Although it´s "silent boat" in this thread since 5 days I´d rather like to "bang" the door one more time. We have worked through the case fairly well, I think. However, why do we stop close to the point where "Pandoras Box" is pulled open ? I like to refer to Turbine D´s list posted in Nr. 1888

Quote:
One way to look at it is this way: (A) leads to (B), (B) leads to (C), (C) leads to (D) and (E) leads to failure. This is the classical failure methodology of a complex system. Looking at engine #2's failure,

(A) = Unknown
(B) = Unknown
(C) = Unknown
(D) = Stub pipe
(E) = IPT Disc burst/major engine failure/damage to aircraft

The trick is to identify (C), (B) and (A), (A) being the root cause.[/quote]


Rolling down that list reversal way I´d like to start at

(E) : IPT Disc burst/major engine failure/damage to the aircraft - symptom caused by "D" through "A"
(D) : Stub pipe fracture /oil spill - probable contributional cause ? sympton
caused by "B" or "C" or "A" ??
(C) : vibration - light on LP and IP shaft, severe on HP shaft - sympton caused
by "B" or "A" ??
(B) : rising oil temperature, wear of bearing ?? - sympton caused by "A" ?????
(A) : the real start and cause of the desaster, this one triggered all the other
recorded events.

I believe there is something technically that I can not cover with my knowledge about sophisticated jet engines, but that engineers might know of. There fore this as my last try to get a logical answer to what went wrong in that Qantas engine.

:D

Yes, we will not rest until RR goes in and removes one of those bearings ... their pick of which one ;)

T-1000 Bearings?
 

Is there one less in the T-1000?

bearing ??
 

Smilin_ Ed, I doubt that there are less bearings in the T 1000 than in any previous 3 shaft engine.
If you want to check, this link gives access to all the big engines
Large aircraft engines - Rolls-Royce

lomopaseo, how sure can one be that a simple switch to a stronger bearing - or several - will solve the obviously existing problem ??
Do you think it´s just bearing ??

Trent 900 Compared To GP7200
 

Annex14

Here are the cutaway views of the two engines. Although one cannot see the details clearly, the general engine layout and differences can be seen. To see the actual details, one has to view the engine and see all the parts laid out on tables.

http://www.pw.utc.com/StaticFiles/Pr...taway_high.jpg

http://www.rolls-royce.com/Images/br...cm92-11346.pdf

Several things strike me: If there is any scale at all between the artistic renditions, the thrust ball bearings in the GP7200 look larger than those in the Trent 900. In the turbine area of the GP7200, the emphasis on slowing a speeding turbine rotor can be seen on both rotors where there would be blade on vane contact, there is no spacing concerns. The stage two turbine rotor (similar to the IPT rotor of the Trent 900) sits close to the structural frame behind it and the blade would contact the struts if this turbine disc were to move back.

Both engines deliver the same thrust, both have the same fan diameter, both have 14 compressor stages, the GP7200 has 6 LPT stages verses 5 for the Trent 900. The GP7200 is slightly longer and weighs 956 lbs. more. It has been reported the GP7200 has a 1% better SFC than the Trent 900, but RR disputes this. Both engines appear to be capable of being mounted on a common designed pylon. The maximum HP spool speed of the GP7200 is slightly higher than the Trent 900, but the maximum fan speed is lower on the GP7200.

For those wondering why the Trent 800 couldn't be used, you can't just clip 5" off the fan without rebalancing the rest of the engine, especially if you are being pressed on SFC to meet aircraft range goals. So two things happened on both engines, incorporation of 3-D highly aerodynamic efficient swept fans, higher HP spool speeds, higher pressures and temperatures, all to generate thrust requirements and deliver good SFC.

I do wonder if the counter-rotational feature (new to a commercial high by-pass engine) had any effects on bearings, frequencies or vibration in the Trent 900.

I enjoy this thread as it is filled with useful info.
To add to Turbine D input
The GP7200 ball thrust bearing carries greater radial loads as well as thrust loads where the T900 has a large roller brg to carry the radial loads with the thrust loads transferred to location bearing adjacent the engine mount. Roller bearings typically have higher load capacity than ball bearings for a given size.

