Lack of lubrication certification for helicopter gearboxes
Thread Starter
Join Date: Aug 2000
Location: UK
Posts: 3,680
Likes: 0
Received 0 Likes
on
0 Posts
Lack of lubrication certification for helicopter gearboxes
Last edited by Thomas coupling; 27th Jul 2017 at 18:31.
If you can, might want to fix the typo in your title. (For a moment I had to think about what ubrication might be). Thanks for the link, I'll need to ponder a bit before any comment.
You asked for comments.
Comments provided.
Because those who wrote the paper won't have to do it. (Yeah, I saw the table at the bottom, and all I can offer is 
.
Regarding paragraph 2.4: would anybody care to offer up a concise version of "where does 30 minutes come from?" (as opposed to 20, or 15, or 10, etc) to bring us up to date? I realize that this has been done to death in various S-92 / Blackhawk / Puma / 225 / 332 threads and bun fights. Reestablishing that old baseline early in this discussion might be helpful. (I see that I may not get my comments in by 31 July, but that's OK, nobody listens to me anyway)
3.2.1. b. makes some points on Failure mode analysis. Like what I see there.
Regarding para 6(a)(1) .. Oh, wait, now it's a 36 minute test. Where does that number come from? (See also 6(e) (1) (iii))
Secondary indications: yeah. Keep moving forward on that thought.
Here's an idea: make the requirement 25 minutes, and substantiate that. Not sure what's so magical about 30 minutes, or 36 minutes. Pick a different number, perhaps?
And every operator of same. (this section explored some financial issues)
Really? If there weren't a social impact, you'd not have been tasked to come up with this.
Comments provided.
The only significant drawbacks of this proposal are economic. This results from the increased cost of certification testing, and potential for increased weight of the gearbox and lubrication system, dependant upon the design solution chosen by the applicant. But the impact to CS-29 helicopter manufacturers is considered to be negligible
. Regarding paragraph 2.4: would anybody care to offer up a concise version of "where does 30 minutes come from?" (as opposed to 20, or 15, or 10, etc) to bring us up to date? I realize that this has been done to death in various S-92 / Blackhawk / Puma / 225 / 332 threads and bun fights. Reestablishing that old baseline early in this discussion might be helpful. (I see that I may not get my comments in by 31 July, but that's OK, nobody listens to me anyway)
3.2.1. b. makes some points on Failure mode analysis. Like what I see there.
Regarding para 6(a)(1) .. Oh, wait, now it's a 36 minute test. Where does that number come from? (See also 6(e) (1) (iii))
Secondary indications: yeah. Keep moving forward on that thought.
Here's an idea: make the requirement 25 minutes, and substantiate that. Not sure what's so magical about 30 minutes, or 36 minutes. Pick a different number, perhaps?
4.1.2. Who is affected This NPA will affect designers and operators of large rotorcraft as well as certifying authorities. The four largest helicopter manufacturers affected by the changes proposed in this NPA are Airbus Helicopters, Bell Helicopter, Leonardo Helicopters, and Sikorsky Aircraft Corporation.
4.5.3. Social impact There are no relevant social impacts for either option.
Lonewolf, re the 30 minute requirement. I believe this first appeared in the UTTAS Material Need document which then became the UTTAS Request For Proposal in terms of the design/test requirements for the aircraft. There was an Army board headed up by a Colonel Bud Patnode which was charged with coming up with the requirements. I talked with him a few times during the fly-off competition, but never about the 30 minute rule. His stories about the internal fights about the skids vs wheels decision were " entertaining ". That group started their work around 1968 I believe, so it is probably safe to say that UH-1 experience in Vietnam influenced the 30 minute number. I do recall that a UH-1 down south ( I was in the Central Highlands 65-66 ) had a transmission oil problem, tried to stretch it and the following aircraft saw his rotor slow down and they went in. There may have been others. Someone with access to the Army Aviation Systems Command, Flight Standards Div records should be able to answer your question definitively.
Last edited by JohnDixson; 28th Jul 2017 at 01:50. Reason: Typo
John, in our neck of the woods (IV Corp) scuttlebutt was a dry transmission was good for either 30 or 60 minutes in the Huey, forget which. Reputedly from a Bell rep. Had occasion to try it out (no pressure) when the filter O ring let go and flew for some period to get to a FSB (<15 minutes, memory as to how long not that good). Occurred during the Cambodia push when spares were impossible to get, hence maintenance reuse of O ring during filter change. Maintenance arrived and I refused to fly it home, they topped it up and flew home. Can tell you a T-53 will run for long enough to take off and get you out of trouble with no oil as well.
