QANTAS A380 Uncontained failure.

[Turbine D]

"Wobble"

I've looked at the recent posts and informative sites regarding vibrations and what can result. I think Bear gives some good information regarding support structures and what may be going on in these newer engines.

First, if you think about the fan and all related rotating airfoils attached to one end of a cylinder, and the LPT rotor attached to the other end, 7 feet away, you have a dumb bell in a static sense. If you pick up the dumb bell at one point near the center, both ends will sag downward, the LPT end more so as it will generally be heavier. So then, to prevent this from happening in reality, a series of static structures are designed and placed to provide the backbone of the engine. These structures must maintain alignment between rotors and stators and their design is dictated by the need for stiffness, more so than strength. The major sources of load for an engine translates into forces on the structures such as maneuver loads (turning), CAT (clear air turbulence), acceleration/deceleration, gyroscopic forces, moments through the bearings into the frame hubs that support the bearings, unbalance in the rotors, internal variable pressures in the engine flowpath and lastly, thermal differences induced by the airfoils compressing the air and then the hot gasses flowing from the combustor through the turbine components. There is also the dynamic behavior of the engine and interactions with the airframe to which it is affixed. All of these features must be combined with design features and analysis to enable the static structure components to handle different types of loads.

For the designer, there are all kinds of analysis programs to assist in determining the adequacy of the structure design, redundant structural model, finite analysis model and a complete engine structural model and a first engine bending model. In addition,designs are verified by static load testing that can include built-in defects to test design robustness. Extreme engine testing can test for structural capability under extreme vibration. Low cycle and high cycle fatigue capabilities must be examined. Based on all of this, limit conditions can be determined and set so that a well designed structure will not show any sign of deformation or contribute to a loss of performance even if the limit is exceeded in a one time event. So as you can see, the engine structures are key to having successful engine capable of meeting the mission advertised to the customers.

Good structures must be capable of absorbing vibrations and not amplifying them and there are many techniques that can be incorporated to accomplish this. I used the word "robust". Care must be taken to not reduce weight of structures that could compromise long term structural integrity although this is often looked at during weight reduction/saving campaigns to improve performance.

Vibrations and rotor unbalances are always present in an engine and they change as the total engine cycles grow over time. So the engine backbone must be capable of handling all of these variables as they occur.

[WojtekSz]

Turbine D:
thanks - it is a pure pleasure to read such posts

Good structures must be capable of absorbing vibrations and not amplifying them and there are many techniques that can be incorporated to accomplish this. I used the word "robust". Care must be taken to not reduce weight of structures that could compromise long term structural integrity although this is often looked at during weight reduction/saving campaigns to improve performance.

could you pls tell a little more on techniques to absorb vibrations? The lower the weight and stiffer the shape the higher is the supporting structure natural frequency but doest is solve the vibration problem?
Looks like multiplatform optimization (technical characteristics vs unit costs vs weight vs project time etc)

[bearfoil]

Turbine D

On the money. A good argument that the 900 is not in the same family as the prior iterations of TRENT would be that the destructive issue at the heart of the dilemma does not affect the others. In using this, RR suggests that the 900 is in fact substantially different from the others.

The problem is distilled as usual into likely courses of action. The question is this..... Will for once the authorities demand a revocation, or will they succumb to pressure to keep the troubled engine in service, although it has obvious service life issues? These issues cause an accelerated loss of dependability, such that mere inspections beg the question!!

Have you searched for the specs on the "C" Modification? Given the reported limits on service, the obvious conclusion is that this powerplant is a "work in Progress". Odd that no one will address the issue, but merely defend the indefensible. I only have the "B1" manual, so my limits are strict. I do know from sources that virtually no oil issues were found in the eighteen months prior to the Burst. I also know from sources what the precise issue is. If so, the fix has some pretty lofty obstacles to a simple cure.

DERG

The seventy cycle limit was required by EASA, not "reported" by RR. No doubt there was a consult, there always is. The consult in August produced the relaxed (lax) inspection cycle amendment in September to allow Spline Wear per se to be "averaged" or normed, using each pair as a contributor instead of the "worst worn" disqualification. It was this very relaxation that allowed the Burst to happen, clearly the wear had proceeded such that certain "pairs" were not even functioning.

