TransAsia in the water?
Join Date: Oct 2006
Location: East Sussex
Age: 86
Posts: 276
Likes: 0
Received 0 Likes
on
0 Posts
The statement that immediate action is required in the event of a fire indication at a critical flight juncture (like just after take off) is, in my opinion, bad advice. Firstly, it may be a false warning and, secondly, even if actually on fire a gas turbine is most likely to still be producing power. No action should be taken until at least 1000ft apart for the NHP to state the fault and the engine concerned and for the HP to positively confirm the diagnosis. The Thompson bird ingestion on rotate at Manchester should be usd repeatedly as an example.
Absolutely pontifex.
If the fire indication is not accompanied by a complete loss of power, get yourself cleaned up and accelerated first.
The other good thing about a fire (if the aircraft systems and checklists are well designed) is you can completely shut it down and feather by simply pulling the illuminated handle, rather than messing about with power and/or prop levers. A lot less chance of getting the wrong one. Certification should be updated to require ONLY pulling a flashing fire handle and nothing else.
If the fire indication is not accompanied by a complete loss of power, get yourself cleaned up and accelerated first.
The other good thing about a fire (if the aircraft systems and checklists are well designed) is you can completely shut it down and feather by simply pulling the illuminated handle, rather than messing about with power and/or prop levers. A lot less chance of getting the wrong one. Certification should be updated to require ONLY pulling a flashing fire handle and nothing else.
Slippery Pete...
Broadly, I agree. But with the proviso which you alluded to: namely, that this advice pertains to modern aircraft only. Maybe. There remain sufficient CV580s, Electras, C-130s, HS748s, and their smaller cousins to narrow your emphasis to the more modern versions only, at best.
In the earlier generation of turboprops the acceleration altitude was typically 400', and that for good reason. I have had a few actual powerplant failures in T-56* engine/Aeroproducts 601 propellors to testify to the need for timely intervention. PW engines too. **But those were primitive installations (and times) compared to the engine/props relevant to this discussion.
* Allison 501-D13 in civilian application
The propellor makes a thrust loss so much more demanding than on a jet: the ability of a windmilling prop/engine to extract all the energy that the good powerplant is producing (made worse with acceleration) is not understood by the jet only guys, and their advice should be taken with a grain of salt, if in fact you can identify it as such. Not all transport aircraft are created equally, nor is all transport aircraft experience.
Anyway...lots of suppositions in this thread, not much emphasis on confirmation bias about a heavily taxed crew already concerned about No.1, then faced with an auto fx on No.2. Most accidents, properly considered, end up being a study in perceptions and indications. I don't expect that this one will be too different.
Broadly, I agree. But with the proviso which you alluded to: namely, that this advice pertains to modern aircraft only. Maybe. There remain sufficient CV580s, Electras, C-130s, HS748s, and their smaller cousins to narrow your emphasis to the more modern versions only, at best.
In the earlier generation of turboprops the acceleration altitude was typically 400', and that for good reason. I have had a few actual powerplant failures in T-56* engine/Aeroproducts 601 propellors to testify to the need for timely intervention. PW engines too. **But those were primitive installations (and times) compared to the engine/props relevant to this discussion.
* Allison 501-D13 in civilian application
The propellor makes a thrust loss so much more demanding than on a jet: the ability of a windmilling prop/engine to extract all the energy that the good powerplant is producing (made worse with acceleration) is not understood by the jet only guys, and their advice should be taken with a grain of salt, if in fact you can identify it as such. Not all transport aircraft are created equally, nor is all transport aircraft experience.
Anyway...lots of suppositions in this thread, not much emphasis on confirmation bias about a heavily taxed crew already concerned about No.1, then faced with an auto fx on No.2. Most accidents, properly considered, end up being a study in perceptions and indications. I don't expect that this one will be too different.
I don't think that the acceleration altitude of 400 feet was chosen as a result of the poor performance of early turbo props, but rather was just an arbitrary number.
