Airbus recalls thousands of A320

https://aviationweek.com/air-transp...immediate-airbus-a320-flight-control-software

"The issue affects A320s and the newer A320neo variants, including A319s and A321s in each case. Most ELACs [Elevator and Aileron Computers] can be fixed by reverting to a previous version of recently-updated software, Airbus said.

But about 1,000 of the oldest affected aircraft need a hardware change to accept the new software. These airframes will need the old hardware re-installed--a process that will take longer."

"Analysis of a recent event involving an A320 family aircraft has revealed that intense solar radiation may corrupt data critical to the functioning of flight controls,"

This is fascinating, reverting software makes me suspect it was something that had software as well as hardware protection against bit-flipping, but it accidently got disabled/deleted in the latest update. The other possibility is the software change has created a new vulnerability that didn't previously exist. What's going on with the older ELACs isn't clear, I'll be really interested to see the details here.

ISTR I ran into this on 777 once, a scenario where a specific bit-flip during finals would almost guarantee a hull-loss. It was simple to just keep two copies of the relevant bit and compare them whenever needed. (What I can't remember is why they were asking me if we should fix it - not my area guv!) It was a really esoteric, low probability error, but for amusement I multiplied the probability I was given by a 1000 aircraft fleet and a 30 year fleet life and the numbers suggested it would occur at least once. (We're now at 30 years in service and a 1200 aircraft fleet).
 
DWG said:
"The issue affects A320s and the newer A320neo variants, including A319s and A321s in each case. Most ELACs [Elevator and Aileron Computers] can be fixed by reverting to a previous version of recently-updated software, Airbus said.
Click to expand...
The XLR is not affected by this issue... I wonder if it's because it features a FBW E-rudder which made Airbus change the flight control computers ?

I know this isn't a rudder issue, but maybe the changes to the computers had an incidental effect, indirectly preventing the problem from manifesting ?

https://www.airbus.com/en/newsroom/...is-up-flight-testing-the-a321xlr-gathers-pace
Notable flight-physics-related changes on the -XLR (compared with today’s A321neo) include a simpler single-slotted inboard flap system (replacing the double-slotted inboard flap of the original A321’s wing); an electronically signalled “e-Rudder” (with related changes to the flight control computers); and uprated landing gear, wheels and brakes.
Click to expand...
 
Titus K said:
The XLR is not affected by this issue... I wonder if it's because it features a FBW E-rudder which made Airbus change the flight control computers ?

I know this isn't a rudder issue, but maybe the changes to the computers had an incidental effect, indirectly preventing the problem from manifesting ?

https://www.airbus.com/en/newsroom/...is-up-flight-testing-the-a321xlr-gathers-pace
Click to expand...
Apparently ELAC 2 normally controls the horizontal tail, so it's possible, but equally likely is the XLR either has a different software version to suit something specific to its handling (perhaps resulting from the flap changes), or they're still on the old version they're reverting to.
 
Note that patching the software is probably not what takes away the most time. Think pulling the plane away from the flight line, having enough dedicated operators available and testing the upgrade.

At this day, many companies have carried the upgrade in record breaking time. It´s probably show a bit of disrespect to call it something easy as snapping fingers (see Joker twit)
 
TomcatViP said:
Note that patching the software is probably not what takes away the most time. Think pulling the plane away from the flight line, having enough dedicated operators available and testing the upgrade.
Click to expand...
The critical path is probably availability of the data loader, with the worst case being aircraft which are away from home base. There are absolutely certainly more qualified avionics techs than loaders, because you don't want an aircraft-on-ground situation when the one guy who knows how to use the PDL is on holiday or out sick today. There's no need to pull the aircraft from the flightline, you can do this kind of thing during daily maintenance. And there shouldn't be a need for extensive testing because you'll have a specified checksum for the build and the chance of the loading procedure not requiring the aircraft to recalculate the checksum and read it back is zero.

(I prepared the downloadable FCS builds for both 777 and Eurofighter, and while I never got to do the on-aircraft loading, a friend did that for both aircraft).
 
Software updates are not part of the turnaround process of an airliner. It´s then an out of schedule dedicated technician that has to be made available. Similarly, the update might require some kind of validation. Etc...
 