[Turbine D]

could you pls tell a little more on techniques to absorb vibrations?

It is not an easy question to answer as each engine is different and solutions are different. Some examples might be the type and design of seals used in the bearing box area (seals can be good dampeners), perhaps the number of struts used or even their orientation relative to engine centerline could be another solution. But the finite element analysis program gives good indications as to which technique or techniques might be most advantageous to use if a problem is apparent. One other thing, you can raise the vibrational frequency of a frame without reducing weight by use of a stiffener ring on the OD of the frame so the frame is not resonant with most vibrational modes in the operating range of the engine. A separate item that affects vibrations is the distance separating the engine mounts that attach to the pylon. The further they are apart, the vibrations become larger should an unbalance condition exist in the rotor system of the engine. All these things must be considered in the frame structure design and analysis.

[barit1]

Vibration control

One technique I have seen is to de-tune the bearing mounting by a combination of viscous damping with spring centering. The viscous part can be accomplished by a squeeze-film clearance between the outer race and static frame hub, pressurized by the oil supply.

Similar in principle to the suspension in your automobile or aircraft, but tuned to a much higher frequency.

[DERG]

The Cost of Safety for The Public

On page 16 of http://www.eng.ox.ac.uk/samp/pubs/clifton_transfer.pdf

A subjective judgement is made by the author: this takes a little while to sink in...please persist

"Applications in which a false negative classification carries a higher cost than a false positive classification" ( THEY MEAN CANCER CELLS IN PEOPLE. This means the medical doctors will persue you in court if the get it wrong)

"Conversely, applications in which a false positive classification carries a higher cost than a false negative" (THEY MEAN AIRCRAFT ENGINES..this means that unless the relatives of 469 people bring a class action in court, life and limb is worth less than an engine repair...OR...the insurance will cover it.)

Here it is in their words:

"Systems for the identification of abnormal operating conditions in gas-turbine engines are of this type: the cost associated with a false positive

classication is high, often involving significant examination of equipment and maintenance processes."

The only way anyone sane would consider this is with the knowledge that the A388 had FOUR engines that would not fail all at once: and uncontained engines failures are rare statistically

Now I just wonder if, after realising the Qantas A388 had 469 lives at risk, they suddenly realised that the costs would be higher than they imagined.
Then again it did have FOUR engines.

No.. this is unacceptable. They set out a scheme that had a priority of financial risk reduction without realising they had 469 lives at risk as opposed to the ONE life they have in assessment of cancer patients.

This would be a very good "moral" assignment for our pals in management.
A good one for the lawyers too. The only way this will be solved will be in court.

This academic team should stay with medical statistics. They themselves are a risk to society if they use their "academic clout" to provide a lever for ill advised and naiive commercial managers.

[WojtekSz]

Turbine D, Barit1
thanks for good examples - it really is a case by case operation then like fine tunung relative natural frequencies so they would not happen together. Does this tuning also include the frequencies induces but burning chambers? These seem to be a source of (possibly the most) significant energy vibrations within the engine case.

Derg, Bearfoil
to me the way chosen by EASA is to FORCE the RR to act faster, even if by now they have realized that the class action would cost them a fortune and lost contracts, is asking for painfully frequent maintenance. This way RR has a chance to improve the solution (especially that the exact reason is known) while still statistically safe and without loosing the face.
After all changes to internal engine structures are difficult to prepare, make and implement fleet wide. New seals would not be able to absorb the energy emitted during TO/GA

The solution - advice to public to stop flying A380 with RR engines for a while - the question is how public can be sure the replacements program have been completed?

Excellent case of raison d'etat...

[DERG]

The only hope we have is the ATSB. They are the only ones that do not have a political or economic interest to consider. I guess the only thing they could do is bannish the engine altogther from use on Australian registered aircraft. I dare say off the record a few would agree with that.

