

Yes. Like the gearbox failures on helicopters that has caused the whole rotor to come off:



In a turbofan i guess that will be the whole fan axle shearing off and the fan shooting forward and out of the cowling and it can then hit the fuselage or wing and make a lot of damage. Seems to me to be a much bigger safety risk than a fan blade coming off and generally being contained in the engine and spit out the rear.

 

[Astronuc]

NDI Process Failures Preceded Boeing 777 PW4077 Fan Blade Off (FBO) Event
UA328 was the third such event.

http://aerossurance.com/safety-management/ndi-failures-b777-pw4077-fbo/

Edit: I had some discussion at work about a more robust NDE method. One of the experts seemed to think a visual examination of the fan blades was possible. Frankly, I don't see that is the case. Does anyone know if it is possible to visually inspect the inner surface of the follow fan blade? Otherwise, one has to rely on a thermoacoustic method. Can that be applied in situ, or does it require removing the fan blades?

https://en.wikipedia.org/wiki/Thermal_acoustic_imaging (P&W developed a method).

 

[cjl]

If the flight was non-stop from Denver to Honolulu, the Boeing 777 likely carried maximum fuel. Aside from the exposed flames, as questioned above, the time and pumping required to jettison fuel coupled with Denver altitude and return distance to airport likely contraindicated dumping fuel.

United Airline quite lucky in that ~3/4 of the route DEN to HNL extends over open ocean. Incredibly lucky that engine debris fell on land, so recoverable, without any known injuries or related damage.

It's unlikely that they had maximum fuel - DEN-HNL is only about 3000 nautical miles, way short of the 777-200's 5120 nautical mile range. They may have been up near maximum takeoff weight when this occurred, but if they were, that would be due to cargo, not due to fuel (aircraft are, in general, unable to take both maximum fuel and maximum cargo at the same time, so anytime they don't need full fuel, they prefer to not take it to allow more room for cargo and/or to allow the aircraft to be lighter and more efficient).

 

[Hayley Mann]

I'm a little afraid to fly. It started when I read the book, I don’t remember what it’s called and who wrote it, but I remember the plot very well. It described the inspection of a Boeing 737 before the flight. I learned from this book that there are different levels of validation. The bottom line is that the engineer, while checking the plane, did not notice a microcrack, because of which the plane was torn to pieces in the sky. I honestly didn’t google the truth it happens or it’s just a fictional fairy tale for a book. But now, every time on the plane, I worry about inattention, myself and other passengers.
I would be glad to hear your opinion and I hope that I'm just winding myself up

 

[Astronuc]

The bottom line is that the engineer, while checking the plane, did not notice a microcrack, because of which the plane was torn to pieces in the sky.

That is true. The pilot crew and maintenance crews did not look carefully at the fuselage. A passenger on the flight noticed a crack(s), but did not mention anything, because she did not want to appear alarmist. The plane was a 737, and the flight was Aloha Airlines Flight 243 on April 28, 1988. The pilots did an extraordinary job in landing the plane. "The safe landing of the aircraft established the incident as a significant event in the history of aviation, with far-reaching effects on aviation safety policies and procedures."

https://en.wikipedia.org/wiki/Aloha_Airlines_Flight_243
https://www.ntsb.gov/investigations/AccidentReports/Pages/AAR8903.aspx
https://www.faa.gov/about/initiativ...200.n73711.near_maui.hawaii.april_28.1988.pdf

It is important to realize that the accident was 33 years ago, and improvements have been made in aircraft design, maintenance, operation and safety. The aircraft was used in frequent short duration (basically commuter) flights. I believe it had one of the highest number (89,680) of flight cycles (takeoff and landings) of any aircraft, which lead to fatigue cracking. It was noticeable near the front passenger door, but obviously, those responsible didn't notice. As a result, maintenance practices were changed.

I often check along the fuselage and around the passenger door when boarding a plane. I also look at the engines and wings.

 

[Vanadium 50]

It is important to realize that the accident was 33 years ago, and improvements have been made in aircraft design, maintenance, operation and safety

This.

As a result, maintenance practices were changed.

Most significantly, these inspections were no longer conducted at night. Sounds like a good idea to me.

 

[Stormer]

A passenger on the flight noticed a crack(s), but did not mention anything, because she did not want to appear alarmist.

Even if she had mentioned it to the flight attendants I seriously doubt anything would have been done about it. After watching some air crash investigations it is pretty clear that the flight attendants have absolutely no knowledge of the plane, and their only job is to just try to calm down the passenger no matter what he or she try to point out. Like passengers pointing out engine fires and the air attendant just says it is totally normal and nothing to worry about. Or passengers pointing out a BIG tear of the wing and the flight attendant just say "that is just the flaps, nothing to worry about" (flaps out mid flight?). And the passenger have to call the flight attendant over MANY times and be extremely insisting until finally they give up and tell the pilot to check it out.

 

[Tom.G]

Many years ago my sister was on a flight that had to make an unscheduled stop for maintenance, something to do with cabin pressure that caused a couple of windows to fog up. This was to be a short delay while a sensor or other small part was replaced.

After de-planing, she was in a conversation with a stewardess and mentioned that many/all of the windows were hot to the touch. Fortunately the stewardess was knowledgeable enough to know this was important and passed it on to 'higher authority.'

Soon, all the passengers had a hotel room for the night while further repairs were done.

