JFZ90

I am not a 100% sure which bearing area instigated the fire but have been leaning towards the roller bearings beneath the IPT rotor. In my mind a fire in the cavity here makes more sense to me at this time. If the web of the disc and the bore area gets heated beyond the material strength capability then it could result in the fracture of the power drive arm at the 580 bolt holes, leading to overspeed and rupture of the disc+the blades impacting the stage 1 LPT nozzle ring. The IPC/IPT shaft does not fail and that is why the LPT shaft is shielded in this area. Also, in this process of failure, as soon as the disc is free from the shaft, the IPC rotor spools down in speed (no longer being driven) and this translates into another problem, a HP compressor stall. I do believe one of the "bangs" heard was this as the highly compressed hot gases are blown both forward towards the IPC and back towards where the IPT rotor was. Any time there is a very significant compressor stall due to major component failure, flames come out both ends of the engine. Who knows what this could have done to the IPC ball bearing area.

Also, if the IP shaft severed, The IPC spool would be free to move rearward and there would have been considerable damage to the IP compressor blades/stators with debris going back into the HPC. Damage has not been reported at all to indicate this, at least at this time.

So those are the thought behind my leaning towards the IPT area as the problem area.

What complete rubbish! Have a look at the diagram on page 70. The HP spool has two bearings, the other spools have three each. Main bearing in the centre and roller bearings at each end.

Forget about the splined coupling, it has nothing to do with QF32!

Please stop the nonsense, you're driving me nuts.

firstfloor

I believe you are mistaken. If you look at the diagram, page 70, you will note that the Aft IP Shaft terminates short of what you call the "Third bearing". As a unit, you are correct, but as separate Shafts you are wrong. The terminus of the Wheel (IP) Shaft is short of the #2 Thrust (Ball) bearing. This leaves the Splines to carry the (aft) Shaft's load to the Thrust bearing. Take a look.

I was unclear in former post, The Aft Shaft has Two bearings. I can see where one may think I stated it had none. I can also understand why people would think it had Three. I do think it would be better for Rolls to have people "forget" about the Splines, but the AD's have not been withdrawn.

In fact, if one looks closely at the diagram, the join of the fore/aft Shafts, with its splines, is likely the source of the Oil combustion evidence in the bearing case for Second and Third Ball bearings. Here, in the Slip area, with splines worn, a rogue gas path would deposit Carbon rich deposits on the LPT Shaft (see pic). Such a "leak" would also explain loss of (some) N2.

The way I look at it is that although a shaft consists of two half shafts, the half shafts are joined at a rigid coupling so that they behave exactly as if they were a single unit. We do not know the secrets of the coupling itself except that it consists in part of a splined joint transmitting torque from the turbine to the compressor and at least one other element transmitting axial loads between the half shafts (because the splined joint alone cannot transmit axial tension).

The fact that wear rates in the coupling of some engines were found to be higher than expected does not mean that the coupling was anywhere near close to failing. I would assume that at some stage the wear develops to the point where movement in the joint goes out of acceptable limits and has to be removed but this would be well before a hazardous situation is reached.

firstfloor (with apologies to wingsfolded)

You have essentially encapsulated the AD itself. There is concern that spline wear can lead to aft migration of the Aft Shaft, so indeed the Splines DO attenuate Axial movement, through the abutment face referenced in the AD. If one reads the Current AD, and prior AD's incorporated by Reference to them, one knows that the wear limit for the Spline is approximately 2mm, as measured on the "crest" of the Spline.

The Splines are helical, and allow for some inter rotation as the Thrust goes from Drive to Driven. Notice that the only (ball) bearing supporting the Aft IP Shaft is the #2 Thrust (Ball) bearing, and that this Shaft is directly supported through the Spline interface.

Nothing has been released as to exactly where the "Oil Fire" took place. Some assume at the Roller Bearings just forward of the IP Disc, others (myself) believe the fire occurred in the cavity containing IPC and HPC ball bearings. The telltale on the LPT Shaft suggests placement directly inside the joint where the Fore and Aft IP Shafts connect. The Aft IP Shaft end does not extend through the width of the Thrust bearing, meaning that all the energy between the two Shafts, is borne through a narrow section of the Splined Joint.

Keep in mind also that the Splines DO attenuate thrust (resist it), though not completely, they terminate at a ring that effaces the inner race of #2 Ball Bearing. Also take note that only a short section of Splines needs to fail to allow aft travel, they do not need to fail utterly.

I can appreciate your concern, but this is not nonsense.

The AD you refer to does explain what would happen if the splines wore through. In practice, no flight engine is allowed to approach that limit. If the splines were worn through the half shafts would rotate relative to one another and undo the axial load coupling (a threaded component most likely). But the splines thenselves only transmit torsional loads - except that the helical shape may serve to unload the axial coupling in some circumstances.

The QF32 related AD's are not linked to the spline wear AD.

Please look at the Picture of the Aft LP Shaft. Take note of the soot ring. Locate that evidence under the Spline joint. Notice the well defined after portion of this ring of soot. From this well defined border, note that the soot seems to lighten and form a less sharp terminus as it moves forward.

High pressure oil fire gases have exited the Spline Joint at the connection of the two IP Shafts. Entering the space between the IP Shaft and LP Shaft, they encounter P30 pressure and quickly are routed forward. They continue to move forward and out the LP cowling, showing the Sooty streaks on the cowl.

The original AD has not been satisfied Publicly. RR is doing work that is under a device, an umbrella if you will, addressing a Duff Tube, as if it caused everything. No reporting of this "work" is ongoing. This engine has deeply problematic issues that degrade internals such that an inspection cycle of every third landing was proposed and enforced.

