Hi Turbine D

Here... As the stress limits begin to be exceeded, the power drive arm fails, releasing the disk and the disk is free to rotate with no control over rotational speed. As the rotational speed increases with no impediment to slow it, it bursts. That, I believe, is what happened on this engine.

I think you are missing something, and it is key to my description of the failure. Using your own description, above, when the Disk is released, it has not the time to increase in rotational velocity. None.

The very instant it is released, it wobbles wildly out of control,, no longer restrained by a symmetric join. It contacts the case, wobbles forward and aftward, and begins immediately to SLOW. It is converting what was controlled and productive rotation into chaotic and catastrophic disintegration of the IPT cave, bursting the case.This is the mechanical Bang. The second bang is the release of P30 gases into the atmosphere, possibly.

Again, once released, it goes non planar and is free to eccentrically smash everything in its path, for a VERY short time, constrained in a Titanium "Barrrel".


There is no overspeed. Out of control rotation, along with the eccentric "orbit" of the disc, shears the blades, and initiates the tri-partite decomposition of the Disc. Failure at the IPT/Drive Arm it was, as reported by Rolls.

What say you, my friend?

barit1

To continue, is it possible that the six seconds of N2 spool drop are connected to the loss of the rigid joint at IPT/Drive ARM? Is the IPT slowing, and converting its energy into a disruption of the contents of the IPT cave? Without the obstacle of the Turbine, is the pressure stage increasing rapidly in pressure, since there is no mechanism to transmit it to the LPT? The HP is still functional (and gains rpm, and 'added' fuel, sent by the ECM, to make up for the loss of the IPT 'barrier').

I don't think it happened that way, the IPT, in circling the case, cannot have stayed for six seconds?

I have given you my thoughts, there is nothing more to add....

TD

I should have asked my question more directly. Once separated from the Power Arm, how can the Disk possibly retain stability sufficient to gain rpm? Won't it start going bs on its cramped environs?

Sorry for the obtuseness, Turbine D.

respect

Wobbling? What is the source of the force required to start it wobbling? Have you ever contemplated the magnitude of force required to accomplish this with a 100kg disc turning 7000 rpm?

Or have you been reading too much Velikowsky?

Lyman,

Do you realize that if you stand in front of a turbofan engine and grasp one of the fan blades with your little finger you can, with very, very little effort at all, turn the fan over, which in turn, turns the LP compressor and the LP turbine? It is a jewel and the same is true with a triple spool engine. So when you think about that, think of what happens in the IP turbine on the Trent 900 when the disk separates from the shaft holding it. We are talking here no more than a couple of seconds. Your talking as if we are in slow motion. The turbine wheel, once the power drive arm fractured, over sped very, very quickly. The thing driving it was the highly hot compressed air that was expanding, but passing through its turbine blades, still attached to the disk. Remember the aircraft was in the climb mode. The rotational inertia for the given mass is unbelievable in this situation, The turbine wheel was driving nothing, a free turbine is a nightmare. Additionally, there was nothing in the way to stop it, it just proceeded to burst and that is what the photos of the found disk fragment depicted.

I would suggest you, once again, look at the engine cross section and note if it moved rearward, there was no substantial material to decrease it rotational speed except for the inner flow path band of the stage 1 LPT nozzle. Whether it wobbled or not is immaterial, in fact it may have gained enough rotational speed while still in place over the IP shaft, but free of the shaft to initiate the speed leading to final failure. Remember, there was little damage to the LP turbine except for the stage 1 LPT nozzle. The majority of debris went outward in a radial direction, not rearward.

Think real time, not slow motion time which is only recorded by high speed cameras.

TD

There is another aspect to this, and that is the rate of heat transfer from the oil fire. Lightweight components heat up very quickly, but massive structures much more slowly. The IP rotor disc is a composite of massive areas (the disc bore) and much thinner areas (the drive arm). Therefore I suspect the drive arm overheated much more quickly than the disc proper.

This reinforces the scenario in which the drive arm failed first (after the oil fire escaped the sump). This released the disc assembly, blades included, and since it was subject to a big gas pressure load on the front face, it was immediately blown aft until it was seated on some static parts, establishing a "false bearing". It's still turning 7000 rpm, and the false bearing is quickly friction-heated producing liquid metal, which has some lubricating properties.

And it's still being driven by core turbine exhaust airflow, and it has lost the normal torque load of the IP compressor.

Guaranteed overspeed, as previously stated.

TD: " The rotational inertia for the given mass is unbelievable in this situation, The turbine wheel was driving nothing, a free turbine is a nightmare. Additionally, there was nothing in the way to stop it, it just proceeded to burst and that is what the photos of the found disk fragment depicted."