The GP7200 has had its fair share of problems, Fan balance issues that have cracked airframe structure and lowered engine mounted accesories life spans. This has generated huge manhour costs and extended ground times but to date has not exploded mid-air. GP7200 1 , RR nil

I see your point except the T800 was a known off the shelf product, now very reliable with the thrust avail to do whats asked plus more. So why the T900 when it requires greater SFC to offset its greater weight before it offers any greater efficiency to the aircraft. So at a glance it does not add up unless huge future promises (SFC) were attached to the T900 model that could not be acheived from current models ?


This is true but RR have been building a contra rotating core for the harrier since the 1960s so I would not call it new to RR and the location IP bearing that supports the fan location bearing turns in the same direction i.e. the ball race speeds have not been a marked increase. This bearing design has been a thorn in RR side since the conception of the RB211 but again is not new technology.
I suspect we will find out once the hype has died down and RR PR machine have massaged and cajoled its customers to accept a modified core engine at a lowered price with the guarantees of increased efficiency ?? (read reliability)
The money will have to come from somewhere and I suspect it will be shared across RR EADS and customers all of whom have committed so much can not afford this to fail. Time will tell.

Turbine D - Bolty McBolt

Once again a very helpful explanation to a layman in jet engine technology.
Especially I like the comparison of GP 7200 and Trent 900. Within a glance it is clear that the construction of those elements that carry the dynamic loads are much lesser concentrated on one small area in the GP 7200 than in the Trent 900. Wether that has a real influence on the problems we have seen on the QF - Trent I only can imagine / assume. But if this problem exists for so many years - Bolty McBolt says since the beginning of the RB 211 - and the production of that type engines goes on with now and than a "big bang" one must ask the question wether there has been ever an in depth investigation towards possible harmonics???

I remember that there have been aircraft desighns that used a long driving shaft for the propeller. Few of these even had counterrotating props. Some of the types ran into kind of severe shaft failures - as far as I remember - caused by kind of harmonics or resonant frequencies ?? I may be corrected at this, too long ago. Doesn´t know if this is applicable in now a days desighn technologies. But what if, . . . . . the counterrotating HP component generates a harmonics that is transferred via bearings the different shafts to other parts of the engine ?? Is that thinkable ?? I must admit, my confidence in RR - engineers abilities exclude such a scenario.
Again I have to cry for help by the engineers !!

DERG

Is that yellow marking just a gag or is it true ??. I have seen some Harriers - must admit on fly by - but havn´t seen something like that. But I can imagine that it might have been indeed a measure of caution.
Jo

there is also another difference concerning spline location
in GP the spline is located directly 'inside' bearing and in T900 the IP spline is 'outside' of the bearing plane. The difference becomes significant under dynamic load when any radial forces and radial play comes into account.

Except the Pegasus was originally a Bristol-Siddeley design, whereas the RB211 from which the modern Trents were derived was from Barnoldswick.

Loving the discussion (despite a lot of it going over my head), shame about the RR-bashing though.

All kinds of vibratory modes/harmonics,coincidence, backward whirl, etc. have been with us for a very long time in gas turbines and Props. Today's engineers have computer models that keep track of the forcing functions and the receptors. Design rules exist to maintain margins between RPM-range drivers and receptors both within and outside the allowed operation limits. In one sense the margins consist of limits on measured vibratory stress while in the case of higher speeds than normal a healthy margin in speed between normal operation and a dangerous vibration (GE CF6, National Airlines)

The problem comes in when operation is either abnormal (damaged parts creating unbalance or aerodynamic burbles) or parts which have lower than expected fatigue margin due to manufacturing flaws. And one of the largest corntributors have been misassembled parts which seriously affect the stiffness of joints between parts.