Megan, the story re the machine ( 1965 or 66 ) was that it was a few minutes. I found the following link for a later incident, which quoted 2 minutes:
https://www.vhpa.org/KIA/incident/700217101ACD.HTM
https://www.vhpa.org/KIA/incident/700217101ACD.HTM
John, I may be able to find someone who can dig that up. Depends on who retired recently.
The lore in the SH-2F (passed down to us newbies in the early 80's) was that one crew lost the oil in the combining gear box and headed back to their little frigate, at the NATOPS recommended power setting. About 30 minutes later they landed and shut down, so there was one data point that it might work for 30 minutes based on an event. But NATOPS was still all about "get it out of the sky" as the rule.
The lore in the SH-2F (passed down to us newbies in the early 80's) was that one crew lost the oil in the combining gear box and headed back to their little frigate, at the NATOPS recommended power setting. About 30 minutes later they landed and shut down, so there was one data point that it might work for 30 minutes based on an event. But NATOPS was still all about "get it out of the sky" as the rule.
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
That EASA notice was interesting reading. Most of the proposed changes to certification of gearbox designs for loss of lubrication operation seem appropriate, but there are a few areas that could be improved.
The most important thing to consider with rotorcraft transmission systems is that the primary function of oil is for cooling the internal components. The amount of oil flow required for "lubrication" is quite small in comparison to the oil flow required for cooling. So the most effective approach to improving loss of lube performance are design changes that improve heat transfer away from critical areas in the gearbox, such as bearing and gear mesh contacts. Or using the recently developed advanced materials for gears, bearings, housings,etc. that can tolerate higher temps.
Another change to loss of lube qual testing would be requiring two or three gearboxes to be tested. And the gearbox test articles used should have been run so they represent the average condition of a gearbox in service.
Finally, there were a couple proposed requirements that seemed a bit optimistic. For example, the Class 3 failure criteria (Imminent Failure) on p.18 states "the efficiency of the gearbox may be reduced by up to 10%.". A decent new build MRGB probably has an input/output efficiency of ~96%. So a gearbox on the verge of imminent failure after the loss of lube test cycle must still demonstrate an input/output efficiency of >86%. I don't think this particular requirement is realistic.
The most important thing to consider with rotorcraft transmission systems is that the primary function of oil is for cooling the internal components. The amount of oil flow required for "lubrication" is quite small in comparison to the oil flow required for cooling. So the most effective approach to improving loss of lube performance are design changes that improve heat transfer away from critical areas in the gearbox, such as bearing and gear mesh contacts. Or using the recently developed advanced materials for gears, bearings, housings,etc. that can tolerate higher temps.
Another change to loss of lube qual testing would be requiring two or three gearboxes to be tested. And the gearbox test articles used should have been run so they represent the average condition of a gearbox in service.
Finally, there were a couple proposed requirements that seemed a bit optimistic. For example, the Class 3 failure criteria (Imminent Failure) on p.18 states "the efficiency of the gearbox may be reduced by up to 10%.". A decent new build MRGB probably has an input/output efficiency of ~96%. So a gearbox on the verge of imminent failure after the loss of lube test cycle must still demonstrate an input/output efficiency of >86%. I don't think this particular requirement is realistic.
In more recent times the 30 minutes was a piece of PR on a certain model that still does not comply to this day.
Oh hang on a minute ................thats what this thread is actually about.
Oh hang on a minute ................thats what this thread is actually about.
Join Date: Mar 2002
Location: Canada
Posts: 690
Likes: 0
Received 0 Likes
on
0 Posts
When asked
"From memory, if pilot action is required "within several seconds" of a system failure, such as a MGB oil pressure loss, isn't that non compliant with Part 29?"
The PM responded on PPRuNe
"And the S-92's oil protection system protected the transmission to the FAR/JAR with such aplomb that we shut the test down at 3 hours, not 30 minutes, with the system still running along (but admittedly pretty close to its end). That is 2 1/2 hours longer than required. The pilot must activate the system, but has several seconds to do so after clear indications, and the checklist does not say "land immediately."
So don't spread any lies
Oh wait, turms out that in NL terms 3 hours becomes 10 minutes.
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
In my post I also mentioned recently developed materials that provide significantly increased high temperature capability. In particular there is C64 steel alloy (AMS 6509) for carburized gears. Since the early 80's X-53 steel alloy (AMS 6308) was the best available material for carburized gears in terms of high-temp capability, with a tempering limit of ~400degF. In comparison, C64 steel alloy has a tempering limit of ~925degF. The tempering temperature of a steel alloy is important because if it is exceeded in service the material will begin to de-temper and lose strength. For rolling element bearings, there is M50NiL (AMS 6278) which is a carburizing grade alloy that can replace the existing standard high-temp material M50 (AMS 6491), which is a through hardening alloy. While M50NiL has similar high-temp capability as M50, it also provides better fracture toughness and lower friction characteristics at boundary contact conditions. Once again, minimizing friction under marginal lubrication conditions is critical for keeping component temperatures down.