That is my point. I think from Edelweiss, and evidence at LPT#1, there was no time for the debris to blow aft. It exited out the opening created by the departing blades, followed immediately by the three main, and hundreds of minor, bits of the disintegrated wheel. If we consider that the wheel, to gain rpm, needed time on shaft (or, 'false bearing'), then debris would have exited back through the LPT proper, making a proper mess. As the wheel was blown instantly back against the Stator vanes platform, the IP blades were shorn, and blew out the case, co-planar with rotational orbit. This is what happened to Edelweiss; in that case, the IPT remained attached (though 'fractured', circumferentially).

Consider, the IPT is not attached to the shaft, per se, and a loss of the drive arm leaves a discrepancy bore diameter/shaft of many centimeters. The bearing is ad hoc, and unable to support rotation. The IPT orbit is instantly eccentric, and the disintegration is likewise instantaneous. It is the relatively undamaged condition of the LP turbine that gives this away; all debris blew out simultaneous loss of drive arm, imho. Similarly, the IPT blades were lost instantly, nothing remained to transmit the gas flow into additional rotation. This is Edelweiss, redux, save IPT loss out the case. Only in Edelweiss, the fracture of the drive arm was not complete, and the wheel was luckily retained.

I know you are convinced of oversped wheel; without blades, and time, I cannot agree. 7000 rpm is well sufficient to blow up the IP system. Is there a conclusion in the report? The initial used the word "may" re: overspeed.

add; [B][Remember, there was little damage to the LP turbine except for the stage 1 LPT nozzle. The majority of debris went outward in a radial direction, not rearward. /B]

Yes, again, my point. Even two seconds would have seen a substantial flow of debris out the tail pipe. Radial exit of virtually all the debris drives my conclusion that the disintegration/exit was instantaneous, not lingering.

For N2 to spool down whilst the IPT was spinning up for seconds, and there is light damage to aft rotational mass, is a reach, imo. What is more likely is a damage trail suggested by the actual AD on this engine. Aft drift of the IP shaft, metal/metal contact, superheated Drive arm, and disintegration, causing aircraft damage and parts on the ground, endangering people below. The AD was written with climb out in mind, hence the reference to population underneath. The 380 serves airports in populated regions, and catastrophic failure puts those below it at risk.

The source of the wear on the rigid coupling is published, and the damage pursuant is also. Drifted shaft, and oil fire are two results of the cause of the AD in the first place. Loss of oil is reported on many a/c, along with oil pressure problems, and preceding incidents.

If the cause of the problem was overspeed, fine; it is not necessary to the explosion, however, and that is my point. I don't see evidence of any kind that isolates this uncontained failure from one predicted by the regulator. For some reason, it has become necessary to propose a new, and unrelated anomaly. Why is that? (Rhetorical). The oil pipe problem was not new, it was not unrelated, and it is not logical to separate the "Oil Fire" from the AD. The wear was caused by vibration of the the Rotating Mass. imho

Thanks

Lyman, I wish you would consider the fact that turbine discs don't just fail at near-normal rpms, absent some significant manufacturing defect. I point to the eight engines intentionally crashed on 9/11. All the turbine discs, from (I believe) three different manufacturers, remained in one piece after the dust had settled. There are plenty of other accidents available for study in which the impact forces failed to break the turbine discs.

Consider also that the IPT on this engine generates (I'm estimating here) 50,000 shaft horsepower, all of which is delivered to the IPC. If the shafting (drive arm in this case) lets go, it's like tromping on the clutch and accelerator at the same time.

No, the only things that I'm aware to cause disc failure are overspeed, overtemperature, or a fault in the disc (either during forging, machining, or some repair process). For you to propose otherwise for QF32 will take some very serious analysis.

Hi barit1, thanks for the reply. I think there is merit to what you say, without doubt, these rotors are phenomenally strong. The design consideration is strictly that to which you allude, resistant to overspeed and overtemp, to a point. We do not need to look past the existing data to find failure. In fact, the failure was foretold by no less than the regulating authority, to which the engine manufacturer is entirely responsible for proof of quality. So I do not feel the need to go beyond, the failure is in front of us.

Let's assume that some combination of above limit rpm and ambient temp caused failure, fair enough?

My purpose from the outset is to frame the evidence within the data provided in the AD as enforced. If some mechanism failed the connection other than high heat, or overspeed, let's agree it is not present in the analysis? I think the metallic spatter is the product of friction stir at the arm/engine frame, as predicted by EASA. Only one quart of oil was missing, the fire was in front of the wheel, and the prediction via EASA tend to override a lingering spin up whilst the core is melting. The Drive Arm, logically was the last failure before disintegration, and the decomposition happened instantly. For the spatter to appear on the aft face of the wheel speaks against the heat migrating from its front? Yet with a shaft transiting aft, the friction can create instant molten metal, whilst the shaft retains integrity. You cite the N2 loss of rpm, I think that was a result of metal/metal from shaft, case, as predicted. I would say the shaft slowed whilst joined to the Turbine, and EEC poured extra fuel in to compensate for the reduction in power applied to the melting metal in the core.