Nothing that RR doesn't already know and adresses

Thanks thats sure a brief but helpful explanation. As I expressed already, would be hard to understand if the RR engineers had not addressed that problem properly.
Yet, we are chewing on that cause (A) - the root cause.
Jo

Bolty McBolt

In my previous post, I erred on the clipped fan and re-balancing scenario. Actually, I should have been referring to the GE90, not the Trent 800. The core of the GP7200 engine comes via the GE90 core. So when the fan diameter was reduced to 116" from the normal ~123" diameter fan used on the GE90, some other things had to be adjusted. The GE90 core evolved out of the "Energy Efficient Engine". A scaled GE90 engine was built to test the concept from which the true GE90 emerged. So GE had good knowledge as to how to scale downward the core to match fan diameter. In fact, the GEnx engine basically uses the scaled GE90 core.

I just think the drive for both the Trent 900 and GP7200 engines was to improve specific fuel consumption (SFC). If you look at the CF6-80C2, the model used on the B-747-400, The sea level SFC is 0.316, at cruise it is 0.605.

If you look at the Trent 800 on the B-777, the sea level SFC is 0.35, at cruise it is 0.56.
If you look at the GE90 on the B-777, the sea level SFC is 0.324, at cruise it is 0.52.
It is estimated the GEnx on the B-747-8, the sea level SFC is 0.27, no figures yet for cruise.
Theere are no SFC published figures out yet for the Trent 900 or the GP7200 engines, except for the comments from the Emirates CEO.

So with the trend of less fuel usage (improved engine efficiency), IMO, the Trent 900 and GP7200 fits somewhere nearer to the GEnx.

Qantas A380 & B747 News
 

Some news out of Qantas:

Video: Qantas quashes A380 LA flight rumours

Blow-up grounds Qantas flight bound for LA | The Daily Telegraph

Turbine D - DERG

I've been following with great interest the discussion in this thread relative to the several failure scenarios that have been set forth re: Qantas T900 uncontained IPT failure. Back in the conversation (#'s 221 and 223) you briefly brought up the topic of oil and possible coking problems during heat soaking. There was a brief suggestion that inclusion of an external oil circulator unit would have been a sensible part of the overall design for longevity.

Could you discuss from your engineering standpoints, the possible or probable effects of short and long term coking on the actual performance of the hot section bearing(s) in this engine? I've been wondering if it's possible that bearing performance from coking could have degraded to the point where it may have overheated, chattered, and or vibrated itself to the extent that it either partially or even totally disintegrated, whilst also impacting the stub pipe. What I'm asking is, is it possible the bearing itself was causal in the sequence of events ending in the IPT overspeed? Maybe I missed someone discussing this very event in an earlier part of the thread?
Failures of this type are not unknown, of course, in automotive and aircraft turbochargers. The simple solution has always been to, one way or another, make sure the shaft bearings are bathed in circulating oil after shut down. Even old dirty mineral oil is ok to stop most coking, just so long as it keeps coming. If this solution is found necessary for T900 bearings I'd almost bet that it could be done for far less penalty than a 50kg bolt-on accessory.

Thanks everybody for all your great contributions to this thread. What a wonderful resource this forum is.

Turbine D
 

Oh NO !!! Not another Trent engine smashed to pieces. Lucky enough this is a contained failure and it happened on the ground.
Guess, finally someone might have to start thinking !!!

The oil coking problem in pipes was detected and explained on some earlier incidents with other type Trent engines. The pipes involved where bearing chamber vent pipes. It was anticipated in investigation that the coking blocked these pipes leading to heating and failure of the bearings.
It was also in the first segment of this thread once mentioned in conection with the QF 32 engine failure. I didn´t even know that oil coking could become a cause of problems in bearings. But I can see your point.

The coking problem is genuine, and may be influenced by operational techniques.

Tactical aircraft oft perform rapid turnarounds, reloading/refueling as fast as possible. If the engine is shut down soon after high-power ops, and if the engine has an oil-fuel heat exchanger in the high-pressure fuel manifold, the hot oil works on the static fuel mass and may boil the fuel, inducing fuel vapor into the hot combustor, and causing a brief post-shutdown fire. Probably not harmful, because of brief duration, but likely exciting.

In the commercial world, a proper cooldown period is important to prevent coking, especially when T900 "normal" oil temps are as high as 180C. In thinking about A380 ops, since only inboards are equipped with reversers, they are likely to see hotter oil at shutdown than the outboards.