The following day, during reboarding, the stewardess thanked her for reporting the hot windows. Much of the cabin air system had failed and was replaced. If the flight had continued the previous day, there would have been many deaths from hypoxia.

Take away: If you see something, say something!

Cheers,
Tom

 

[Astronuc]

Take away: If you see something, say something!

Yes! Especially these days. If something doesn't look right, then one should speak up.

If I saw something, as an engineer, I'd speak up, and I'd speak with the pilot rather than crew, or if I had to encounter the crew first, I'd ask to speak with the pilot.

In the case of the Aloha 737, it was apparently quite evident that cracks were visible at the line of rivets. Back in 1988, airline crews might have been less inclined to heed concerns of the public. My understanding that this has changed over the decades.

 

[Klystron]

Approaching Atlanta late 1970's in a Lockheed L-1011 wide-body transport with another USAF member when things became rather dicey on the sparsely populated flight. Anticipating bad weather we had asked the senior flight attendant, obviously a seasoned professional, which seats to take for the most comfort. Without hesitation she seated us in the rear of the aircraft across from the crew's seats.

The L-1011 had an unusual 'tristar' configuration with a large rear engine. Our rearmost seats lacked headroom not compensated by the engine noise. As night fell a major storm developed and the reports from the first officer over the intercom became ominous. Even old hands were seen reviewing emergency directions provided at each seat. The flight attendants strapped in on orders from the FO; the Captain fully engaged in piloting that beast in a terrific tailwind.

That landing, the last before the hurricane closed ATL, sticks in my memory as the most outrageous I have ever experienced even in storms and missile attacks in Southeast Asia during the war. The lightning flashed purple through the ports, illuminating our strained faces. I thought my head would crash through the overhead as the transport bucked and side slipped. I swear, the landing gear bottomed out on touchdown as we bounced, not rolled, to a stop.

Deplaning on shaky legs, the entire crew bid us passengers farewell and good luck amid fervid compliments to the pilots and crew. The senior flight attendant, cool and composed, remarked in a mid-Western accent, "Now boys, was that enough excitement for ya?"

 

[DaveE]

Approaching Atlanta late 1970's in a Lockheed L-1011 wide-body transport with another USAF member when things became rather dicey on the sparsely populated flight. Anticipating bad weather we had asked the senior flight attendant, obviously a seasoned professional, which seats to take for the most comfort. Without hesitation she seated us in the rear of the aircraft across from the crew's seats.

The L-1011 had an unusual 'tristar' configuration with a large rear engine. Our rearmost seats lacked headroom not compensated by the engine noise. As night fell a major storm developed and the reports from the first officer over the intercom became ominous. Even old hands were seen reviewing emergency directions provided at each seat. The flight attendants strapped in on orders from the FO; the Captain fully engaged in piloting that beast in a terrific tailwind.

That landing, the last before the hurricane closed ATL, sticks in my memory as the most outrageous I have ever experienced even in storms and missile attacks in Southeast Asia during the war. The lightning flashed purple through the ports, illuminating our strained faces. I thought my head would crash through the overhead as the transport bucked and side slipped. I swear, the landing gear bottomed out on touchdown as we bounced, not rolled, to a stop.

Deplaning on shaky legs, the entire crew bid us passengers farewell and good luck amid fervid compliments to the pilots and crew. The senior flight attendant, cool and composed, remarked in a mid-Western accent, "Now boys, was that enough excitement for ya?"

Odd that they wouldn't have seated you closer to the center of mass.

 

[Klystron]

I argued the exact same thing. Sit near the pivot point. But:

We were in uniform IMS. The unpopular war had just ended. Also, while I was married, one of the attendants was flirting with the other NCO on the aisle whom she engaged in talk. Upon reflection, that flight may have been crowded due to other passengers missing flights. USAF MAC long distance transports often had few passengers but this was Delta Airlines IIRC.

 

[Tom.G]

Without hesitation she seated us in the rear of the aircraft...

Odd that they wouldn't have seated you closer to the center of mass.

Not at all odd.

During an uncontrolled contact with the ground, there supposedly is a survival advantage to have all available crushable material in front of you.

And some/many of those cabin-crew seats are rear facing, backed against the bulkhead to the passenger area.

I would certainly opt for that!

Cheers,
Tom

 

[Twigg]

Out of curiosity, how do you tell the difference post mortem between fatigue and a blunt force pigeon at mach 0.8? I feel like that's the kind of question that decides who has to pay a lot of money.

Incredibly lucky that engine debris fell on land, so recoverable, without any known injuries or related damage.

I'm not far from places where debris was found. I checked the lawn just in case, but didn't find any bits of aerospace-grade unobtanium to play with (Not that I wouldn't have reported it, just horsing around)

 

[berkeman]

Out of curiosity, how do you tell the difference post mortem between fatigue and a blunt force pigeon at mach 0.8?

Pigeon guts/blood?

 

[Twigg]

Ok, I feel a little silly not putting that one together lol. Thanks!

 

[Astronuc]

difference post mortem between fatigue and a blunt force pigeon

Besides bird remains, a fatigue crack has a unique pattern of striations, usually propagating from an inclusion or some internal/surface defect finally reaching a critical crack size and rupture, whereas a catastrophic fraction would not have the striations but a ductile fracture. There would be some plastic deformation of the blade as well. One blade failed near the root, while the blade following in the rotation was struck by the first failed blade and broke about half-length. The second blade would have fractured from impact with the first blade.