TomcatViP said:
Software updates are not part of the turnaround process of an airliner. It´s then an out of schedule dedicated technician that has to be made available. Similarly, the update might require some kind of validation. Etc...
Click to expand...
You don't need a dedicated technician, you just need an avionics tech who's been trained to use the PDL, which all the maintenance bases will have, because these things do get updated. Even if they had to call someone in, it's just a case of them working through every aircraft on the flightline one after the other - plug the PDL in, run the installer, verify the checksum, done. No different to any other software update.
 
TomcatViP said:
... is a dedicated technician per se. Enough of that @DWG , thank you.
Click to expand...
No, a dedicated technician is one who is dedicated to ONLY running the PDL.

As opposed to Joe Line Dog who can fix almost everything and was trained how to use the PDL and get the checksum out of it.

I'd honestly expect that the various operators are willing to loan a PDL if someone has an aircraft on ground someplace.
 
Scott Kenny said:
No, a dedicated technician is one who is dedicated to ONLY running the PDL.

As opposed to Joe Line Dog who can fix almost everything and was trained how to use the PDL and get the checksum out of it.
Click to expand...
Absolutely.

(I should point out the friend who was doing FCS installs on Eurofighters and 777s isn't even an avionics tech, he's a software engineer like me. For that matter I've certainly done it on Eurofighter boxes, though only in the lab, so not using representative loader hardware. Don't think I ever did it on 777 boxes, but I've done software installs on various development boxes going all the way back to when we were flashing EEPROMs - though the legendary case in which we sent an update by faxing the relevant section of paper tape and the guys at the other end splicing it in with scissors was before my time).

The process isn't that complex.

Load the old build into the PDL (apparently L103+ vs L104)

Walk to the aircraft, plug the PDL in in, press the start button, wait 20 minutes, check no errors/checksums match, done. Unplug and repeat.

Apparently we were down to less than 100 aircraft needing updates as of Monday. Which are presumably the few aircraft needing hardware updates.

There's an interesting possible case of cosmic ray induced data corruption on an A330 here (though two out of three rare events happening in the same box would make me sideline that one box out of pure caution) https://www.flightglobal.com/qantas-a330-upset-pinned-on-rare-data-spike-effect/103468.article

* L104 has some interesting safety upgrades to reversionary modes, so hopefully we can get that functionality back quickly
 
DWG said:
There's an interesting possible case of cosmic ray induced data corruption on an A330 here (though two out of three rare events happening in the same box would make me sideline that one box out of pure caution) https://www.flightglobal.com/qantas-a330-upset-pinned-on-rare-data-spike-effect/103468.article
Click to expand...
"But the ATSB found that the failure mode had been observed only three times in 128 million hours of operation - twice with the same specific unit, the one fitted on the Qantas A330."​
Oh yeah, there's something NOT RIGHT with that unit. Back to the manufacturer with ye'!

DWG said:
* L104 has some interesting safety upgrades to reversionary modes, so hopefully we can get that functionality back quickly
Click to expand...
I suspect that they're going to have to do a whole new build.
 
Some more details on the quality escape affecting some panels that seems to result, not from an improper gauging at raw material supplier level, but inadequate working methodology of a tier-1:

The affected parts have the wrong thickness following a process of stretching and milling carried out by Seville-based Sofitec Aero, the presentation showed.
Click to expand...

https://www.reuters.com/business/ae...ons-fuselage-flaw-hits-deliveries-2025-12-03/
The good news however:

Unlike the emergency recall of thousands of Airbus A320s for a software change over the weekend, the fuselage problem is not being treated as an immediate safety matter, sources said.
Click to expand...
 
TomcatViP said:
Some more details on the quality escape affecting some panels that seems to result, not from an improper gauging at raw material supplier level, but inadequate working methodology of a tier-1:



https://www.reuters.com/business/ae...ons-fuselage-flaw-hits-deliveries-2025-12-03/
The good news however:
Click to expand...
I've seen some comments on machinist's groups on FB that Airbus is demanding crazy surface flatness out of their suppliers. To the tune of "show us how you set up your master piece to meet this requirement"
 
Scott Kenny said:
I've seen some comments on machinist's groups on FB that Airbus is demanding crazy surface flatness out of their suppliers. To the tune of "show us how you set up your master piece to meet this requirement"
Click to expand...
As an infrequent flier, I highly/deeply appreciate that/when both candidates A(irbus) and B(oeing) are held to the very same exacting standards.
 
martinbayer said:
As an infrequent flier, I highly/deeply appreciate that/when both candidates A(irbus) and B(oeing) are held to the very same exacting standards.
Click to expand...
You're an aerospace engineer, right?