And so it is possible that something in QF shutdown timing technique is aggravating the coking problem. I suspect QF and R-R are already examining this issue. :8

DERG:The two universal, perennial complaints of the engineering profession:

"To solve the problem, we need more data" - and:

"There's so much data, we need more time to analyze it all" :O

DERG

Howdy. Rather than dredge up my old posts, I'll rephrase and update. The "Pipe" is NOT the victim of some inattentive Bench Monkey.

First, notice the "Land", or "Ledge" that supposedly resulted from a "counterbore" gone awry. Look closely at the "corner" enunciated at the transition from ledge to Wall.

It is COVED. It shows no sharp delineation twixt the bore and the "Shoulder". This is 100 per cent the result of wear, and given its locale, almost certainly caused by vibration.

Next, note the Striae on the wall of the inner surface of the "normal" pipe. My opinion is a pulled extrusion process assuredly excluding any secondary operation, or Blunder. Gently Helical signatures from a normal process of machining. Now, look at the area of the Pipe proxima; to the "ledge". There are circumferential grooves, the result of metal/metal wear, as this is undoubtedly the joint locus of two fittings. Look closely and see that the grooves are at bias to the ledge, and irregular, random even, and certainly not the result of any machining process.

There is a great deal more. Suffice that ATSB allowed for a wide open gate for later correction: "....MAY have initiated a sequence of failures that led to burst...."

bear

I think you are incorrect in your suppositions, which do not seem to be shared by anyone else.

The 'cove' at the bottom of the incorrect counterbore is the runout radius of the counterbore tool. Radius at this position is essential to avoid a stress concentration factor. Careful design would never tolerate a sharp corner .

The 'bias' in the tooling marks is caused by the feeding in of the tool. It is a shallow angle helix determined by the feed rate of the cutter.

We are all ignorant as to the design surrounding this failed part, but it is probably the female element of an O ring seal. As such, gross frettage of the nature which you assume is unlikely.

Time to move on.

DERG:

I'm not so sure about the labeling on the photo. The region labeled "fatigue cracking" can also be interpreted as a brittle fracture. BUT - The part clearly should have never reached the assembly floor, and should have been a manufacturing engineer's object lesson many months prior.

But the immediate subject of discussion was oil coking, and coking can readily occur without such a manufacturing fubar. :*

Keep in mind the Splines are not situate centrally within the boundaries of the Thrust bearings balance points. There is a reason this powerplant is light, and 'efficient'. The two go together and rely on fine balance and gentle transition through the Thrust range.

Derg, in a nutshell, you have it. Within your parameters add Contra Rotation, higher Wheel Speeds, and ineffective EEC at failure boundary. There was a Punch List at introduction that should have undergone remediation prior to Service, IMO.

There were surmountable obstacles to be sure, but the time to address them is prior to On Wing Commercial service. Not cute to allow actual passengers to piggy back a development programme.

I agree that the labeling is odd in the photo. I tend to believe the actual slow, but ever progressing fracture surface is in the forefront of the photo, out of focus. The "clean" fracture area does look brittle to me, the final break, so to speak. IMO, this part is a nickel-base alloy, an Inconel or Hastalloy type material. Normally a part like this would be investment cast to a near net shape and subsequently machined as required to final desired tolerances. I am sure vibration played a big role in this component failure besides whatever the configuration and how it got to be that way. I am sure oil was slowly seeping from this fracture into the cavity and one wonders how long and how much was there before the final event took place.

It has been silent as to the progress of the failure investigation, no updates from the ASTB since December 23. It makes one wonder about progress being made in going through the total failure progression scenario, identifying the true root cause or causes.

