TransAsia in the water?

[BlueVolta]

So well said.
If it flies, just let it fly until a decent acceleration altitude that gives good ground clearance.
The engines are certified in RTO for 10 minutes, so at least 1000 ft allows for some margin if something goes wrong.
At 400 ft AGL, there is very little margin to correct an error in a turboprop as drag and yawing movement will make the aircraft hard to control.

[edmundronald]

Skyhighfallguy -

Some interested party - eg. ICAO - or even a local authority could do a couple of "no fault" random check sim tests a year for every airline. That would give "people in high places" an idea of what is really going on re flying skills.

Politicians, their wives and children take airplanes too ...

[parkfell]

Clearly the competency testing following the accident has revealed some interesting results. I think their current TRE/TRIs should have their LPC/OPCs conducted by the Regulator until further notice.

Random observations when line pilots have their six monthly recurrency simulator checks are also necessary. This must include a good measure of teaching interesting situations from which a lot can be learnt.

The previous year's MORs will provide a rich source of scenarios. Just requires a bit of imagination and thought.

An interesting one ( a golden oldy) when I was on Saab 340 was ICING and the behaviour of the AP.
I was lucky as a Saab TP involved in Saab icing trials ( when the ATRs did not cope with icing in the USA) supervised by the FAA, often did my six monthly sim checks. I learnt a lot from him.

[Valdiviano]

Pardon my ignorance. Could the C C / third pilot have pulled the PL back?? And from then on things went pear shaped?

[Hotel Tango]

MPN11: Yes, but there's only a very minute fraction of passengers who read PPRuNe. So as I said, the majority are simply not aware.

[Uplinker]

Notwithstanding the actual cause of this accident, which remains to be determined; I have thought for some time - and it has been touched on here - that all LPC's should be conducted by TRE's from ICAO or some other suitable neutral agency - not by TRE's from the pilot's own company.

Examining by some company TREs is top standard, but in other companys, it seems to be very poor.

OPC's could still be conducted by company TRE's, but LPC's should be examined by TRE's entirely separate from the company, and unknown to the candidates, otherwise low standards might be allowed to prevail. An airline has an economic interest in keeping training costs to a minimum, and this could lead to substandard line flying and substandard pilot skill levels. Think about the recent crashes where pilots have made totally inappropriate control inputs when approaching a stall, or failed to monitor their airspeed, or identified the wrong engine etc. What is the cause of these examples of poor flying? It generally seems to be poor training and/or poor testing.

Company TRE's are checked by the authorities, but are then allowed to conduct LPC's autonomously until their next check.

I do of course realise that this would need a complete redesign of the logistics and economics of LPC SIMs, but I would have thought it would increase standards and safety.

[clunckdriver]

For those sugesting that a few of us from the industry could influence change by joining forces and talking to the regulators, I have bad news for you. In the country that I live and fly in the regulator is a major part of the problem, along with totally unqualified Provincial employees who have been able to hijack the training standards and flight training industry, promoting the dumbing down of training and causing huge monetry costs to the industry, all this done on the watch of the previous director of flight training who in retirement now thinks what he has done "might have been a miss- step". Now we have a Minister of Transport allowing of- shore pilots to fly Canadian aircraft without the required training and qualification required by our own regulations! {Having heard/witnesed one of these crews in "panic mode" a week ago when the weather was a bit claggy, we had in fact diverted, I just cant imagine what their background/training is having heard and witnesed their perfomance, in what was an almost normal Canadian Winter event}The regulator in this part of the globe is now looked upon as a bloody joke and have not attracted the right folks to join them in many moons, in fact of the five employees I had to fire years ago, four are now "inspectors". The only thing which will change whats going on is the insurance industry, and a major house cleaning of our Civil Service, Im not holding my breath for this to happen!The rot in pilot training and standards can, in this part of the globe, be laid at the feet of the regulator, untill qualified, independant folks are hired into a broken system, these sad events will not stop, the race to the bottom is a world wide problem, not just an Asian one.

[lomapaseo]

Slippery Pete

If this is simply as it appears... Autofeather #2 (perhaps ATPCS error, doesn't matter) followed by crew pulling torque off and eventually shutting down the operating engine, then the lesson to be learnt is very simple.

