225 cleared to fly in UK & Norway
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
AnFI,
I did not read that the planet gear spherical roller bearings from the two suppliers actually had different "conformal curvatures" (or osculation ratios). What I recall was that they had different roller widths, different contact angles, different retainer designs, and different contact stress levels at similar analysis conditions.
I did not read that the planet gear spherical roller bearings from the two suppliers actually had different "conformal curvatures" (or osculation ratios). What I recall was that they had different roller widths, different contact angles, different retainer designs, and different contact stress levels at similar analysis conditions.
Interesting Video from Airbus showing the exact failure that many of us here zeroed in on from the first photos of the broken pinion. Alarming that blame is cast on handling damage. Spalls move away from dents and get bigger, further away. If it initiated at a dent, it would still be there. This was no dent and they know it. Further, although spalling is the initiating trigger, they neglect to discuss where the fatigue load comes from. The same fatigue load that caused multiple fractures I would suggest, perhaps not coincidentally 90 degrees apart. They probably do not understand the loads but here is another thought: On such a raceway, fatigue damage will very likely initiate sub-surface. the fatigue itself could cause a spall, not just the other way around. Which means propagation toward the teeth may already be underway by the time the spall forms, let alone gets big enough for particle detection. The problem: Both suppliers parts are subject to the load, and thus subject to the failure. Hopefully they guessed at a low enough life to preclude that. But all the same, is this a hope based argument rather than a data driven one? It would be a shame to see a third preventable event, when they should by rule understand the loads.
Why would the regulators accept that handling damage argument if the evidence is not there?
If there is fatigue caused spalling, surely it should be possible to replicate the failure. We've had more than enough time to run a 10,000 hour test if needed.
Yet replication has not been achieved, which indicates that some aspect of the failure mechanism is still being overlooked.
If there is fatigue caused spalling, surely it should be possible to replicate the failure. We've had more than enough time to run a 10,000 hour test if needed.
Yet replication has not been achieved, which indicates that some aspect of the failure mechanism is still being overlooked.
Why would the regulators accept that handling damage argument if the evidence is not there?
If there is fatigue caused spalling, surely it should be possible to replicate the failure. We've had more than enough time to run a 10,000 hour test if needed.
Yet replication has not been achieved, which indicates that some aspect of the failure mechanism is still being overlooked.
If there is fatigue caused spalling, surely it should be possible to replicate the failure. We've had more than enough time to run a 10,000 hour test if needed.
Yet replication has not been achieved, which indicates that some aspect of the failure mechanism is still being overlooked.
They don’t claim both gearbox that failed fell off a truck.
Just trying to throw enough conjecture around that the fact that they do not know what caused the failure is overlooked.
It’s the modern world: do something, to make everyone feel better, even if it wouldn’t have fixed the problem that caused the reaction in the first place.
After scum hijacked several airliners in flight, we were all forced to chain our helicopters up. (Fines issued for unlocked machines)
After a madman in Vegas shoots people from 450m away with an arsenal of modified automatic rifles, lever action shotguns are banned in QLD (with the Vegas atrocitie cited as the justification by the Premier)
The rotor detaches from two similar AH machines. OH times are reduced even though there is no evidence this was the issue.
At least they did something
Join Date: Sep 2007
Location: Birmingham
Posts: 244
Likes: 0
Received 0 Likes
on
0 Posts
An interesting and very slick film about a unique situation. Airbus have now removed the part involved in the two occurrences. Additionally, they have introduced new sensors and sensibly restricted operations to minimise the possibility of recurrence of any of the identified failure modes. Operators can go one step further of course; by not using the 225 at all they can guarantee no recurrence. However, even that will only eliminate the risk in the 225. MGB failure can happen in all helicopters not just Super Pumas; managed risk is part of every life in helicopter ops. But most take the view that in the current economic, HSE and industrial relations landscape, completely removing the 225 is their safest bet. Unless and until the actual failure mode can be identified and removed from the equation Super Puma ops are likely to remain a minority sport.
Last edited by birmingham; 2nd Nov 2017 at 12:00.
I know the “Super Puma” name is bandied around by the uniformed press.
This is a Professional Pilots forum.
Lets be clear L, L1 model Super Pumas are unaffected.
L2 Super Pumas, and EC225s share the gearbox design, that on two occasions, has catastrophically failed.
This is a Professional Pilots forum.
Lets be clear L, L1 model Super Pumas are unaffected.
L2 Super Pumas, and EC225s share the gearbox design, that on two occasions, has catastrophically failed.
The problem is that the gearbox has had it's load lifting requirements expanded from 6,300 kg. from the original 330 A to 11,200 kg. in the same dimensional gearbox. You can only stuff new wine into old bottles so far.