Oil Systems & Coking
 

radken

I would both hope and believe that coking on the failed engine was not a related cause given its rather short cyclic life on wing. Coking in the longer engine service life is a problem and with the higher operating oil temperatures, could begin to occur sooner as cyclic life progresses. Everything depends on accurate metal and air temperature predictions for bearings, sumps and surrounding cavities, particularly in hot turbine areas of the engine. Heat transfer analysis is extremely difficult due to the complexity of the oil flow in and around the bearings. Obviously there is prior history to fall back on, but in a new engine design (IMO, the 970 is one), rig testing of the bearing cluster would be performed to determine bearing heat characteristics and instrumented engine test data would be used to determine air and heat in the surrounding areas. Then the heat transfer analysis program would be refined and run to determine if the design is adequate or not. It is important to keep all oil wetted surfaces below 400℉ and that includes anticipated soak back heat generation. If one does not accomplish this, coking can occur in sumps along the walls and in tubing associated with the sump. The big worry in the 970 would be the roller bearings under the IPT rotor.

But, probably the most difficult item to face is to get all the vibrations/harmonics identified and dealt within the oil tube piping system. This usually winds up being iterative process during the engine testing phase, identification and correction.

To give you some ideas on maximum oil temperatures on various engines:

RB211 Series 335℉
Trent 700 374℉
Trent 800 375℉
Trent 900 385℉
Trent 1000 365℉

CF6-50 320℉
CF6-80C2 320℉
PW4000 350℉
GE90 270℉
GEnx 320℉

A main coking generation factor is, what happens when you shutdown a hot engine quickly as barit1 points out and there is no longer oil flow?

Hope this helps...

It is pretty meaningless to list oil temperatures and attempt to make some sort of comparison unless also listing where in the system the temperature is measured and how this measurement relates to the highest temperature in the system.

KBPsen
 

Hmmm . . . easy speech !! Can you offer a better or more detailed listing ??
Jo

Sure it is easy. It is as easy as listing a series of maximum continuous oil temperatures without at the same time listing where these temperatures are measured, how it relates to the maximum temperature experienced in the system and what oil specifications are used for the particular engines.

It's a bit like comparing turbine temperatures without specifying whether you are talking about TIT, ITT, EGT and what material the turbine is made of.

KBPsen

All the RR engines, RB211, Trent 700,800,900 and 1000 are combined scavenge temperatures. So are the PW4000, CF6-50, CF6-80C2 and the GP7000. The two I am not sure of would be the GE90 and GEnx. These two engines, particularly the GE90, may have a different measuring location, not sure.

The point is that the three spool engines generally run a higher oil temperature compared to two spool engines. But, none of the reported temperatures may relate to the highest temperature experienced (the hottest bearing sump) although I think the engine manufacturers know the relationship.

oil - coking - vibration
 

Turbine D
What is the meaning of the 400°F you mention in your post Nr. 257 ? It is too low, as far as I know, for self ignition temperature of oil. Is it than a border temperature for coking ?

In that connection, you mentioned several times the plenum in front of the IPT disk. It is supposed by best knowledge, I assume, to be the space the oil feed tube that finally failed runs through. At the same time it is the supporting structure for the bearing chamber of the IP and HP roller bearings.

How likely would you consider the possibility to be that, although the engine had that bearing problem, was in repair shop from Sept. 2009 untill Dec. 2009, was boroscoped in June 2010, no one had a look into that plenum chamber and detected that cracked feed tube spilling oil into that space ??

If the engine was checked including that section - what I would expect to happen with such a major repair - and there was no oil, it means the leak developed later. If however, this section wasn´t checked at all - means neither in the repair shop nor on the boroscopic inspection - the oil leak and spil could have started quite a long time before there was that oilfire.

The ATSB investigators have come to the conclusion there was an oil fire - so for us thats fact. Question is, when started it and what damage did it ?

If - worst case - the oil leak existed undetected for a longer period of time it appears likely that over the many hours of usage quite a bunch of oil coking has happened. I would expect to have the lower part of that plenum chamber substantially filled with a swamp of coke and oil. For a while the cooling air fed into that chamber might have prevented heating beyond SIT of the coke-oil mixture. It was said that this temperature is substantially lower than plain oil SIT.
So when the fire started that way after engine start at SIN would that fire accomplish a burn through?? or is the material the plenum is made of heat resistant. I think it has to in that hot region of the engine.

So if in fact no burn through occured, just weakening of the lower half of that supporting structure, what consequences would that have on the precise fixture of the bearing chamber ?? Is my assumption correct that the chamber triggered by vibration may have started some kind of motion that made the disk tumble ??