There is no question that this mistake has cost RR a lot of money. They will be very carefully attending to these T900s. If another one goes bang they basically have lost all reputation. I would ride on the A388 with the RR Trent because I know that it is safe....well the probabilities are that it is safe..cannot see another novel event happening.

There is a saying about accidents, very much used on this thread; "when all the holes in the cheese line up" Yeah....we had a new engine design using an inexperienced supplier (FAG) and a duff bunch of academics advising the usual bag of dull crayons in management.

It really has been the way NOT to do things. RR made every mistake in the book. As always the general public will be mostly unaware of the background hustle. The deals will be made out of court and business will go on as usual. As Bear says:

"Will for once the authorities demand a revocation, or will they succumb to pressure to keep the troubled engine in service?"

GE will be watching all this in total disbelief. Of course the Schaeffer/FAG bearing manufacturer has opened a plant in Canada specifically to supply GE with bearings for their aerospace gas turbines. I don't think the USA or Canada will be as trusting as RR was.

The unbelievable part to me in ALL this is how RR were led to believe that the telemetry would give them " one and a half hour warning" of a serious engine problem. Just how a bunch of spophistic academics can lead a company like RR to the depths of total incompetence is truly amazing.

[forget]

DERG

I am a civil engineer ...... I learned about turbines as a result of this accident.

(In 12 weeks!)

DERG

It really has been the way NOT to do things. RR made every mistake in the book.

Am I the only one who finds these claims a little incompatible.

[DERG]

That's why they hire us to head up projects, we take no prisoners. We can add up too.

[descol]

It has probably been mentioned somwhre - ... can anyone let know which of the A380 carriers are using RR and which are GE ?
Thanks

[DERG]

Hiya

RR trents used by Lufthansa, Singapore, Qantas.
Rest are Engine Alliance (GE)

[bearfoil]

Those on the ground who have not sold off the Titanium are using hundreds of TRENT bits as paperweights.

[DERG]

Tell you what Bear..

Turbine D has been a hell of a help on here. Generally the "nay sayers" and "meddlers" have stayed well clear. We did a GOOD JOB on unravelling this mess..

[Turbine D]

WojtekSz

Does this tuning also include the frequencies induces but burning chambers? These seem to be a source of (possibly the most) significant energy vibrations within the engine case.

I think you will be surprised to learn that this is not a major concern in the combustor, and that is not to say vibrations are not created by the mixing and burning of the fuel. Vibration responses are looked at during the design phase using a sophisticated analysis program but, the major concern is that of LCF (eventual crack initiation) due to the thermal gradients throughout the engine cycle, i.e., cold to hot to cold. The reason vibration is not such a problem is the unique way it is handled. The "flame" is contained inside the combustor liner. The liner is "free floating" inside the combustor pressure vessel meaning it is not firmly attached to the pressure vessel in any way. The liner itself is cantilevered by bolting it to the stage 1 HPT nozzle ring. At the turbine nozzle attachment point, a 360° leaf seal is used to prevent hot gas leakage, but it also provides dampening. The actual liner is constructed from several individual pieces. The dome (front end where the fuel nozzles inject fuel) is one piece. It is bolted to the inner and outer liner walls with a series of self locking bolts. The bolts serve as dampers and prevent vibration from transferring to the fuel nozzles that are attached to the main fuel supply tube ring. So there is vibration but it is dealt with by unique design mechanisms to prevent it from being a problem.

[WojtekSz]

Turbine D:
thanks - your clear explanation + some web search produced much better understanding that the combustor tech can be very simple (available to enthusiasts) or complicated (professional big power aeroturbines). But the vibrations may be contained and to certain extent controlled.


Is LCF based on using cooling airflow, protective coutings, disposabe elements to prevent temp buildup or polishing and rounding the edges to prevent cracks? Most probalbe a mix of the above but are there any other interesting techniques?