Join Date: Sep 2007
Location: Birmingham
Posts: 244
Likes: 0
Received 0 Likes
on
0 Posts
I know the “Super Puma” name is bandied around by the uniformed press.
This is a Professional Pilots forum.
Lets be clear L, L1 model Super Pumas are unaffected.
L2 Super Pumas, and EC225s share the gearbox design, that on two occasions, has catastrophically failed.
This is a Professional Pilots forum.
Lets be clear L, L1 model Super Pumas are unaffected.
L2 Super Pumas, and EC225s share the gearbox design, that on two occasions, has catastrophically failed.
Can I safely assume that none of these military a/c are fitted with the problem box even the later model H225Ms?
Last edited by birmingham; 3rd Nov 2017 at 16:46.
Apologies - but even to the familiar the endless branding and re-branding can confuse. I'm only discussing affected aircraft sometimes referred to as Super Puma Mk 2 i.e. AS332 L2 EC225 (or should that be EC225 L2?) or H225. Not being military I have no idea what MGB is fitted to the models formerly known as Super Pumas and branded since the 90s as AS532, Cougars, EC725M, Caracal, Super Cougars or H225M! I haven't included national military derivations (I think!) or locally built versions for the sake of sanity!
Can I safely assume that none of these military a/c are fitted with the problem box even the later model H225Ms?
Can I safely assume that none of these military a/c are fitted with the problem box even the later model H225Ms?
In Australia and nearby, AS332L and L1 are operating on oil and gas operations.
The clients, workers and pilots know the difference.
Not sure I'd describe that as testing.....
Maybe it was a testing flight for the pilot, as all that hand flying be can hard to do when you are used to just pressing buttons 
The aircraft certainly wasn't tested! Perhaps the crew were scared something bad might happen
The aircraft certainly wasn't tested! Perhaps the crew were scared something bad might happen
Last edited by Apate; 7th Nov 2017 at 19:47.
Would it be pointless to run some of these gearboxes to destruction?
Documenting the failure paths in order to provide some improved understanding of these disasters. Perhaps we might even learn something new.
Documenting the failure paths in order to provide some improved understanding of these disasters. Perhaps we might even learn something new.
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
Consider what would be required to bench test say 5 EC225 main gearboxes to their service life limit. There is the huge cost to manufacture the gearbox test articles. There is the significant cost to operate the test stand for a couple thousand hours required for each test procedure. At something like 2000 hours per gearbox test, it would take over one year to complete the 5 tests operating non-stop.
Join Date: Jun 2016
Location: Aberdeen
Posts: 90
Likes: 0
Received 0 Likes
on
0 Posts
Since it is known exactly which component failed (the 2nd stage planet gear/bearing race) it is more scientific to conduct tests on examples of just that component in isolation and simulate the operating conditions.
A test rig used to examine propagation of rolling contact fatigue cracks in bearing steel is shown in this technical paper. Whilst that test rig is designed for external bearing surface test pieces (because it is easier, for one thing) it should be possible to design a rig to test internal surfaces too although space would be much more restrictive.
Rather than just testing to destruction, such a rig allows tests to be paused and the test piece removed for examination at intervals to measure and chart growth of the defect.
In the above test the focus is mainly on crack initiation at or close to the surface and is a result of cyclic contact stresses, often increased in magnitude (up to 4.76GPa) to accelerate the fatigue process. Whilst lubrication is provided in this test it is considered only at the surface.
There are many other technical papers available which investigate lubrication effects not only at the bearing surface but also with seeping into and pressurization inside a crack. This paper (also available here) compared different lubricant additives with the same synthetic base oil to investigate their effects on lubricant seepage into a developing crack. Not wishing to get too technical, it should be clear from a brief scan of these papers that the effect of a lubricant is far more complex than just some oily stuff that you either have in your gearbox and are OK with, or have lost and are (or very soon will be) in trouble without.
Lubricants work mechanically, hydraulically, thermodynamically and chemically. Lubrication is a science in itself and I find it very surprising that it receives little or no mention in investigation reports to date.
I previously queried whether a dramatic reduction in AH statistics of spalling or pitting circa 2009 may have been due to some change in lubrication/lubricants. Having since read up a little on tribology and the different effects of some lubricant additives on the surface (mostly beneficial) and inside cracks (sometimes detrimental) I would not be at all surprised if the next report investigates this area in considerable depth.
Last edited by Concentric; 9th Nov 2017 at 07:57. Reason: Added unrestricted link.
Thank you, Concentric, for your insightful comment.
It makes sense to focus on the component that failed, rather than the entire assembly.