DERG

In your post Nr. 262 you mention now twist, vibration and kinetic energy. So if I understand that post correct, you point to the same causes that I mentioned in my former post about shaft failure with specific engine / propeller installations.
But is that something to consider ??
What I have thought also about several times is balance. I know balancing the plain disk is not the art, mounting several tenth or hundred of vanes on that disk and still keep it in balance is the tricky part.
Probably everyone will say: no impossible ! But yet I would like to ask is it technically possible to balance a compound of pieces so secure that it will run without any imbalance at all possible rpm´s ??? If not ?? If only balanced to bearable amounts of vibration for cruise power ?? That´s the power setting longest used on a flight and appears more important than anything else.

Remember lomopaseo said that all kind of vibration and harmonics are common items in engine development and use that normally are covered by computer programs. That should suggest that those sketched possibilities are not existing. Correct ???

I hope this becomes another trigger to push the efforts on this circle closer to the truth behind the engine failure.

A said before, everyone would be wiser if the pictures of the inner parts of the engine would be released. But I am sure ATSB will come along sooner or later with a more profound report than a prelim ever can be.

Annex14

Oil begins to coke at a temperature of 440℉ to 450℉. The 400℉ is what you attempt to design to in the sump oil wetted wall areas to prevent coking as a safety margin, but, also having an idea as to what the maximum soak back temperature would be/could be through a heat transfer analysis program that examines that entire engine area.

Yes, that would be my opinion.

The bearing problem was with a ball bearing cage and race that I assume to be in the fan/IPC area of the engine. To get at that, the engine would have to be torn-down in a modular fashion. It is conceivable the IP Turbine module containing the plenum wasn't looked at from an internal point of view. It is hard to say what happened for sure. It could have been weeping or slightly seeping or not yet leaking at the time of the bearing replacement. The boroscope inspection was for the spline wear, again, in the Fan/IPC/HPC area of the engine.

The cooling air really isn't that cool in the IPT area of the engine. I would suspect sections of the frame, Casing, hot gas path airfoils (including outer and inner bands and the plenum itself would be made of different materials depending on expected temperatures. The plenum material would have to be a medium temperature material capable of being formed (hot forming?) given the very contoured shape fitting in the cavity in that area. It may also have to be a weldable material. In two spool engines, this would be a "turbine mid-frame", something that was finally eliminated as it was such a pain to both make and then subsequently dealing with never ending problems in service.

IMO, the fire was intense enough to burn through the plenum wall, exposing the bore of the IPT disc to temperatures beyond the disc material's load carrying capability. I am sure the bearing structure supported by this frame was adversely affected by all of what was taking place. It is very possible the bearing upset together with the weakened disc (stretching) could cause an uneven spin plane from normal.

Thanks! That helps a lot in understanding the relations involved.
PM sent

.

DERG
The torque discussion is interesting. I read somewhere that the Trent 900 engine series accelerates from ground idle to maximum TO rpm in something like 5.3 seconds. Seems like a lot of twisting/torque going on in a hurry.

All rotors are balanced after blades are inserted into the disc prior to actual assembly into the module build. Each blade has been weighed to the gram and may be inserted into the disc based on weight distributions. The balancing machine is like that experienced in an auto tire shop except it is more sophisticated and goes to higher speeds.

After final assembly of a new engine, it is filled with oil and put on a test stand to run through its paces while being closely monitored. Vibs are examined quite closely as well as EGT margin. The process takes about 3 hours of engine running time to complete. If all is good, the engine has the oil removed and it is packed up for shipment. If it is not good, the engine is returned to dis-assembly to figure out what the problem/problems are. This doesn't happen often but when it does, the engine is known as a "hanger queen".

DERG & bearfoil:

You are drawing a bunch of speculative conclusions based on rather incomplete data. I know that speculation is intrinsic in R&N, and R-R is holding cards close to the vest, but to read ulterior motive into their silence is over the edge IMHO. :ouch:

barit1

With respect, my conclusions are based on information from professionals who have/are working for the principals here. Consider what you read speculation or not, it is most definitely not "over the top". No signs of Oil problems were found in these engines for upwards of a year prior to this Burst. Oil Fire is not the cause of the Burst, nor is it due repetitive malfunctioning of ridiculous claims of "Misbored" pipes.