These are informed from the AD, directly. How can N2 unwind for seconds, as a result of separation of IPT? Again, the IPT without the Drive Arm connection is a sizable herd of horses loose in the barrel. The Orbit would have to remain precisely free of obstruction for the Wheel to accelerate. The Drive Arm/IPT resemble a Bell on a bearing at the small end. If the connection is lost, the mass is not only free to wander, chaotically, but the rim of the Bell is quite distant from the bearing attach, any imbalance would instantaneously throw the entire asembly into an extreme eccentric.

I am not arguing for its own sake, but am trying to fit the results into the predictive regulation published by the engineers before the fact.

See you in JB?

take care, and thanks again, I appreciate your patience with my stubborn streak.

Kucing,

You sent me a private message, but unfortunately the website won't let me reply to you. If you send me another private message with your email address I will try to answer your queries.

Lyman,

Lets go through some facts again.

From the latest ATSB Interim-factual Report:

"The investigation has found that the intermediate pressure (IP) turbine disc failed as a result of an overspeed condition, liberating sections of the IP turbine disc that then penetrated the engine case and wing structure. The disc failure was initiated by a manufacturing defect in an oil feed pipe that resulted in a wall thickness reduction in an area that is machined to receive a coarse filter. That section of the oil feed pipe sustained a fatigue crack during engine operations that lead to an internal engine oil fire that weakened the IP turbine disc. In turn, a circumferential fracture was induced around the disc, allowing it to separate from the IP turbine shaft. The unrestrained disc accelerated to critical burst speed. This lead to the No 2 engine failure and subsequent significant penetration damage to the airframe structure and systems."

Your quotes directed to barit1:

The way a three spool turbine engine works is this:

1. The low pressure turbine (LPT) drives the fan and low pressure compressor.
2. The intermediate turbine (IP) drives the intermediate compressor.
3. The high pressure turbine (HPT) drives the high pressure compressor.

So, if the IP turbine disconnects from the shaft resulting from a circumferential fracture of the power drive arm, it is no longer driving the intermediate compressor. Therefore, N2, that is measured at the intermediate compressor begins to decrease, in this case from 94.5% to 93.2% at UTC 0201:00. A 100% speed for the IP turbine rotor means it was rotating at 8,300 RPMs. At 94.5% speed it is rotating at 7844 RPMs. Burst speed would be approximately 10,375 RPMs, assuming a 25% margin above maximum operating RPMs. It didn't have to speed up (overspeed) very far to get to the burst point. The rate of speed up would be determined by the expanding air passing through the IP turbine blades received from the HPT blades. It was probably less than 7 seconds from normal operation to burst.

No it doesn't. It only resembles that way on the disc after the failure. Prior to failure, it is a bent downward plate (towards the engine centerline), parallel to the disc face, a circumferential feature that attaches to the end of the shaft by means of a series of bolts that pass through bolt holes in the power drive arm.

The mass you suggest, the IP turbine rotor disc and blades attached is free to very rapidly accelerate until such time the disc bore stretches which is indicative of pending failure and then moves back. At that point, IMO, the only contact would be near the base of the turbine blades embedded in the disc slots contacting the stage 1 LPT nozzle. That rubbing would create metal splatter observed on the aft surfaces of the disc, similar to what happens during an inertia welding process, an initial false bearing without adequate pressure to slow the rotation of the disc. It was not chaotic at all until the burst occurred. Again, IMO, eccentric motion had little effect on what was about to happen.

about as chaotic as a freely spinning bicycle wheel being dropped from a window ledge or ladder

Try it some time and film it before it hits the ground

I suggest that once the circumferential fracture separated the wheel from the power arm, the disc was already exceeding a (somewhat lower) burst speed.

Are we looking at the same schematic? The turbine disc has "return" (sleeve) feature that is the drive arm. The circumferential fracture separated the wheel from the arm, therefore the wheel is no longer restrained in a circular orbit. Show me where the wheel has any interest whatsoever in accelerating? The massive pressure of the exiting HPC gases thrusts the wheel instantly into the LPT Stator. The turbine blades are sheared, and the wheel performs a very rapid eccentric dance out the case, post disintegration. Attached to nothing, how do you say the wheel remains stable to resist the flow of gas, and thus accel? It has no axle. No stability.

Lomapaseo. A rapidly spinning bicycle wheel illustrates exactly my point, its gyro expression of high energy slams it into the stationery parts of the core.

I still suggest that the slowing of the IPC demonstrates the axial migration of the IShaft aft ward into the "stationery parts of the engine" (AD). Integrity of the shaft is required until disintegration of the I Turbine, IMO. I suggest that this slowing of N2 (evidence from phonic wheel, and diminished pressure), is what precipitated the introduction of extra fuel into the burners, and caused N3 to increase, perhaps even introducing unburned Fuel into the IP cavity. That can't be a good thing...