To minimize friction in rolling element bearings under marginal lubrication conditions, a recent development is applying amorphous carbon coatings (DLC) to the rollers/balls. This adds some cost, but otherwise should have minimal impact on an existing design. The only potential issue is how to detect debris generated by this coating.
Megan, the story re the machine ( 1965 or 66 ) was that it was a few minutes. I found the following link for a later incident, which quoted 2 minutes:
https://www.vhpa.org/KIA/incident/700217101ACD.HTM
https://www.vhpa.org/KIA/incident/700217101ACD.HTM
NTSB said it was a transmission failure from a lack of lubrication.
Zalt,
Did the Cougar Crew comply with the Checklist?
How many times did the Co-Pilot tell the Captain the Checklist's last entry said to Land/Ditch?
Was there a 30 Minute Run Dry Time required by Certification?
Did the Aircraft receive Certification?
Have you read the Cougar Report in detail?
Under Section 4.2 ACTION REQUIRED, The Accident Report lays out the basic problem with all of this 30 Minute Run Dry Time issue. The Authorities amended the Requirement to allow the use of the definition "Extremely Remote" and by doing so undermined their own Certification Standard.
The FAA also allowed the use of the Pilot Activated Isolation system and in the Cougar case that was of no use as the entire contents would have leaked out even had the System been activated due to the Filter being the cause of the leak.
http://www.tsb.gc.ca/eng/rapports-re...6/a09a0016.pdf
Did the Cougar Crew comply with the Checklist?
How many times did the Co-Pilot tell the Captain the Checklist's last entry said to Land/Ditch?
Was there a 30 Minute Run Dry Time required by Certification?
Did the Aircraft receive Certification?
Have you read the Cougar Report in detail?
Under Section 4.2 ACTION REQUIRED, The Accident Report lays out the basic problem with all of this 30 Minute Run Dry Time issue. The Authorities amended the Requirement to allow the use of the definition "Extremely Remote" and by doing so undermined their own Certification Standard.
The FAA also allowed the use of the Pilot Activated Isolation system and in the Cougar case that was of no use as the entire contents would have leaked out even had the System been activated due to the Filter being the cause of the leak.
http://www.tsb.gc.ca/eng/rapports-re...6/a09a0016.pdf
Join Date: Mar 2002
Location: Canada
Posts: 690
Likes: 0
Received 0 Likes
on
0 Posts
SASless
I'm loathed to 'answer' questions you have your own answers to! But yes I have read the accident report in great detail.
You might not have realised it but you have conceded that if Sikorsky had actually designed the S-92A to run for 30 minutes after a loss of lubrication, the crash would not have occurred as they would have reached Cape Spear.
This latest EASA NPA does not change that at the time of certification a test with a loss of lubrication (not just a bit of the oil) was already required (see TSB footnote 84).
Despite your question below about what the FO said, in fact the FULL CVR transcript was not published by TSB so we can't examine the whole discussion between the crew and also with their ops centre. Perhaps Sikorsky's sale of the 3 float bag configuration to an operator flying over the Grand Banks may have prayed on their minds. In March in the Atlantic with 3 float bags a capsize and some fatalities was very possible.
The Canadian press had a field day when it emerged that the original marketing material used in Canada HAD claimed the MGB could run for 30 minutes after an oil loss. The TSB also have a section in their report about social media. I wonder which site they were thinking of
oh: Perhaps that material and the Program Managers claim of 3 hours was an influence in delaying a decision despite the RFM?
Perhaps three S-92A incidents just the previous year with oil system anomalies that did not involve oil loss may have also clouded their judgement.
A Norwegian accident report 4 1/2 YEARS AFTER the Cougar accident considered the Sikorsky checklists on oil loss were STILL confusing http://www.aibn.no/Aviation/Publishe...ts/2015-11-eng
I must correct myself. The S-92A gearbox that failed on 6 August 2002 (see TSB page 100) when the test was done the normal way followed by the rest of the industry failed after 11 minutes not 10 as I implied below. A long way less than 30! Its only after that that Sikorsky "revisited" the requirements as the TSB put it and focused on what would not be "extremely remote".
The Cougar accident shows that one failure mode claimed to be "extremely remote" was not. But the oil loss in Canada was not the first such oil loss on the S-92A. The previous year one occurred in Australia. That should have rung loud alarm bells that a key certification claim was false. If we apply the logic that some apply to other helicopters and their gearboxes, the S-92A should have been grounded until the root cause of that failure to meet certification requirements was fully understood....