This will all out. It won't go away. If you rely only on long cites of lab work and legal claims to allow posting, you will have to wait.

Spline Damage Due To Torque Stress
 

Derg:

actually both two-spool and three-spool engines have shafts inside shafts
when discussing different load components acting in any joint it is important to evaluate actual values of these load components. If not possible to measure than standard evaluations are taken assuming that the loads would be of similar type to existing known solutions (trent 800 et al). The twist load is definitely calculated and appropriately taken care of. Even if combustion may not be constant process but the force to the shaft is transferred via hot air which is a lot more compessible than stiff disc so the total torsional shaft load is very much smoothered by this.

not really and for two good reasons:
(1) difficulty to define the vibration characteristic of the vibrating parts - especially when we have to take into account non ideally stiff (so non linear) elements like spline coupling and bearings
(2) difficulty to define the source, type and size of excitation: bearings, other rotating elements and their relative influence

yes, the phenomena is well known but the new solutions are being developed using new techniques: instead of making models and improving them RR try to digitally model and improve the digital model. But if it is easy to model the KNOWN forces like thrust and imbalance tolerences than the vibrations and wear are the real problems, especially if they develop into not foreseen way.

WojtekSz

I am not sure I understand what is meant here, could you explain more clearly?

Turbine D

I get from Wojtek that he is offering a "Fluid Drive" argument against torque problems that may result in stress wear or fracture. I'm not sure it obtains to this discussion, however, at the cusp of Torque "reversal", there is enormous stress on the rigid coupling, having to translate gobs of torque from components that were at one moment driving in one direction, but quickly changing to "driven" mode. Vibration could mimic this reversal of torque, but in more and shallower "events". Load, Unload, Load, etc. A resonant frequency would cause the same, but even more chaotically. At the "Null" point, where torque is balanced (neutral) any wear or excess tolerance in these bearings would cause chaotic vibration in erratic "Ovoids", a chattering mess of blurred steel balls, raceway and cage. Can you hear the din of bearings "floating" as the result of intense and unplanned for motion?? We've all heard it. Rattle, shriek, and stop.

A little off the (interesting & enlightening) technical discussion, but what is the current situation with the A380 involved ?.

Is she still in Singapore ?. Will she be fully repaired there, or patched up & flown to France for full repairs ? Or is she unrepairable given the substantial wing damage and machined tapered thickness slab construction of the wing skin ?, - (mentioned & youtube illustrated many threads ago).

I recently read a book entitled "The Somerset & Dorset Railway, Then and Now" by Mac Hawkins. An interesting comment is made regarding Winsor Hill Tunnel (Near Shepton Mallet), which was used, in 1968, after the line was closed and lines lifted, by Rolls Royce for destructive tests on the Olympus engine destined for Concorde.

To quote the book, "Up to the late 1980's the tunnel's portals were obscured by massive steel doors, built a little in front of the stonework and supported by a frame. These where constructed as an anti-blast measure by Rolls Royce in 1968, who used the tunnel for destructive tests on the Olympus engine for Concorde. They ran an engine without oil, expecting it to blow up within 20 minutes or so, but in the event it laster for well over two hours !. The tunnel's use for this purpose was only over a few days, planning permission having been sought from Shepton Mallet RDC as a matter of course, in case an explosion caused a change in the local topography"

TurbineD:
torsional load on the shaft and on the spline coupling comes form hot gases from burning process. Bearfoil calls it fluid drive but the fluid is still very compressable compared to any real fluid. This has significant smootering efect. The burning process creates vibrations but these are small compared to overall value of forces acting on a blade (and additionally dampened by the dovetail blade fixing). And than we have a relatively heavy disc where this vibrating loads from many blades are totalled into significantly smoother torque on the coupling

Bearfoil:
i can agree to certain level of vibrations from the burning process but i do belive that it must be very much similar to the very process eqisting in all other aero engines flying all over the world for some time already. We are looking for reasons that are UNIQUE to T900, right?