How did one quart of oil incinerate this engine?

Sorry, but in the T900 series, the space downstream of the IPT is vacant for quite some distance; The LPT stator (if any **) is well downstream. The IPT blades remained largely intact, providing the "unbalanced torque" (i.e. freewheeling, unloaded) to accelerate the disc to burst speed.

NOTE ** One aerodynamic improvement in the Trent engine is the use of contra-rotating spools. I do not recall positively which ones rotate opposite to others, but I think the IP and LP are opposite. IF this is the case, R-R had the opportunity to ELIMINATE the first stage LPT stator, because the swirl flow upon leaving the IPT accomplishes the same thing. Less weight, less pressure drop, better efficiency, but also the increased opportunity for an IPT overspeed per the present discussion.

Lyman,

Your quote:
Again, from the ATSB Interim Factual Report

Quite honestly, I never saw this report until yesterday. It describes what I have thought and tried to express to you all along. However, you believe differently with no evidence to support your theories. So it remains a standoff, your theories verses factual information and some turbine engine experiences on my part. No sense in continuation of this discussion.

Simple, it started a fire in a very sensitive area of the engine that generated heat on top of what was already present, which when combined, exceeded the disc material capability for the temperature experienced.

Hi barit1,

On the Trent 900 engine, the LP & IP rotors turn counter clockwise with the HP rotor turning clockwise. This enables elimination of the stationary nozzle between the HPT blade wheel and the IPT blade wheel. The only thing between the two is a sheet metal frame structure that contained the failed oil feed pipe that started the fire. The stage 1 LPT nozzle is still there, but has a significantly larger diameter due to the conical nature of the LPT.

Hi barit1 I think it is HPC that spins contra. The HP shaft is nested inside the well of the IP shaft and the bearings for each have a ~20000 rpm differential due rotation opposition.

TD: From the schematic, the bearings for the aft terminus of the IP shaft are on the shaft side, the Power Arm is outboard of the shaft's end, as it returns foreward, surrounding the bearings. Thus any separation of the Power Arm from the wheel eliminates the cvarriage for the Turbine Disc, via separation, The Disc has no structure left, it is not as if the wheel is inboard, and merely spins on, without bearings.

I have read the report, and it can be read in at least two ways, leaving unclear some of the evidence/conclusion chain of thought.

I have spent hours reviewing the schematic, and to me, it is simple. I might be missing some element, or the drawing may be incomplete or inaccurate. The mate for the Power Arm, Wheel, Shaft, and bearing is in the three lam sandwich at the most aft end of the shaft. If these retaining bolts sheared, the entire Arm/assembly is lost, there is no possibility for any retention of the Wheel's mass for any length of time.

Could you post the Drawing? I have limited computer skill, and perhaps if addressing the same drawing, we could meet in our thoughts? I don't doubt either of you gents, and am in deep water with you two, my mechanical skill is better than my theoretical, and I stand to be corrected, always....

Lyman,

The innermost shaft is the LP, the outermost is the HP and IP shaft is between the HP & LP. The bearing supporting the HP rotors are separate (independent) from the IP rotors.

The IP disc accelerated to burst while still over at least part of the two rear roller bearings.

I don't see how you can read the ATSB Report two ways...

Lyman & Turbine D:Thank you for reconnecting my aging neurons! My argument stands amended.

Lyman:I assure you that the intact disc (minus the drive arm flange) is only unsupported for a few milliseconds, as the gas pressure drives it aft to a false bearing surface in the LPT inner duct. Now it is once again supported against a fairly low-friction surface, akin to the thrust bearing on your car's crankshaft against which your clutch reacts.

Without a significant torque load (normally the IPC, which is now un-driven and winding down), the IPT has no chance to do anything other than spin up to destruction. Over the decades, I have seen the results of like turbine disconnects; the sequence is always fairly similar, and the results never pretty.

You seem to keep referring to the IP main shaft, but is there any mention of a pertinent shaft defect in this accident report?

You seem to keep referring to the IP main shaft, but is there any mention of a pertinent shaft defect in this accident report?....

************************************************************ ***

No, but I am not sure of the direction you want to go. The Shaft did not fail, at least not prior to disintegration. If you look at the section through this space, you see that the Turbine Plane is foreward of the aft bearing (shaft). The attach to the shaft is aft of the bearing plane. So the wheel is connected via a bearing that is remote from its geometrical plane involving a 180 degree return to the Drive Arm. I assume the circumferential fracture is located at the seam between wheel and ARM. The bore of the Wheel at that point is the same diameter as the ARM, does the Wheel slip over the ARM, as it moves "AFT"? Nice trick! Or does it slam into the Stator platform (LPT) which provides some false bearing support? OR, is the circumferential fracture at the aft terminus of the ARM, where it flats to enter the three way join? If the true bearing architecture gives way to a new and "false" support, where is it located? On the Stator? This surface is of course irregular, ad hoc, and incapable of maintaining tolerance that prevent the Wheel from going wildly eccentric, here's why: The bore of the IPT is much larger than the diameter of the IShaft, of course. Unlike a standard slip fit or "Press Join" to an axle (with a keyway, or splines) and without restraint, the wheel cannot maintain any local integrity....