I'm loathed to 'answer' questions you have your own answers to! But yes I have read the accident report in great detail.
You might not have realised it but you have conceded that if Sikorsky had actually designed the S-92A to run for 30 minutes after a loss of lubrication, the crash would not have occurred as they would have reached Cape Spear.
This latest EASA NPA does not change that at the time of certification a test with a loss of lubrication (not just a bit of the oil) was already required (see TSB footnote 84).
Despite your question below about what the FO said, in fact the FULL CVR transcript was not published by TSB so we can't examine the whole discussion between the crew and also with their ops centre. Perhaps Sikorsky's sale of the 3 float bag configuration to an operator flying over the Grand Banks may have prayed on their minds. In March in the Atlantic with 3 float bags a capsize and some fatalities was very possible.
The Canadian press had a field day when it emerged that the original marketing material used in Canada HAD claimed the MGB could run for 30 minutes after an oil loss. The TSB also have a section in their report about social media. I wonder which site they were thinking of
oh: Perhaps that material and the Program Managers claim of 3 hours was an influence in delaying a decision despite the RFM? Perhaps three S-92A incidents just the previous year with oil system anomalies that did not involve oil loss may have also clouded their judgement.
A Norwegian accident report 4 1/2 YEARS AFTER the Cougar accident considered the Sikorsky checklists on oil loss were STILL confusing http://www.aibn.no/Aviation/Publishe...ts/2015-11-eng
I must correct myself. The S-92A gearbox that failed on 6 August 2002 (see TSB page 100) when the test was done the normal way followed by the rest of the industry failed after 11 minutes not 10 as I implied below. A long way less than 30! Its only after that that Sikorsky "revisited" the requirements as the TSB put it and focused on what would not be "extremely remote".
The Cougar accident shows that one failure mode claimed to be "extremely remote" was not. But the oil loss in Canada was not the first such oil loss on the S-92A. The previous year one occurred in Australia. That should have rung loud alarm bells that a key certification claim was false. If we apply the logic that some apply to other helicopters and their gearboxes, the S-92A should have been grounded until the root cause of that failure to meet certification requirements was fully understood....
Not to read too much into it - Leonardo claim that the 139 has proven to be good for 60 minutes and the 189 for 50 minutes on certification tests.
So it's not as if it cannot be achieved. They obviously put a bunch of work into it.
So it's not as if it cannot be achieved. They obviously put a bunch of work into it.
Join Date: Sep 2011
Location: Europe
Posts: 59
Likes: 0
Received 0 Likes
on
0 Posts
Zalt, I couldn't agree more with your post above. You can see from my previous posts that this is a subject of great interest to me. How could an aircraft continue to fly after it had actively demonstrated that the extremely remote 'loop hole', under which it had been certified, was actually not actually extremely remote after all. And, worse still continue to fly without immediate modifications, after the earlier incident, leading to a complete loss of an aircraft and multiple fatalities.
It is also extremely worrying that the Canadian navy are getting a militarized version of this same aircraft. How extremely remote is extremely remote when people are firing all sorts of s**t at you. A couple of 0.50" rounds through your MGB casing over water and what happens next?
It is also extremely worrying that the Canadian navy are getting a militarized version of this same aircraft. How extremely remote is extremely remote when people are firing all sorts of s**t at you. A couple of 0.50" rounds through your MGB casing over water and what happens next?
Join Date: Oct 2012
Location: SE England
Posts: 111
Likes: 0
Received 0 Likes
on
0 Posts
If a total loss of oil occurs, you comply with the RFM and land immediately.
Given that as far as we are aware from the poor statistical sample of about 2 occurrences - one in a test cell and one for real -
both achieved ~ 11 minutes. Lets put a margin in for error of 50% which is generous and round it down - 5 minutes.
Also not forgetting that on the Cougar flight most of the 11 minutes would have been power off or low power in a rapid descent from altitude.
As the failure was primarily in the TR output you could probably have a failure regardless of the power level.
Another issue on the Cougar incident was the pre-occupation with the fact that the oil temperature remained normal. This is a fallacy that has been perpetuated for years that low or no oil will give you high oil temperature. Theory and experience would prove the exact opposite or it has no correlation. Yet it is still preached.
These guys were looking at an "extremely remote" event and making up for it as they went.
The 92 has had quite a few "extremely remote" events and incidents where "cry wolf" was patched up and lulled a lot of people into the wrong frame of mind.
"Extremely remote" has been redefined if you are looking for an example as it clearly does not meet the alternative.