Because windmilling drag increases with acceleration, there is a very strong case for NOT commencing acceleration until AFTER the memory items of any fire/manual feather drill have been completed. Even on the much simpler jets, all too often crews in simulator checks make a right pig's ear of the drill by attempting to do memory items simultaneous with acceleration. The result is inevitably that flap speeds are exceeded and the optimum final climb speed is not established.
Wrong as it may be from a certification point of view, I tell them to do one or the other in sequence - either finish the memory items then accelerate even if it means the aircraft is well above minimum acceleration altitude, or if it is less urgent, complete the acceleration then complete the memory items in the final climb segment.
Where it gets interesting is if the failure/fire is introduced just as acceleration is commenced. It is very hard for crews to resist trying to do it all at once, rather than sticking with whatever has been initiated.
Although we train for the classic failure at v1, more emphasis is needed on dealing with failures at random times.
I refrain from comment on this particular accident or the airline's training regime until the report is finalised.
Because windmilling drag increases with acceleration, there is a very strong case for NOT commencing acceleration until AFTER the memory items of any fire/manual feather drill have been completed. Even on the much simpler jets, all too often crews in simulator checks make a right pig's ear of the drill by attempting to do memory items simultaneous with acceleration. The result is inevitably that flap speeds are exceeded and the optimum final climb speed is not established.
Wrong as it may be from a certification point of view, I tell them to do one or the other in sequence - either finish the memory items then accelerate even if it means the aircraft is well above minimum acceleration altitude, or if it is less urgent, complete the acceleration then complete the memory items in the final climb segment.
Where it gets interesting is if the failure/fire is introduced just as acceleration is commenced. It is very hard for crews to resist trying to do it all at once, rather than sticking with whatever has been initiated.
Although we train for the classic failure at v1, more emphasis is needed on dealing with failures at random times.
I refrain from comment on this particular accident or the airline's training regime until the report is finalised.
Join Date: Mar 2010
Location: L.A.
Age: 56
Posts: 579
Likes: 0
Received 0 Likes
on
0 Posts
Ratherbe:
To many CAAs still want to see fast hands.
To many CAAs still want to see fast hands.
When I went east I was told to leave everything, and throw it into a very tight 500' circuit and land. No checks, no shutdown. Invariably many pilots, who were not allowed to hand fly anyway, missed the final turn completely and landed on a taxiway or the apron.
I was horrified, but had to comply.
Join Date: Aug 2008
Location: Sydney
Posts: 79
Likes: 0
Received 0 Likes
on
0 Posts
Again? Better outcome this time around.
Incident: Transasia AT72 at Taipei on Feb 21st 2015, engine loses power
Incident: Transasia AT72 at Taipei on Feb 21st 2015, engine loses power
Transasia Avions de Transport Regional ATR-72-212A, registration B-22806 performing flight GE-507 from Taipei Songshan to Makung (Taiwan) with 20 passengers, was climbing out of Taipei's Songshan Airport when the crew received abnormal torque indications for the left hand engine (PW127) suggesting loss of engine power, stopped the climb and returned to Songshan Airport for a safe landing 19 minutes after departure.
The airline reported that the aircraft is undergoing an extensive examination even though so far no mechanical issue has been identified and the indication is believed to have been false. The passengers were taken to Makung by a replacement aircraft.
Transasia Avions de Transport Regional ATR-72-212A, registration B-22806 performing flight GE-507 from Taipei Songshan to Makung (Taiwan) with 20 passengers, was climbing out of Taipei's Songshan Airport when the crew received abnormal torque indications for the left hand engine (PW127) suggesting loss of engine power, stopped the climb and returned to Songshan Airport for a safe landing 19 minutes after departure.
The airline reported that the aircraft is undergoing an extensive examination even though so far no mechanical issue has been identified and the indication is believed to have been false. The passengers were taken to Makung by a replacement aircraft.
Loss of correct torque indication is not that unusual. Fluctuating torque could be anything from gauge or transmitter problems to damage within the engine.
Dispatch without torque indication would be a no go, so a return to major maintenance base would also be a smart move. It seems that the crew decided that it did not warrant shutting the engine down, or that would have probably been added to the report in typically dramatic terms. So, what is hard to imagine?