I have searched for the proper schematic, and the usual places are not available, so I do hope I am not remembering the T5, or 7.... So my words are memory based. Bottom line? Can you help me with a better description of this "false bearing". Also, how do the blades remain attached , my opinion is they exited forward as they were pushed out the pine trees by the stator?

While i have seen an (alleged) cross-section of the T900, I cannot seem to locate it just now. Maybe someone can post a link so I can point out the pertinent features.

But the essential point I'm making is that the uncoupled disc assembly will establish a false bearing INBOARD of the IPT airfoils, so they will remain in place as the disc assembly is driven overspeed.

post 1415 previous thread, looking

Lyman,

I only mention the IP shaft because you think it move rearward, somehow separating from the coupling in the compressor section. It did not!

Your quotes:
And,

The IP disc is a smooth bore disc. It fits snuggly around the end of the IP shaft and is held in place by a series of bolts through bolt holes in the power drive arm. If you review the photos in the ATSB reports, the initial failure occurred at the circumferential bolt holes releasing the disc to overspeed. From a design point of view, it is never good to have bolt holes in the area they are in because of the fact bolt holes are built in stress risers. So this was the weakest point once the disc overheated due to the oil fire.

At some point in time, seconds after the power drive arm failure, the disc, still containing the turbine blades, moved rearward, contacting the inner band leading edge of the LPT stage 1 nozzle ring. There was no IP blade to LP nozzle contact to slow the accelerated disc. At this point in time, the disc was highly accelerated, near burst and had stretched so that the "pine tree" features of the disc were enlarged, the turbine blades were now loose. The contact with the LPT nozzle created the metal splatter and then the disc burst. The turbine blades may have come out of the several disc fractured sections in two directions, radially and rearward as the disc departed the engine.

I agree with barit1, this is a classic disc burst, have seen it happen in a test cell, not a pretty sight.

barit1 & Lyman,

Here is the Trent 900 engine cross-section for your information. Also, one thing I forgot to convey, the disc burst speed I gave was for a normal operating temperatures. With the oil fire, the disc burst speed would be perhaps significantly less...

http://i1166.photobucket.com/albums/...t900u6cu-1.jpg

TD

Hi gents. I do not misunderstand your conclusions. Matter of fact, it is very difficult to disagree with them. The Problem.

The drawing is not the one I had in my other computer, which is presently in maintenance. I'd like to condense the crux of my disagreement around the Shaft/Arm/Wheel system.

I caught flack in the first thread for stating that the Wheel and Arm were made up of three parts. I was unclear. Let me start by saying that in this assembly I see three elements. The rim, and blades, the face, or disc, and the Arm, or Drive.

The first design concept I see is the "wraparound" feature of the Arm. Attached outside the bearing load, it performs a 180 to wrap the Shaft as it extends forward to capture the Disc. Is this common? I am familiar only with engines that include rotors inside the "Load frame" iow, attached directly to the shaft. As it is, the Wheel drives a load that is outside the bearing plane "The Attach" So I see potential problems with balance, and resonance.

As barit1 says, and TurbineD agrees, these rotors are of immense strength. Now that is subjective, and since failure is obvious, here, I refer to my earlier comment that due fire, the designed fail (rpm) was reduced, perhaps "substantially" as Turbine D points out.

My intention is to direct the comments (if you agree) toward the method of attachment of the Arm/Wheel, and if you are familiar with any potential pitfalls?

Because although TurbineD disagrees with my "Bell" shaped comment, I think we can see that the Arm and disc create a shape that is similar to bell, mushroom, umbrella, etc. and that this shape must resist a somewhat more complex strain profile than the Wheel/Axle tradition.

Your participation is optional, of course, but I am serious, and believe in addressing some of these concerns; at least one person will gain some knowledge, here. And tht would be me....

ad. For instance, as I say, the load is outside the bearing loadframe. Standard would be within at least two bearings, rather than attached at the end of a shaft supported only one side of the load. A corollary might be "whirl mode", where a great deal of mass and energy is "extended" past a point of support, such that the bearing is stressed in ways that are common to rapidly spinning systems, but ordinarily are suspended between two mounting points....

Please note that according to the drawing, the wheel contacts the LPT platform/stator with its blades prior to the more core oriented partition. So the false bearing would have to "wait" for the wheel to arrive post stator contact?

Isn't this IPT blades contact a design feature, to prevent OverSpeed? Scrub off the blades to prevent a drive surface for the gas path? If so, what failed, in this case?