These issues? Every impact that can be imagined on Splines that historically have a virtual 100 per cent record. Metal in the Oil? Residue? Failed Oiling Couples? How much of Rolls Royce's authority as to compliance and airworthiness is to be respected (Other than by EASA) when they don't disclose serious issues to their clients? Rolls is doing workarounds under the cover of one tube that may not have been onboard. It is outrageous that EASA leaves to the Manufacturer all authority for airworthiness.

FAA FOIA

That’s hardy a forensic analysis Bearfoil. Since they are based on very nearly zero evidence, your conclusions have to be taken as extremely dubious if not completely mad (I regret to say).

firstfloor & bearfoil,

To see the shaft couplings in perhaps more detail, go to this site and click on the Trent 500 icon. Granted it is a Trent 500 and not a Trent 900, but I bet not much has changed relative to the general layout/architecture.

Portfolio of Technical Illustrations, Aviation Cutaway Drawings (Ghosted Drawings) & Technical Art. Info Graphics from Flightline Arts, UK.

Hope this is helpful.

Yes, that's nice stuff. Thankyou. Simplest illustration to follow is front end of the Honeywell AS907 showing nut and spline. Axial load through the nut, torsion in the spline. Very common and simple type of joint - never goes wrong, apparently.

Well allmost never :)

http://fromtheflightdeck.com/MEL/PPRune/shaft.jpg

TheTrent 500 cutaway is very good, one can clearly see the bearing box and the scavange oil pipe going downward from the box. Here how such a box works in principle:


http://i298.photobucket.com/albums/m...yrinthseal.jpg
Bleed air is used to keep the outer box under pressure, the labyrinth seals has air pushing any oil back into the inner bearing box.

Trent 900 cutaway
 

here is a similar cutaway
http://www.rolls-royce.com/Images/br...cm92-11346.pdf
I like that, shows all the technical delicacy of that desighn.
Jo

Turbine D:

Most helpful. Every pixel of added resolution gets up closer to understanding what is going on.

Are those widgets in the center of the pix on the large piping what we used to see as clamps on automobile radiator hoses?

OE

lomapaseo wrote, quoting firstfloor:

lomapaseo, I have to apologize for thinking that you could not see the problem that splines could present. And a sense of humor as well :)

If I may impose on that sense of humor to partially quote you from back at #1816 back at page 91, on 12th December:

I assume that the pix you posted was not damage due to a can opener being left in the turbine :).

OE

Old Engineer

They sure look some on automobiles I have owned:)

Turbine D

Ferpe - That sump ("bearing box") sketch looks a lot like one I recall seeing in some GE CF6 orientation literature, but I'm sure the Trent is similar in concept. One complication might be bearings on concentric shafts, though.

Yes, the between HP and IP shafts there is a space where the oil can go outwards (red 1 in picture) and follow the IP shaft forwards, there seems to be a labyrinth between the IP and HP shafts at the line from 2 in the picture.

http://i298.photobucket.com/albums/m...bearingbox.jpg

I guess they use the same principle here, a labyrinth sealing with a higher pressure in the space between the HP and IP shaft then the inner bearing area.

Annex14

Thanks for the Trent 900 cutaway. Although it is only a sketch as the Trent 500 cutaway was, it gives some indications of changes that were made, particularly in the area of the IPT rotor and Stage 1 LPT nozzle that has been puzzling me.

First, I am pretty convinced that the fire caused by oil leakage (broken feed line) occurred in this area and was the instigator of the subsequent events leading to catastrophic failure of the components in this area of the engine.

In the Trent 500 engine series, there have been two failures in this same area. One failure has been identified as to coking instigating a fire causing the IPT rotor to fail at the disc rear drive arm 580 bolt holes. The second failure (Qantas B-747 out of SFO) is still under investigation by the ASTB, no cause yet given. However, in both cases the failures were uncontained. In both cases the IPT rotor moved rearward wiping out the Stage 1 LPT nozzle ring and the casing holding the nozzle ring as well as eliminating all the IPT rotor blades. In both cases, the IPT disc did not rupture but there was considerable damage to the LPT nozzles and blades as quite a bit of the debris went through the turbine. So why did the IPT rotor disc not rupture on these engines, but did on the Qantas A-380? All three aircraft were in a climb out mode, with the A-380 at an initial stage, the other two at 20,000+ feet altitude. In making the assumption that all three experienced a fire at or around the IPT rotor, the answer may be some changes made in the Trent 900 engine verses what is present in the Trent 500 engine. In particular, note the gap between the IPT rotor blades and the leading edge of the airfoils of Stage 1 LPT nozzle on the Trent 900. Compare it to the gap present on the Trent 500.

In the design of rotor/stator interfaces in this area of the engine, two important considerations must be addressed, fire and shaft breakage. If the rotor is released due to one or the other, the design should assure that the airfoils come in contact first as a breaking mechanism to prevent disc overspeed and subsequent possibility of a disc burst. The reason for this is a term called "False Bearings". If the disc or a portion thereof contacts first, there is sufficient energy present to melt the contact surfaces and momentarily make a low friction bearing long enough to allow overspeed to occur. This liquid metal bearing is referred to as the "false bearing". Often, the LPT nozzle airfoils have their leading edges bowed outward to assure first contact with the IPT rotor blades should a disconnect occur.

Because of the closeness of these airfoils in the Trent 500, blade to nozzle contact prevented IPT disc overspeed. But in the Trent 900, note the Stage 1 LPT nozzle airfoils are more rearward precluding contact and breaking effect before overspeed occurs.

So my thought is the oil fire and temperatures in the cavities "softened the IPT rotor disc, causing the rear drive arm to fail, the rotor moved aft, but the blades didn't contact the nozzle airfoils first causing a disc overspeed condition and subsequent burst.

Turbine D