Software link suspected in Airbus A220 engine blowouts

[Mark in CA]

Reuters is reporting that A220 engine problems may be related to a recent engine software update.

Exclusive: Software link suspected in Airbus A220 engine blowouts - sources (https://www.reuters.com/article/us-airbus-a220-exclusive/exclusive-software-link-suspected-in-airbus-a220-engine-blowouts-sources-idUSKBN1X31ST)

[infrequentflyer789]

Reuters is reporting that A220 engine problems may be related to a recent engine software update.

Exclusive: Software link suspected in Airbus A220 engine blowouts - sources (https://www.reuters.com/article/us-airbus-a220-exclusive/exclusive-software-link-suspected-in-airbus-a220-engine-blowouts-sources-idUSKBN1X31ST)

Airbus mangle compressor and spit-out system. Acronym left as exercise for the reader... :E

[atakacs]

If true it would be quiet a story...

[pattern_is_full]

To err is human - to really screw up requires software.

[RickNRoll]

How could software cause that? Does the software patch just allow the engines to run harder so that a new failure mode has been exposed.

[tdracer]

How could software cause that? Does the software patch just allow the engines to run harder so that a new failure mode has been exposed.

Educated guess is that they changed the stator vane scheduling in an effort to either optimize engine performance or get a little red line speed margin - if they got it wrong it can set up harmonics in the blades at certain rotor speeds.
Pratt has had that problem before - on the PW2000 they came out with some compressor tweaks as part of a performance improvement program. Turns out if you did derated climb, you could get a harmonic in (IIRC) the fifth stage compressor which could cause the blades to fail - interestingly it was only a problem with derated climb - full rated climb it was just fine.

[b1lanc]

How could software cause that? Does the software patch just allow the engines to run harder so that a new failure mode has been exposed.

If they took a lesson from VW/Audi it would be increase thrust when certfication authorities aren't looking.

[J.O.]

An interim AD was issued today by Transport Canada. It limits the N1 setting in climb above FL290 to no more than 94%. It also limits use of engine anti-ice to no higher than FL350.

[radiosutch]

https://www.swissinfo.ch/eng/business/high-risk-_swiss-airbus-planes-must-limit-their-top-speed-at-altitude-/45330720?utm_campaign=own-posts&utm_content=o&utm_source=Facebook&utm_medium=socialflow&fbclid=IwAR3d6ti3eMmppr4dT9KGWF248wu7SjdfWDbBC0E9UQinnJevbZf v9lac9lI

[Chris2303]

Reuters is reporting that A220 engine problems may be related to a recent engine software update.

Exclusive: Software link suspected in Airbus A220 engine blowouts - sources (https://www.reuters.com/article/us-airbus-a220-exclusive/exclusive-software-link-suspected-in-airbus-a220-engine-blowouts-sources-idUSKBN1X31ST)

If it really is a software update that has caused the problem surely the software can just be rolled back?

[tdracer]

If it really is a software update that has caused the problem surely the software can just be rolled back?

Depends on what else is in the new software. Could be the old software has an even worse issue...

[futurama]

flight critical Level A software and 'open source' don't belong in the same sentence.
Why not?

Open source software is already used as a basis for (and/or incorporated in) safety critical systems in other industries, including nuclear, medicine, automotive, etc. Why not in aviation?

Just because the source is open doesn't mean the process behind it can't be as rigorous (or even more rigorous) than closed source development.

If the source code and associated documentation on MCAS had been mirrored to (say) GitHub, maybe someone from the industry or academia could have commented on its flaws before hundreds of people perished.

[tdracer]

First off, there is a great deal of proprietary information contained in FADEC software - how can you make it open source without revealing your trade secrets to your potential competitors? A lot of that information can't even be exported without special licenses (we needed to keep special agreements in place between Boeing and Rolls Royce just to allow us to talk with each other about the engine control s/w).
The flaw in MCAS was it's hazard classification (no worse than Major). Had it been properly identified as potentially Catastrophic, it never would have been implemented the way it was.
Or are you suggesting Boeing, Airbus, Pratt, GE, and Rolls provide China (and everyone else) with all the information needed to build state of the art aircraft?

[nebpor]

Why not?

Open source software is already used as a basis for (and/or incorporated in) safety critical systems in other industries, including nuclear, medicine, automotive, etc. Why not in aviation?

Just because the source is open doesn't mean the process behind it can't be as rigorous (or even more rigorous) than closed source development.

If the source code and associated documentation on MCAS had been mirrored to (say) GitHub, maybe someone from the industry or academia could have commented on its flaws before hundreds of people perished.

This is a fallacy - there are now numerous examples of core Internet infrastructure built on open source / free software, where everyone thought everyone else was auditing it - the reality is that no one was auditing it. The best example is possibly OpenSSL, which had a long-running and serious vulnerability. I am not knocking free/open source software, but some things cannot be crowd sourced and need clear ownership.

[DaveReidUK]

If the source code and associated documentation on MCAS had been mirrored to (say) GitHub, maybe someone from the industry or academia could have commented on its flaws before hundreds of people perished.

Why would Boeing have paid any more attention to third-party comments than it did to its own engineers ?

[Ben_S]

People tend to contribute to open source software because they personally have a use for it and want to make it better. It is unlikely Bill from Torquay will have anything to contribute to specialist software to control a P&W engine on an A220.

I'm inclined to agree that non pilots and aviation people like myself shouldn't be allowed to post when nonsense suggestions like the above are posted.

[Nialler]

Why not?

Open source software is already used as a basis for (and/or incorporated in) safety critical systems in other industries, including nuclear, medicine, automotive, etc. Why not in aviation?