Is the bearing mount portion of the joint connected to the race? So, can the outer race support the wheel, post separation from Arm? I would offer yes, but there would be a lot of play between the race and the inner surface/drive arm?

Note for comparison the architecture of the HP Rotor, supported on one face by the HP cap bearing, and on the other face by the shaft proper? Here, the loads are functionally within the two bearing planes, and the one sided torsional affect is snubbed?

Similarly, although the Fan is supported outside the bearing run, it has two bearing planes, which provide a 'moment arm' to capture the oddities this disc encounters with the atmosphere, and FOD? Connected as it is to the IPT terminus, and the radius of the bearing system it suggests a very robust carriage for the bulk of the a/c thrust.....

TD, barit1, what would be your opinion on the Drive Arm subject to these torsional anomalies whilst under the climb load of the gas path? Would undue fatigue weaken, and/or heat the system? Mind, the HP bearing and the IP bearing rotate opposite each other, at times with a differential of 20000 rpm.

First off, my expertise (such as it is) does not include rotor dynamic vibration, but I think having the plane of the IPT disc directly over the aft IP bearing (i.e. zero overhung moment) might have some advantage re whirl mode. But I have never seen this arrangement employed before.

Second, I wonder just how accurate the flowpath detail in Turbine D's cross-section is, especially in the IPT > LPT area. I could see where the inner band of the first LPT nozzle stage could contact the IPT blades at the TE root, perhaps initiating blade failure. This would seem to be in conflict with the failure sequence in the powerplant report. I think Lyman is in agreement here.

I'm puzzled by the words (my bold) in the quote above.

Do you have a link to the section of the report that you refer?

Typically it is only in a final report which may contain an analyis where critical details are explained and supported.

My read so far is that a local (oil fire) overheat occurred sufficient to release the IPT disk from the compressor load at the drive arm.

This overheat may include only the innermost portions of the turbine disk.

Does the referenced report detail any estimates of disk body material strength from Bore to rim of the disk?.

If not then the strength of the disk may have been only marginally impacted by the oil fire.

Regardless of the answer to my question, the released disk from its drive arm is an important detail in its progression. If it was only completely released in the tangential direction (torque loss only) then it may not have permitted the free disk from moving aft into the aft nozzle vane clusters.

Whether the disk moves aft or not, it would still free wheel up in speed at a very rapid rate while remaining centered about the engine centerline. The tendancy is for it to seek restraint both upon the remaining drive arm/shaft pieces as well as its still intact blade tips against the outer case structure.

The time between the separation from its drive arm and burst would be dependant on several factors

1) the average material strength and average temperature in the compartment (local strength losses due to a localized fire would not be significant as the disk stretches and redistributes its strain)

2) The residual pressure trace across the turbine blades vs time

3) any loss of turbine blades in this stage

Rubbing friction itself against the aft nozzle vanes and between the blade tips and case would be negligible in such an event.

Likewise any false bearing between the blade roots and the aft nozzles would be negligible.

I haven't seen any factual information that suggests that blade to vane contact would be expected in the airfolis for this design.

In the end it's a F=MA over time that assess how far aft the rotor might have moved vs the speed vs time of the free wheeling disk under a decreasing gas load.

If any of the above has been convered already in a report I would be pleased to read it.

Meanwhile I am quite happy with the recommendations provided by this investigation todate (Fault analysis vs quailty control at the manufacturers)

Lomapaseo,

@ TurbineD...."I assure you that the intact disc (minus the drive arm flange) is only unsupported for a few milliseconds, as the gas pressure drives it aft to a false bearing surface in the LPT inner duct. Now it is once again supported against a fairly low-friction surface, akin to the thrust bearing on your car's crankshaft against which your clutch reacts.

I think TurbineD is correct here, and from the statement, aft migration puts the Turbine's airfoils into the LPT Stator ring, where they are ejected from their slots. This eliminates the ability of the gas path to turn the IPT. The amount of aft migration is failure dependent; but if the circumferential fracture happens about the bolt flange, as he proposed, their is nothing to prevent the drift from continuing well past the Stator into the inner web of the partition.

The HP is receiving fuel in excess of the rpm ratio to flow, (via EEC), I think raw fuel may have entered the cavity to exacerbate the fire post separation, making matters worse. In any event, where do you put the fracture?

@lomapaseo.....Whether the disk moves aft or not, it would still free wheel up in speed at a very rapid rate while remaining centered about the engine centerline. The tendency is for it to seek restraint both upon the remaining drive arm/shaft pieces as well as its still intact blade tips against the outer case structure.

Noting the room between the bearing box and the drive arm, I would disagree that the Turbine remains centered about its shaft, (the engine's centerline). Gas path pressure and gyroscopic forces work against this type of ad hoc stability, IMHO. So I picture the bore portion as contacting the bearing box in a random and chaotic fashion, not conducive to the Turbine's retention, and destructive of its ability to remain in the engine. This impacts the blades against the platform, and they are lost, immediately.