As you point out, that minimizes the cost and maximizes the test flexibility.
The possibility that a change in the lubricating compound helped suppress the failure symptoms in this design, although very plausible, is quite novel. How does the certification process deal with such a possibility?
It makes sense to focus on the component that failed, rather than the entire assembly.
As you point out, that minimizes the cost and maximizes the test flexibility.
The possibility that a change in the lubricating compound helped suppress the failure symptoms in this design, although very plausible, is quite novel. How does the certification process deal with such a possibility?
Join Date: Dec 2008
Location: N/A
Posts: 845
Likes: 0
Received 0 Likes
on
0 Posts
riff raff
just a hypothetical question
How many gearboxes would you have to run and for how long under what conditions
to demonstrate a catestrophic in service failure rate of < 10^-9 per hour?
etudiant "As you point out, that minimizes the cost and maximizes the test flexibility." minimising cost !
"focus on the component that failed, rather than the entire assembly."
so if you wanted 10^-9 for the gearbox and there are 20 planet gears then would they have to be better than 5x10^-11 /hr reliable eh?
how would anyone ever establish that? is it even possible?
what is the acceptable catestrophic failure rate for a gearbox? is there one?
just a hypothetical question
How many gearboxes would you have to run and for how long under what conditions
to demonstrate a catestrophic in service failure rate of < 10^-9 per hour?
etudiant "As you point out, that minimizes the cost and maximizes the test flexibility." minimising cost !
"focus on the component that failed, rather than the entire assembly."
so if you wanted 10^-9 for the gearbox and there are 20 planet gears then would they have to be better than 5x10^-11 /hr reliable eh?
how would anyone ever establish that? is it even possible?
what is the acceptable catestrophic failure rate for a gearbox? is there one?
Last edited by AnFI; 8th Nov 2017 at 18:49. Reason: maths
AnFi what are the alternatives?
AH currently has a multi billion dollar program near death. In that context, blowing up a hundred gear boxes would be a cheap solution, if it helped restore the type to market acceptance.
The logic for the planet gear reliability calculation is yours, but note that we had 2 catastrophic failures in about 2 million hours of operations iirc, so we are a long ways from 10 to the ninth.
I'm no regulator, but I'd think that the acceptable rate of catastrophic gear box failure is zero. I'd hope regulators would not be bamboozled by sweet talk about 9 sigma reliability. We don't tolerate aircraft whose wings fall off either.
AH currently has a multi billion dollar program near death. In that context, blowing up a hundred gear boxes would be a cheap solution, if it helped restore the type to market acceptance.
The logic for the planet gear reliability calculation is yours, but note that we had 2 catastrophic failures in about 2 million hours of operations iirc, so we are a long ways from 10 to the ninth.
I'm no regulator, but I'd think that the acceptable rate of catastrophic gear box failure is zero. I'd hope regulators would not be bamboozled by sweet talk about 9 sigma reliability. We don't tolerate aircraft whose wings fall off either.
Testing is great. Challenge is "How" to test.
Interesting points by Concentric, but I fear that the focus on the spalling has led people to examine what we typically understand from spalling events on a bearing raceway, and that this event was NOT a typical spalling event due to ordinary Hertzian contact stress. Tell us more about the other fractures, and then we can talk about the likelihood of spall propagation to fatigue crack initiation and propagation happening on three separate regions at similar timelines. Maybe more likely that general rim fatigue began in multiple locations and the subsurface initiation produced a spall. Not the other way around.
So we rumor folks do not know the details of the ring stress loads on this planetary arrangement, maybe AH does but probably not. If you don't, then a test would also probably miss the load, and the test worthless in real benefit (Finding and correcting the problem). Most test stands lack something, even if just the ability to represent a larger load for test purposes. A piece-part test can do that, but then lacks the rest of the appropriate interface. A classic problem for engineers, not so easy.
Interesting points by Concentric, but I fear that the focus on the spalling has led people to examine what we typically understand from spalling events on a bearing raceway, and that this event was NOT a typical spalling event due to ordinary Hertzian contact stress. Tell us more about the other fractures, and then we can talk about the likelihood of spall propagation to fatigue crack initiation and propagation happening on three separate regions at similar timelines. Maybe more likely that general rim fatigue began in multiple locations and the subsurface initiation produced a spall. Not the other way around.
So we rumor folks do not know the details of the ring stress loads on this planetary arrangement, maybe AH does but probably not. If you don't, then a test would also probably miss the load, and the test worthless in real benefit (Finding and correcting the problem). Most test stands lack something, even if just the ability to represent a larger load for test purposes. A piece-part test can do that, but then lacks the rest of the appropriate interface. A classic problem for engineers, not so easy.