Does it really matter how flat the web of an H beam is?
 
Scott Kenny said:
You're an aerospace engineer, right?

Does it really matter how flat the web of an H beam is?
Click to expand...
I am indeed a German Diploma Aerospace Engineer (thankfully, I graduated from University of Stuttgart well before Germany switched to that wretched *international* master/bachelor system, so with my Dipl.-Ing. degree {and [no] thanks to the Bologna Process - they should just have stuck to making Italian food...} I'm part of a dying breed, which no doubt some forum members may consider a most welcome trend) with an initial focus on *conceptual* reusable space transportation systems (and you can find some of my associated work on this august forum) by training, but I really never got down to the level of exploring how flat the web of an H beam is, so... is this like angels dancing on a pinhead?
 
Last edited:
martinbayer said:
I am indeed a German Diploma Aerospace Engineer (thankfully, I graduated from Stuttgart University well before Germany switched to that wretched *international* master/bachelor system, so with my Dipl.-Ing. degree {and [no] thanks to the Bologna Process - they should just have stuck to making Italian food...} I'm part of a dying breed, which no doubt some forum members may consider a most welcome trend) with an initial focus on *conceptual* reusable space transportation systems (and you can find some of my associated work on this august forum) by training, but I really never got down to the level of exploring how flat the web of an H beam is, so... what - is this like angels dancing on a pin?
Click to expand...
No, the part flatness I was talking about was on H beams. And it was at a level where the machinist writing about it could not personally figure out how to get it and was turning to social media for ideas.

Airbus is either trying to force out some bidders to get their preferred contractor, or the engineers are making demands without knowing what the machinists and machines are capable of achieving.
 
Scott Kenny said:
Does it really matter how flat the web of an H beam is?
Click to expand...
Yes. If said H-beam is non-conformant to the specification that you signed up to meet in the contract.

There are waivers and deviations ya'know. "Parts made acceptable" etc.
 
aim9xray said:
Yes. If said H-beam is non-conformant to the specification that you signed up to meet in the contract.

There are waivers and deviations ya'know. "Parts made acceptable" etc.
Click to expand...
Is the specification realistic?

Because we're talking about parts being in spec while still in the work fixtures and then out of spec once you release the part from the clamps.
 
Scott Kenny said:
No, the part flatness I was talking about was on H beams. And it was at a level where the machinist writing about it could not personally figure out how to get it and was turning to social media for ideas.

Airbus is either trying to force out some bidders to get their preferred contractor, or the engineers are making demands without knowing what the machinists and machines are capable of achieving.
Click to expand...
Trust me, as an aerospace engineer I *really* got the part that the flatness you were talking about was on some more or less random H beams the *first* time you mentioned it, because I am not a fornicating imbecile, and I gladly concede that your insight on this particular issue quite likely *vastly* exceeds mine and consequently apologize for not being able to make any constructive contribution to this very particular discussion, simply due to my lack of pertinent experience on H beam flatness and lack of relevant information, but more elaboration on EXACTLY *WHAT* particular H beams you are referring to would *VASTLY* facilitate our online discussion. PRECISE DETAILS PLEASE. Ron Swanson out.
 
Last edited:
aim9xray said:
Doesn't matter, it was signed up to. "You did read the fine print?"

Is it realistic? Different issue. NOW we are into theology.
Click to expand...
Once again, not knowing any engineering and scientific details, please, on what this pseudo/fake discussion is all about - Engineering has no room or patience for any pseudo voodoo hoodoo hokum. Science. Rules. Supreme. THEOLOGY HAS ABSOLUTELY NOTHING TO DO WITH IT.
 
Last edited:
@martinbayer @Scott Kenny : we can easily say that this kind of dissociation b/w design and manufacturing suppliers are often the result of a failure to properly incorporate real matching tolerances during the design phase or selecting a supplier without the proper vetting by engineering. Both are, sadly, fairly commune occurrence in 21st century Europe due to the absurd division of labor as Martin hinted...
 