Why the #### are these things being shutdown close to the ground? At one turboprop operator I worked for, they were adamant the shutdown with a V1 failure was to be done at acceleration - sometimes as low as 500'. It was mind bogglingly stupid, but it still happens in turboprop sims the world over.

The only reason a shutdown should be actioned so low is for engine failure followed by failure to auto feather (a desperate situation indeed, yet one which had NEVER occurred since type certificate issue, worldwide, on the airframe I used to fly).

The case of a fire may also warrant an immediate shutdown, but should be with fire handle and fire handle alone.

Manipulating power levers and fuel shutoffs just after takeoff when one propeller has auto feathered and the aircraft is still completely climb-able and controllable is just nuts.

That's why these damn things have auto feather in the first place.

If it's climbing, sit on your hands and keep it flying. Anything and everything else can wait.

I hope desperately this was not a case of shutting down the only engine which was developing power.]

Agree ...... but

I'm still open minded that the aircraft having lost the thrust from one engine may have been continuing to lose flying performance from as yet the unexplained symptoms from the remaining engine. The crew may have reason to believe that they could only regain the full power on the remaining engine by restarting it.

This is only speculation on my part but surely it should be vetted in the onward investigation

[ZFT]

For those suggesting that a few of us from the industry could influence change by joining forces and talking to the regulators, I have bad news for you. In the country that I live and fly in the regulator is a major part of the problem, along with totally unqualified Provincial employees who have been able to hijack the training standards and flight training industry, promoting the dumbing down of training and causing huge monetary costs to the industry

Rest assured it's not just your part of the world. Many posters on this thread have been quick to criticise 'Asian' regulators but the principle regulators, FAA, EASA, Transport of Canada etc who should be leading by example are all guilty of dragging the industry down with ever increasing bureaucratic nonsense that adds no training value but does add significant cost. Costs that have to be recovered somewhere.

In our case we spend MORE time and MORE effort satisfying the regulatory requirements on CMS and SMS processes and procedures than we do on the OM and TM and the regulator spend more time auditing the CMS, SMS and associated Compliance Matrix than he does auditing the OM and TM.

It is somewhat ironic that today when the training tools have never been so good, the trainers (generally) so well trained that we seem to be having so many issues.

This particular accident may well highlight a regulatory issue with the manufacturers EASA approved delta course which (for marketing and/or cost reasons?) is quite limited with only 1 X 4 hour FFS session consisting of Severe Icing, Stall, EFATO, Go Arounds, both single and both engines exercises with 3 other days on low level training tools.

The FAA is even worse with NO regulatory requirement for ANY training to move from any 'Classic' ATR to the -600 and viva versa!!

What hope is there for the minor regulators that tend to follow (look up to) the majors?

So yes, I would concur that the regulator is a major part of the problem.

[skyhighfallguy]

To change the subject ever so slightly.

Last weekend I read that the authorities were to release transcripts of CVR and data from FDR on wednesday. This does not seem to have happened.

I must ask, what did they find out on the CVR that would be so damning as to preclude release of transcript? Was it rote? Something like: engine failure, FEATHER ONE.

Even though one was working OK?

It must be a public relations mess to have 14 percent of your pilots fail a written/oral exam AFTER a crash, and after you have warned them that a test is coming. Add to that a complete failure in the cockpit (Potentially), and the PR boys must be working over time.

[ThreeThreeMike]

Originally Posted by no-hoper

No2 did a complete autofeather.Normal readings before AF and after.
The only signal to start the AF is torque below 18%.


ATCPS received a wrong signal.

Until the engine is examined I'd say your conclusion is premature. It's a bit of a stretch to make this statement based on a FDR trace.

Again, this is a moot point. Engines fail, and pilots are expected to deal with it. Regardless of the validity of the indication, the aircraft was by all indications flying normally and climbing on the other engine until the apparent improper manipulation of the flight controls.

[Smott999]

Re CVR call outs

I don't know the veracity of this website or the translation, but if this is "all" there is in the way of CVR call-outs it seems pretty sparse doesn't it? The FDR notation of closing #1 appears to be the initial reduction, and there were several more reductions over the ensuing seconds which are not mentioned....eventually we get CVR callout re #2 flameout....by this time #1 is shut down or nearly so....