The problem is, any turn back these days is cause for all sorts of speculation and innuendo. Particularly when the airline concerned is already in the spotlight for its recent accidents.
Dispatch without torque indication would be a no go, so a return to major maintenance base would also be a smart move. It seems that the crew decided that it did not warrant shutting the engine down, or that would have probably been added to the report in typically dramatic terms. So, what is hard to imagine?
The problem is, any turn back these days is cause for all sorts of speculation and innuendo. Particularly when the airline concerned is already in the spotlight for its recent accidents.
Join Date: Feb 2005
Location: flyover country USA
Age: 82
Posts: 4,579
Likes: 0
Received 0 Likes
on
0 Posts
I'd find it hard to reconcile torque fluctuation without some other parameters also varying - either core rpm & turbine temp (gas generator issues), or else Np (prop governor/control).
Torque fluctuation, by itself without other confirming parameters, is almost certainly a sensor or indication problem.
Torque fluctuation, by itself without other confirming parameters, is almost certainly a sensor or indication problem.
Join Date: Aug 2008
Location: Sydney
Posts: 79
Likes: 0
Received 0 Likes
on
0 Posts
Update
Incident: Transasia AT72 at Taipei on Feb 21st 2015, engine loses power
Just throwing a theory out there, maybe the airline just got a bad set of sensors installed across a range of their aircraft and nothing was ever wrong?
Would a bad sensor(s) deliver the kind of FDR report that came out with the accident aircraft?
On Feb 22nd 2015 Taiwan's Civil Aviation Authority reported that the engine was working properly, the sensor system however delivered false indications. Parts of the sensor system have been replaced, examination and testing continues.
Would a bad sensor(s) deliver the kind of FDR report that came out with the accident aircraft?
Join Date: Feb 2010
Location: Norden
Posts: 138
Likes: 0
Received 0 Likes
on
0 Posts
About the sensors from AVH:
By no-hoper on Sunday, Feb 15th 2015 21:59Z
@Anne. Indeed there is a small buckle for a couple seconds.I think it is related to the closing of the bleed valve.
I checked the wiring manual for the signal required to start autofeather: The signal of one
torque sensor (2ea installed per engine) via the autofeather unit and displayed as the torque pointer will start autofeather.
The MEL for a u/s sensor requires ATCPS off,same for the autofeather unit.
Note for both of them: select off to prevent uncommanded autofeather.
By no-hoper on Sunday, Feb 15th 2015 21:59Z
@Anne. Indeed there is a small buckle for a couple seconds.I think it is related to the closing of the bleed valve.
I checked the wiring manual for the signal required to start autofeather: The signal of one
torque sensor (2ea installed per engine) via the autofeather unit and displayed as the torque pointer will start autofeather.
The MEL for a u/s sensor requires ATCPS off,same for the autofeather unit.
Note for both of them: select off to prevent uncommanded autofeather.
Join Date: Jan 2008
Location: sweden
Age: 83
Posts: 12
Likes: 0
Received 0 Likes
on
0 Posts
Unreliable torq indication may indicate upcoming problem.
Surviving eight engine failures with engines like deHavillandGoblin, Allison C-18, Lycoming VO-540, General Electric CT58, PT6A-50 and PT6T-3, I may say the TQ indication in four of the occasions did show unreliable and erratic readings a few seconds before any other instrument or whistle went on.
These four failures were all related to serious internal engine oil problem.
Surviving eight engine failures with engines like deHavillandGoblin, Allison C-18, Lycoming VO-540, General Electric CT58, PT6A-50 and PT6T-3, I may say the TQ indication in four of the occasions did show unreliable and erratic readings a few seconds before any other instrument or whistle went on.
These four failures were all related to serious internal engine oil problem.
Join Date: Dec 2014
Location: United States
Posts: 105
Likes: 0
Received 0 Likes
on
0 Posts
Taipei, Feb. 25 (CNA) TransAsia Airways ATR 72 pilots who did not pass a competency test following the carrier's second deadly air crash in less than a year will receive a second test in mid-March, the Civil Aeronautics Administration (CAA) said Wednesday. The 10 pilots who failed the evaluation after a TransAsia ATR 72 turbo-prop plane crashed on Feb. 4 were suspended from flying and have been undergoing remedial training since the Lunar New Year holiday, CAA officials said.