By the way, Airbus is also complaining about supplier quality issues on A320 fuselage pannels

https://www.airbus.com/en/newsroom/...commercial-aircraft-delivery-target-maintains
Amsterdam, the Netherlands, 3 December 2025 - Based on a recent supplier quality issue on fuselage panels impacting its A320 Family delivery flow, Airbus SE (stock exchange symbol: AIR) is providing an update to its commercial aircraft delivery guidance for 2025.

The Company now targets around 790 commercial aircraft deliveries in 2025.
Click to expand...
 
TomcatViP said:
@martinbayer @Scott Kenny : we can easily say that this kind of dissociation b/w design and manufacturing suppliers are often the result of a failure to properly incorporate real matching tolerances during the design phase or selecting a supplier without the proper vetting by engineering. Both are, sadly, fairly commune occurrence in 21st century Europe due to the absurd division of labor as Martin hinted...
Click to expand...
One of the reasons Dassault moved it's design bureau to Merignac in 2021
https://www.usinenouvelle.com/artic...se-dote-d-un-nouveau-bureau-d-etudes.N1149607
[Press Release] Dassault Aviation’s Mérignac Plant Gets a New Design Office

The new building at Dassault Aviation’s Mérignac plant will allow the group to bring closer together the teams that develop the aircraft and those that build them.
[...]
The creation of this building, part of the manufacturer’s “Piloting Our Future” transformation plan, aims to bring together the teams who design the aircraft and those responsible for their final assembly. The goal is to efficiently integrate, from the design phase onward, all the professions involved in a product’s lifecycle (industrialization, procurement, manufacturing, and support) in order to optimally meet the needs of clients. This approach is also intended to achieve objectives in performance, quality, timelines, and risk management.
Click to expand...
 
martinbayer said:
Engineering has no room or patience for any pseudo voodoo hoodoo hokum. Science. Rules. Supreme. THEOLOGY HAS ABSOLUTELY NOTHING TO DO WITH IT.
Click to expand...
Thank you. That was precisely my point. We are in violent agreement.

It is too late (as a practical [non-engineering] matter) to challenge the specification after you have been found in non-compliance halfway through the production run. Now you are deep into MRB waiver-land. And talking through your contracts weenies about negotiating for "consideration". (Best case.) That's real-world.

Let me restate with more words ... Arguing the specification at this point is as actionable and practical as theology.
 
aim9xray said:
Thank you. That was precisely my point. We are in violent agreement.

It is too late (as a practical [non-engineering] matter) to challenge the specification after you have been found in non-compliance halfway through the production run. Now you are deep into MRB waiver-land. And talking through your contracts weenies about negotiating for "consideration". (Best case.) That's real-world.

Let me restate with more words ... Arguing the specification at this point is as actionable and practical as theology.
Click to expand...
Correct.

But when the spec does not appear to be achievable by shops that were specifically invited by Airbus to make the parts, you gotta start questioning why that spec exists in the first place.
 
TomcatViP said:
Or why that shop ;)
Click to expand...
Again, I'm talking surface flatness specs that go from "in spec with the part secured to the fixture" to "out of spec as soon as the clamps come off."

Ten-thousandths of an inch across multiple feet of distance. Hundredths of a millimeter per meter, for those that don't think in SAE.

Flatness specs that you only meet within a narrow temperature range around room temperature, because thermal expansion in an H beam will distort the part.

In places that it does not make sense to develop the fixtures/clamping procedure to meet the spec, like the web of an H beam. The big flat parts (not sure about technical name) are easier to justify and achieve such extreme flatness.
 
I learned early in my engineering career how easy it is to over-tighten a tolerance (almost always unnecessarily) to the point it is nearly impossible to make. Our machinists drove home the point by making me make the parts I drew.

Then they took that part out of drawing tolerance by holding it in their hand for a couple minutes. Another lesson learned.

It would not surprise me in the least if the people at the supplier who deal with the contracts didn't bother talking to the people making the parts or their engineers. I/my employers have been shafted many times by poorly written contracts where those doing the writing and agreeing never bothered to ask anyone with a technical background.
 

Similar threads

Back
Top Bottom