In any case, do you not agree, that from the drawing, the LPT Stator platform makes first contact as the IPT drifts aft?

From the drawing, I note a build up of the LPT vanes platform at the area that makes contact first with IPT should it drift aftward. The portion of the IPT that makes contact with this strengthened area is the blade root circle. This in anticipation of wheel release, the design is meant to remove the airfoils in just such an emergency as this explosive burst.

As the design consideration, it demonstrates an improvement in concept, an anticipated mitigation. I suggest that it worked; as bad as this incident appears, if IPT retains its blades any longer than apparent here, the a/ c may have been lost.

Lyman,

Your quote: No, No, No, that is what you don't want to have happen. What you want to have happen is the turbine blades to stay with the disc which is holding them so that if the disc becomes liberated and if and when it moves back, those turbine blades clash with the turbine nozzle airfoils behind slowing the disc acceleration before it gets to a burst situation. What would then happen would be a bunch of ground up airfoil pieces exiting through the LPT and out the exhaust pipe, but the disc would slow down and remain intact. That is what happened on several older model Trent engines where blade to nozzle contacted prevented disc bursts.

In the case of the Trent 900, when the disc became free, overspeed initiated and when it move back, the contact with the stage 1 LPT nozzle occurred at the disc rim rubbing out the turbine blade retainers thereby creating the potential for the blades to come out and making no airfoil to airfoil contact slowing the disc speed. Remember, there was little damage to the LPT other than the stage 1 LPT nozzle with most of the remnants, disc sections included, exiting radially, not good. I will look again at the photos in the ATSB reports to be sure of this scenario as I am doing it at the moment from recall. In the meantime, look at the GP7200 engine cutaway below.

You will note that if the stage 2 HPT disc (similar position to the IPT disc) became liberated for one reason or another, if it would move back, the turbine blades would clash with the stage 1 LPT nozzle slowing the disc speed. It is an important design feature to inhibit runaway disc speed.

lomapaseo,

All I can say about the intensity of the fire from the burning oil is what the ATSB said in their Interim-factual report:

I do think there is some merit in what you suggest, the disc stayed centered for a short period of time during the acceleration.






http://i1166.photobucket.com/albums/...72/mc03_lg.png

I take it then that the airfoils are to be sacrificial? I do not see how the IPT blades can clash with the LPT vanes without transiting the Stator "Platform".

They cannot get there from there. They have to slice through 10 centimeters of platform to access the Vanes. Now that may slow the Wheel, but the blades are lost anyway. If it is as you say, why provide an impediment to instantaneous clash post separation at the ARM? It is not sensible.

I was sure I asked more questions, and I'll stand by the scrub foils for now. May I choose one question to ask? How about the open space between the Bearing box and the bore of the IPT? Sufficient to restrain the IPT and offer a bearing for overspeed? Without allowing aft drift, or Disc eccentrics, wobble?

I prefer the TRENT7 pic to the more pedestrian T9drawing......Are they exactly similar? I see that they are not, the foils are pitched the same way. No contra.

TD......"You will note that if the stage 2 HPT disc (similar position to the IPT disc) became liberated for one reason or another, if it would move back, the turbine blades would clash with the stage 1 LPT nozzle slowing the disc speed. It is an important design feature to inhibit runaway disc speed."

Come again? I just said that, re: TRENT 900? How does HPT climb through the IPT to get at LPTNozzle? Blade clash is a bad thing for the 9, but ok for the 7?

Lyman,
Your quote: The (GE/Pratt) Engine Alliance GP7200 is a 2 spool engine verses the Trent 900 that is a 3 spool engine. By showing you a cutaway of the GP7200 engine, I wanted you to note only this:

If the stage 2 HPT blade rotor, for whatever reason, moves rearward, disconnected from the shaft, the turbine blades would clash with the stage 1 LPT nozzle airfoils, slowing the disc rotational speed down below that of burst overspeed.

In the Trent 900, if the IP turbine rotor, for whatever reason, moves rearward, disconnected from the shaft, the turbine blades cannot clash with the stage 1 LPT nozzle airfoils to slow the disc rotational speed down below that of burst overspeed.

The difference between the two engines lies in each of their basic designs. The Trent 900 engine is also different from previous (older) Trent designs where blade to nozzle clashing is possible and did occur in several failure instances. So, the Trent 900 is unique in its apparent design and positional relationship between the IP turbine rotor and the LPT stage 1 stator (nozzle). I am not going to speculate as to why the Trent 900 design was developed the way it is.

I don't know how to make this any clearer to you, take your time to understand what is presented, I know it is somewhat complex...