Another 45 TransAsia ATR 72 pilots who passed the exam will be tested on a flight simulator in Bangkok by the end of February, and those who fail the simulator evaluation will also be suspended from active duty, according to the agency.
The airline currently has 13 other ATR 72 pilots training overseas, and they will be given the same two-round evaluation as their colleagues after returning to Taiwan, the CAA said.
Another 45 TransAsia ATR 72 pilots who passed the exam will be tested on a flight simulator in Bangkok by the end of February, and those who fail the simulator evaluation will also be suspended from active duty, according to the agency.
The airline currently has 13 other ATR 72 pilots training overseas, and they will be given the same two-round evaluation as their colleagues after returning to Taiwan, the CAA said.
autofeather question
It has been 10 years since I went from turbo prop to jet, and I cannot remember the details. Also I flew the F50 and DHC8 and not the ATR. My questions:
What does the autofeather do? (I know the basis function but what else):
-what triggers autofeather?
- does it reduce fuel flow to the "failed" engine?
- is the "objective TQ" increase a power up trim on the "working engine" or just a adjusted position of the torq bug requiring manual power increase on the working engine?
It looks like at 02:52:35 pretty much simultaneously :
- MW ENG 2 (flame out warning)
- start of increase in Beta ENG2 (feather in progress)
-ENG1 Bleed valve close (more power available on working engine)
- drop in all engine parameters for ENG 2 (FF, NH, NL, ITT) without PLA ENG2 movement (so that is a big one, will an autofeather of a working engine lead to over torque or reduce fuel flow?)
Looks like an autofeather but before the torque started dropping of. (is torque the only trigger for autofeather?)
From there on it looks like the pilot started reducing power on ENG 1 (02:52:43-02:53:15), tried to increase power on ENG 2 (PLA2 increase 02:53:15), and ENG 1 shut off at 02:53:20.
My cent (two cents corrected for inflation since it has been 10 years of the props)
I don't think any of the indications were crossed, it would have been apparent during start.
To me it looks like while they were looking at the problem they kept reducing power on the working engine, all engine indications on ENG 2 are consistent with the PLA change and subsequent shutdown.
Because I don't know if the autofeather reduces fuel flow after it senses failure and tries to feather I don't know if ENG 2 failed with a working autofeather or if the autofeather system feathered a working ENG 2 BUT:
The pilots should know how/what the autofeather does.
I was always trained to use power reduction as a confirmation for the failed engine, if you get a stall warning after you reduce power the power should have been pushed up (followed by a sheepish chuckle and "darn, I guess it was the other one").
It doesn't really matter what was wrong with ENG 2, it seems clear the pilots shut down a working engine for no apparent reason. I would love a well translated CVR readout, seems appropriate considering the leaked/released FDR data (and required testing of the remaining pilots) pretty much hangs them already.
What does the autofeather do? (I know the basis function but what else):
-what triggers autofeather?
- does it reduce fuel flow to the "failed" engine?
- is the "objective TQ" increase a power up trim on the "working engine" or just a adjusted position of the torq bug requiring manual power increase on the working engine?
It looks like at 02:52:35 pretty much simultaneously :
- MW ENG 2 (flame out warning)
- start of increase in Beta ENG2 (feather in progress)
-ENG1 Bleed valve close (more power available on working engine)
- drop in all engine parameters for ENG 2 (FF, NH, NL, ITT) without PLA ENG2 movement (so that is a big one, will an autofeather of a working engine lead to over torque or reduce fuel flow?)
Looks like an autofeather but before the torque started dropping of. (is torque the only trigger for autofeather?)
From there on it looks like the pilot started reducing power on ENG 1 (02:52:43-02:53:15), tried to increase power on ENG 2 (PLA2 increase 02:53:15), and ENG 1 shut off at 02:53:20.
My cent (two cents corrected for inflation since it has been 10 years of the props)
I don't think any of the indications were crossed, it would have been apparent during start.