ALL 10:41:14.6 GE 235 flight recordings start
CVR 10:51:12.7 Songshan tower issued takeoff clearance
CVR 10:52:33.8 Songshan request GE 235 Contact Taipei Approach
FDR 10:52:38.x Engine Warning (2)
CVR 10:52:38.3 The main warning sound
FDR 10:52:42.x Closed on the 1st engine throttle
CVR 10:52:43.0 A crew mentioned engine recovery (One engine crew referred to recover)
CVR 10:53:00.4 Both crew mentioned engine flameout procedure
CVR 10:53:06.4 A crew mentioned engine recovery (One engine crew referred to recover)
CVR 10:53:07.7 Members mentioned confirm engine #2 flameout
ALL 10:53:09.9 Stall warning sounds (Until 10:53:10.8)
ALL 10:53:12.6 Stall warning sounds (Until 10:53:18.8)
CVR 10:53:19.6 A crew mentioned Engine #1 already feathered and the oil pressure was reducing
ALL 10:53:21.4 Stall warning sounds (Until 10:53:23.3)
FDR 10:53:24.x Off the 1st engine
ALL 10:53:25.7 Stall warning sounds (Until 10:53:27.3)
CVR 10:53:34.9 Contact Songshan tower crew call “mayday mayday engine flameout”
ALL 10:53:55.9 Stall warning sounds (Until 10:53:59.7)
ALL 10:54:06.1 Stall warning sounds (Until 10:54:10.1)
CVR 10:54:09.2 Crew began to call engine restart (Members began to call again to drive)
ALL 10:54:12.4 Stall warning sounds (Until 10:54:21.6)
FDR 10:54:20.x Engine 1 Restart (Reboot the 1st engine)
ALL 10:54:23.2 Stall warning sounds (Until 10:54:33.9)
CVR 10:54:34.4 The main warning sound
CVR 10:54:34.8 Unknown sound
ALL 10:54:36.6 Recorder stops recording
ASC explained that the warnings started about 37 seconds after takeoff (at an altitude of about 1,000 – 1,200 feet )

[de facto]

OPC's could still be conducted by company TRE's, but LPC's should be examined by TRE's entirely separate from the company, and unknown to the candidates, otherwise low standards might be allowed to prevail.

of course realise that this would need a complete redesign of the logistics and economics of LPC SIMs, but I would have thought it would increase standards and safety.

I believe it to be a very good idea....for example Star alliance TRE should be going from one airline to the other...basic knowledge of company SOP and off you go.

[noalign]

@ZFT

The FAA is even worse with NO regulatory requirement for ANY training to move from any 'Classic' ATR to the -600 and viva versa!!

Are you basing this upon a reading of both the FSB for the ATR and AC 120-53B, a particular operator's differences program, or what?

[skyhighfallguy]

Doesn't it boil down to this?

Rotten Training.

Questionable Basic flying skills and fundamentals of understanding of energy management.

Terrible over sight by local FAA equivilent.

Money being spent to buy planes but not train pilots to a real understanding as well as proficiency in the art of flying in normal and non normal situations.

I would add this: is it proper to sell a gun to a 5 year old? is it proper to sell a complicated airplane to an airline that can't handle the training and repair? It makes sense to me that some countries, some airlines are not ready to handle things to high western standards. PERHAPS the answer is to wet lease planes and crews from the manufacturer who will bear all training costs and mx costs and responsbilities/liabilities for the wet lease cost.

[RatherBeFlying]

Fast Hands

I suspect the Viscount was the first aircraft with autofeather. I saw lots of folks in the sim lose an engine and just keep on flying without doing much of anything.

Unless the engine was on fire, a restart would be attempted at height.

The pistons needed to be feathered manually to have a chance of climbing which put the emphasis on fast hands, but they came with an FE who often caught a sagging engine before it got seriously sick.

Too many CAAs still want to see fast hands. Maybe it's still needed for those piston twins still out there, but definitely not for turboprops with working autofeather and a 50/50 chance of ending up with no engines producing power

[Tu.114]

It seems to me that there is some exaggeration going on here at the moment.