So let me ask you then re: your response. Does the GP have an IPT turbine that is affixed with the same architecture as the T9? Because if it is fit on to the shaft via keyway or splines, the overspeed would happen , should the key(s) shear, or the splines grind smooth. It is the false bearing that interests me.

So one question then. Can you show me the false bearing locus? Because any aft movement at all of the IPT in the TRENT puts the aerofoil ring into the LPT Stator. Isn't that accomplishing the same thing as the GE in wheel loss? Design aside, I see no room for false bearing to establish, without IPT/LP1Nozzle conflict. At 500 pounds, the IPT would take out anything in its way, and I do not see how you propose the Bearing box to contain it to capture gas path, and resultant spin up? The relative bore/diameter of this 'new' bearing/system does not look like it is up to keeping the Wheel from axial drift, and/or blade clash with LPT Stator.

I do not see the LPT vanes at all in the GE/PRATT.

Lyman

Your quote: The GP7200 engine does not have an IPT rotor, it is a two spool engine and the HPT consists of two rotors, Stage 1 and Stage 2.

Your quote: Perhaps this view of the GP7200 will enable you to see the Stage 1 LPT nozzle in relationship to the Stage 2 HPT rotor.

http://i1166.photobucket.com/albums/...taway_high.jpg

Your quote: I don't think it does on the Trent because of the angular lean of the Stage 1 LPT vane, refer back to the Trent cross section. The impingement between the two on the Trent is close to the blade root and disc pine tree holding the blades and the inner band extension of the Stage 1 LPT nozzle ring. If you look at the recovered parts photographed in the ATSB report, you can note 2 items of interest. On the recovered LPT vane segment, the inner band is gone. Also, the turbine blade retainers on the recovered portion of the IPT disc are all missing. So you have relatively one smooth surface contacting another relatively smooth surface. Beyond, these observations, nothing more can be said with any certainty. Even with all the recovered parts, lab analysis, etc., plus the assistance of the engine designer and builder, it is difficult to reconstruct the minuscule details of exactly what happened in the few seconds following the release of the IPT rotor disc and blades to the degree you are attempting to without the complete data the ATSB possesses and the knowledge RR has on this engine.

TurbineD

As always, thanks for your patience and expertise. A quote from you....

You will note that if the stage 2 HPT disc (similar position to the IPT disc) became liberated for one reason or another, if it would move back, the turbine blades would clash with the stage 1 LPT nozzle slowing the disc speed. It is an important design feature to inhibit runaway disc speed.

In either engine, the Power Turbine, (IPT in the TRENT, and HPT2 in G/P) send the gas path into the (stationery) nozzle of the engine's LP (turbine) cavity. In the G/P, the vanes are elongated and occupy the plane directly behind the blades of the Power Wheel. The Trent's Vanes appear shorter in length, and are arranged at some distance aft of the preceding rim of the STATOR. This means, as you say, that blade clash will not occur instantly, and the effacement in aft drift occurs with two smooth surfaces, rather than Blade/Blade.

I am speculating as to design. To me, it (TRENT9) represents a sequential approach to Disc runaway. The Wheel is slowed, and borne temporarily at Blade roots/Stator platform, (your "On the recovered LPT vane segment, the inner band is gone. Also, the turbine blade retainers on the recovered portion of the IPT disc are all missing.") So you have relatively one smooth surface contacting another relatively smooth surface. This wipes the retainers, and the blades loosen in the IPT. As the blades begin to depart the wheel, the IPT captures less of the power of the gas path, and the blades shred, to fill the cavity with shrapnel. This shrapnel, I see as a benefit to further loss of blades and resultant loss of rpm. As the IPT travels further back, the Platform degrades into more robust areas of the Platform, and eventually the Blades/Vanes efface.

The IPT, off arm, does not have a solid support at the bearing box, though it does have concentric structures post fracture. This allows an eccentric braking action, though the vibration and noise must be extreme.

At the last, the IPT foils (if any are left) scrub the Vanes of the LPT nozzle, removing the last of them.

As you point out above, blade/blade is the G/P approach to arrestment, RR similar, but with a primary contact at blade roots, Stator platform. Again, I believe that the desire is to slow the wheel, but I would add that equally important is to defeat the gas path mechanically, a fuel cut is not possible in this time sequence.

So I can only say that the approach is different in the 900, but to me represents a step ahead of that taken by GE/PRATT.

I do not discount the overspeed, but I note that you agree the design limit for separation of the wheel into three parts may have happened at lower than maximum rpm for integrity. Yes?

Thanks again

Turbine D
 

Turbine D is your man here. He should be working for the ATSB.:ok:

Lets not forget that this was a T972 on very heavy work cycles esp out of LAX ro Oz. RR made a severe error in an oil pipe bore. Qantas are persuing them in court.

This engine was not a T970 or a T1000.

Turbine D is one of the very few people who had the quals and tactile experience to assess this engine failure.

And as for No. 1 engine running on a for a full hour .. what can you say..God was with them.