[Turbine D]

WojtekSz

In answer to your combustor question concerning LCF, just about everything you mentioned is done to improve combustor life. Primarily, I am more familiar with combustors that are not lined, meaning nothing is disposable, but other combustor designs use tiles that can be replaced. The liner walls are cooled with air through a series of small holes to provide a film of cooled air. The flame side of the liners are thermal barrier coated to reduce the magnitude of thermal gradient, e.g. cold - hot - cold. Attention is paid to rounding all edges to minimize potential crack initiation sites. Regardless of the extremes this component sees, they last a long time. Life is very good on engines that fly long distance routes, 10-14 hours in the air, fewer cycles and long hours. There are set periods of time/cycles where boroscope inspection of the combustor is done on wing. If an engine comes off wing regardless of cause, the combustor is inspected. At a specific set time, the combustor is removed and weld repaired as necessary and reassembled with time starting as if it were new. This can typically be done 4 times before it is scrapped.

I should add, the newer combustors utilize twin swirlers that perform better to mix the air with the fuel creating a leaner burn, lower temperatures, improved fuel burn and longer combustor life.

[DERG]

Basic Designs

TURBINE D

"I should add, the newer combustors utilize twin swirlers that perform better to mix the air with the fuel creating a leaner burn, lower temperatures, improved fuel burn and longer combustor life."

So we have a cool end where the big intake fan is, and a hot end where the exhaust gas is released. We have your analogy of the central shaft running from fornt to back...about 7 feet..in the case of the T900 upon which all of the various twirlers and compressors are affixed. Then we have the ancillaries the power take off shaft down to the gearbox and another shaft to start the engine. Much appreciate your input..thank you!

OPEN FORUM

Deduction 1.

Looking at the latest GE design we see that the big intake fan is made from carbon fibre. GE also use a fibre case to house the engine. So I think I am correct is assuming the GE products are using carbon fibre technology where as RR are using advanced metallurgy for the blades and old technology for the case.

Deduction 2.
Because the intake fan is lighter in the GE engine it is easier to balance and has inherently less kinetic energy to be lost in vibration. So the big mass at the front end is now "tamed".

Deduction 3.
Running temps. The 180C and 196C max temp of the T972 that failed is higher than other civil aviation engines to date. We have read the data in the Oxford University document and it is noted that 10% or more of RR engines fail the final passing out test due to imbalance. And a higher rejection rate is detected when the engines are overhauled at the RR appointed agents. The high running temps are cause by vibration inherent to this design.

Deduction 4.
Manufacturing tolerances. The RR design is difficult to manufacture beacuse of the nature of the concentricity of the drive shafts and the fact that is has three distinct compression stages. GE, I believe, use two stages.

Deduction 5
Given that the RR product is inherently difficult to balance it is clear that the existing in service methods of detecting an inbalance are inadequate.
On one drawing I see only one vibe detector noted but this was a schematic and there may be more than one.

Deduction 6
RR agents themselves cannot detect vibration or allow engines that are outside the norms back into service.

Deduction 7
RR are aware that the 3 spool design has inherent flaws but are unwilling to adopt a new design because the costs would be to high.

Deduction 8
As we see the fractured oil feed pipe as published in the ATSB report we must assume that the quality checks on parts from outside suppliers is poor.

Deduction 9
Communication failure. As the fractured oil pipe is clearly of poor manufacture we must deduce that the department which does the invigilation does not realise the importance of that part. In other words they have no concept about the vibrations inherent within the machine.

Deduction 10
Management Character. The commercial managers were easily persuaded that the monitoring system using the probablility model was robust. Because the work was done by assumed experts they did not apply rigour in seeking confirmed evidence. As far as we know they did not hire MIT for instance.

We know that a company called Schaeffer with the brand names FAG and INA were given the contract to build the entire bearing set and support structure. This was the first time this company had built this..the entire assemble with supports... a complex and critical part for a civil avaition aerospace gas turbine.

Deduction 11
The failure to be able to shut down the #1 engine on the stricken Qantas A388 was unforseen.

All repiles most welcome.

[lynn789]

its been suggested that previous RR design errors have been caused by the RR legend of mechanical perfection influencing RR engineers

[DERG]

Yeah I can believe that.