It is true; most turboprops around at the moment have Autofeather, and it is also true that it works as advertised most of the time. In an engine failure after V1, memory items are initially limited to the two questions:

1. Is the propeller feathered?

2. Is there a fire indicated?

(3. Did the uptrim work on the good engine - type dependent)

Answering those just requires checking of the instruments and no manipulation of any lever. And if both conditions are fulfilled, there is no more to be done until acceleration and cleanup. Then the book is taken out and the lists are worked off. So the only difference to a jet is the positive need to have the propeller feathered.

Workload starts to get a bit higher if either of the two conditions are not met. This is not because of some authority wishing to "see hands fly" at all costs; the aircraft just needs to be brought into a condition covered by performance calculations by feathering the propeller and a possible fire needs to be fought (this, again, is just as it is on a propeller-less aircraft). No flying hands or frantic actions are involved; such a procedure can and indeed should be handled in a calm and proper way just like any other abnormal procedure on any aircraft there may be. There is always time e. g. for the PF to properly verify that the PNF has his hands on the appropriate power lever, condition lever and T handle.

Also, please do not forget the role of the manufacturer in creating such a procedure. It may well be possible for him to allow flying the aircraft with one windmilling propeller until acceleration, but if climb gradients are still to be met, this will require reducing the max takeoff weight by offloading passengers and other payload and therefore make the aircraft unsellable to airlines.

Just to show some figures: on the DH8-300, an unserviceable Autofeather was releasable under MEL. This meant that an engine failure after V1 would have required manual feathering of the engine, which would have taken longer than with the autofeather working. And this alone caused a MTOW reduction that would have forced me once to offload 30 of the 40 passengers on board AND everything that was in the cargo compartment, had I accepted this aircraft. With no feathering at all until acceleration altitude, the restrictions would even be more severe and I would be surprised if the MTOW still allowed taking fuel on board.

[Lonewolf_50]

Originally Posted by skyhighfallguy

Money being spent to buy planes but not train pilots to a real understanding as well as proficiency in the art of flying in normal and non normal situations.

There are a few countries in this world who do something similar in the realm of military hardware. They purchase some pretty good kit but their training methods don't prepare their operators on how to use them. The results in that case become obvious in a rather grim manner. I had never thought to apply the analogy to air transport ... but am disturbed to see that this might be the case.

[IcePack]

1/2 the trouble is that so many non pilots inc the CEO's & accountants, have such "chips" on their shoulders & consider pilots expensive premadonners. Flying is easy isn't it not like my job etc unfortunately it is not. On occasions it is the most demanding of jobs which needs continuous training & practice. But that would undermine those non flyers mind set of expensive premadonners.
This has led to using the aircraft to practise for the sim, which is the opposite for which sim's were invented for.
So give this crew a break. If they were at fault the onus falls fair an square at the feet of those who follow the perception that flying is easy and those check airmen who go along with that mindset.

[Slippery_Pete]

Tu.144

I agree with most of what you've said. The need in a modern turboprop with a V1 failure is one of monitoring, not throwing switches and levers.

As for workload when a failure to autofeather occurs, or when a fire and failure occur simultaneously... Can anyone actually tell me the last time a modern turboprop had this occur? Engine failures between V1 and MSA are very, very rare these days - couple this with a simultaneous fire or failure to autofeather... I simply can't remember this having happened to any airline turboprop airframe type, the world over, for the last twenty years. As for monitoring torque push on the live and auto feather status, let the PNF worry about it. Just fly the OEI attitude and look at the VSI. Either it's climbing like it should (autofeather and torque push have occurred) or it isn't climbing like it should and something is drastically wrong. The correct attitude and the VSI will tell you. No need for slick hands or fancy callouts.

Yes, autofeather failure or simultaneous fire and failure needs to be trained for, but let's keep it all in perspective. It's a terribly, terribly unlikely thing to occur at takeoff.

If it's flying and climbing, it's flying and climbing. Get the thing safely accelerated and cleaned up, burn the good engine to its OEI time limit before reducing to MCP, get it to MSA, accelerate some more... And then, with the AP on if type allows, two pilots can concentrate on carefully pulling the dead engine into feather/fuel cutoff.

I don't for a second imply that this is definitely what has happened in this case. However, regardless of the cause of this accident, the way turboprop crews are trained on this has really been bugging me for a long time.