Just because the source is open doesn't mean the process behind it can't be as rigorous (or even more rigorous) than closed source development.

If the source code and associated documentation on MCAS had been mirrored to (say) GitHub, maybe someone from the industry or academia could have commented on its flaws before hundreds of people perished.
Sorry, but this is right in my bailiwick. Open Source can not be used in many critical industries unless it is fully supported with a meaningful support contract in place. Take the financial industry, for example. Certainly in the EU all software used in banks etc must have fully licensed and supported software on *all* platforms. This is fully audited on a regular basis. It is absolutely a regulatory requirement for holding a licence.

I would be horrified at the thought of such critical systems as transport, medicine and others were allowed to cut corners by using Open Source software in key areas.

[NWA SLF]

Different engine type but FADEC related. Was sent software to update test machine, newly developed diesel engine that I said didn't perform right. Drove a machine up a hill until it stalled, then rolled backward and the engine began running - backward. Downloaded the record of my test and sure enough, running poorly but running backward. Emailed engine group who responded that could never happen but they would look at my data. Hear for months until they finally found time to run on a test stand and confirmed, required mechanical and software changes. Always nice to hear software people say that can't happen.

[flight_mode]

After 27 years of the Avro I'm sure Swiss are relaxed about these engine problems. There's half as many :}

[Australopithecus]

Can any A220 pilot comment on the ramifications of a 94% N1 limit? WRT the anti-ice above FL350: I may have used anti ice in cruise that high twice in 41 years, but I wonder if this engine is susceptible to ice crystal icing?

[tdracer]

Can any A220 pilot comment on the ramifications of a 94% N1 limit? WRT the anti-ice above FL350: I may have used anti ice in cruise that high twice in 41 years, but I wonder if this engine is susceptible to ice crystal icing?

Engine anti-ice is not directly effective against Ice Crystal Icing - ICI just bounces off of unheated surfaces. The surface needs to be warm (significantly above freezing) before ICI melts then re-freezes - heating the inlet won't have any positive effect.
On the CF6-80C2 we had pilots turn on Anti-Ice, but it was for indirect effects. First off, at least on the Boeing installations, turning on anti-ice increases the idle speed (the CF6-80C2 ICI issue was typically during an idle descent - ice would accumulate in the compressor then shed when they accelerated to level off and the ice shed would quench the combustor - so higher idle was goodness). Second, the extra bleed for anti-ice made the engine run 'richer' which provided better flameout protection when the ice shed.

[Dave Therhino]

They are trying to keep the fan shaft (really the LP compressor) out of a speed/flow condition that, with the recently implemented different stator vane schedule, puts an LP compressor blade stage into resonance. I suspect the reason for the anti-ice restriction at high altitude is that anti-ice puts a heavy bleed load on the engine, changes the flow/speed relationship within the compressor, and gets them into resonance as well. If they allowed anti-ice bleed at high altitude they'd have to further reduce the fan speed limit I suspect. High altitude climb is where fan speeds are typically highest, and reducing the allowed fan speed there is a heavy performance penalty.

[Australopithecus]

So would they even be able to make FL350 at typical weight using 94%? That wouldn’t work for the aircraft that I've flown, but perhaps the values are different for this engine?

[lomapaseo]

They are trying to keep the fan shaft (really the LP compressor) out of a speed/flow condition that, with the recently implemented different stator vane schedule, puts an LP compressor blade stage into resonance. I suspect the reason for the anti-ice restriction at high altitude is that anti-ice puts a heavy bleed load on the engine, changes the flow/speed relationship within the compressor, and gets them into resonance as well. If they allowed anti-ice bleed at high altitude they'd have to further reduce the fan speed limit I suspect. High altitude climb is where fan speeds are typically highest, and reducing the allowed fan speed there is a heavy performance penalty.

I would not look too keenly on a pilot -engine control action which leads to an engine failure in a single flight. There must be more to this.

[Dave Therhino]

I would not look too keenly on a pilot -engine control action which leads to an engine failure in a single flight. There must be more to this.

I don't understand your comment because I'm not sure how pilot-to-engine-control interaction came into this. Could you elaborate or restate it please?

[lomapaseo]

I don't understand your comment because I'm not sure how pilot-to-engine-control interaction came into this. Could you elaborate or restate it please?

Turning on ant-ice

[Dave Therhino]

Agreed - having the selection of anti-ice cause an engine failure in a very short period of time would create a threat of common cause failure of both engines on the same flight, and shouldn't be acceptable. I don't know how long this particular engine can withstand the resonance issue that is occurring, but I've seen some compressor blade vibration issues where it takes dozens of flights before the issue (a bent blade from ice impact in the case I'm thinking of) causes the blade to fail. This issue may be a case where the failure results from the cumulative exposure from a large number of flights, in which case probability of failure of both engines on the same flight from common mode failure is much lower. Again, while I'm aware that this is a blade resonance issue in part due to a vane schedule change, I don't have detailed knowledge of the issue and am speculating on the details.

[TheLambtonWorm]

BEA.AERO have found three of the parts from the Swiss plane.

http://twitter.com/BEA_Aero/status/1192101236595662850

[Annex14]

I am wondering what might be the difference in engines on the BTI and LX fleet of A 220. While the former - according FlightRadar24 - operate the entire flight envelope of the a/c type, the latter still crawls around at altitudes not higher than 29000 ft. Any explanation ?

[atakacs]

I am wondering what might be the difference in engines on the BTI and LX fleet of A 220. While the former - according FlightRadar24 - operate the entire flight envelope of the a/c type, the latter still crawls around at altitudes not higher than 29000 ft. Any explanation ?
Muss say it is somewhat baffling...