To me it looks like while they were looking at the problem they kept reducing power on the working engine, all engine indications on ENG 2 are consistent with the PLA change and subsequent shutdown.
Because I don't know if the autofeather reduces fuel flow after it senses failure and tries to feather I don't know if ENG 2 failed with a working autofeather or if the autofeather system feathered a working ENG 2 BUT:
The pilots should know how/what the autofeather does.
I was always trained to use power reduction as a confirmation for the failed engine, if you get a stall warning after you reduce power the power should have been pushed up (followed by a sheepish chuckle and "darn, I guess it was the other one").
It doesn't really matter what was wrong with ENG 2, it seems clear the pilots shut down a working engine for no apparent reason. I would love a well translated CVR readout, seems appropriate considering the leaked/released FDR data (and required testing of the remaining pilots) pretty much hangs them already.
Last edited by hans brinker; 27th Feb 2015 at 00:16.
Join Date: Jan 2008
Location: Hotel Sheets, Downtown Plunketville
Age: 76
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
autofeather question
NSTB ASRS REPORT NO 823457 of Feb 2009 involving an ATR72 reads as follows.
During takeoff, the #2 propeller went into feather. I aborted the takeoff and returned to the ramp. Maintenance told me to do a run-up with the Automatic takeoff Power Control System off, because he suspected that it was sending a wrong electronic signal. I did the run-up with the Automatic Takeoff Power Control System on and off and both engines were operating normally. I concluded that since the aircraft's electrical system had been reset, the situation was corrected. I discussed the situation with the First Officer and we both felt that it was safe to fly.
Events
Anomaly.Aircraft Equipment Problem : Critical
Detector.Person : Flight Crew
Result.Flight Crew : Rejected Takeoff
Assessments
Contributing Factors / Situations : Aircraft
Contributing Factors / Situations : Human Factors
Primary Problem : Human Factors
During takeoff, the #2 propeller went into feather. I aborted the takeoff and returned to the ramp. Maintenance told me to do a run-up with the Automatic takeoff Power Control System off, because he suspected that it was sending a wrong electronic signal. I did the run-up with the Automatic Takeoff Power Control System on and off and both engines were operating normally. I concluded that since the aircraft's electrical system had been reset, the situation was corrected. I discussed the situation with the First Officer and we both felt that it was safe to fly.
Events
Anomaly.Aircraft Equipment Problem : Critical
Detector.Person : Flight Crew
Result.Flight Crew : Rejected Takeoff
Assessments
Contributing Factors / Situations : Aircraft
Contributing Factors / Situations : Human Factors
Primary Problem : Human Factors
Join Date: Dec 2007
Location: Wales/now Canada
Age: 82
Posts: 37
Likes: 0
Received 0 Likes
on
0 Posts
During a turboprop take-off, V1 could occur with 8,000 to 10,000 ft of runway remaining at an international airport. If an engine failure and a fire warning occured just after V1 ,would it not be prudent to cut all power and use the remaining runway for a controlled landing and shutdown if necessary. Just a thought from an old timer.
Join Date: Feb 2011
Location: Leeds
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
What does the autofeather do? (I know the basis function but what else):
-what triggers autofeather?
- does it reduce fuel flow to the "failed" engine?
- is the "objective TQ" increase a power up trim on the "working engine" or just a adjusted position of the torq bug requiring manual power increase on the working engine?
-what triggers autofeather?
- does it reduce fuel flow to the "failed" engine?
- is the "objective TQ" increase a power up trim on the "working engine" or just a adjusted position of the torq bug requiring manual power increase on the working engine?
The AF doesn't cut the fuel
we get 2.5 minutes at this setting, as i said its a different type but similar engine (PW126)
Join Date: Apr 2009
Location: Sentul
Age: 55
Posts: 41
Likes: 0
Received 0 Likes
on
0 Posts
You are a real wise old timer; I like the way you think.
However, anal retentive powers that be do not allow this prudent de isin making anymore. We must " follow the book ". Otherwise you have arm chair experts gunning for your six.
An f/o as PF touching down at 3001 feet beyond the threshold on a 14000ft LDA runway at well below MLDW will get the captain in big doo-doo, thanks to FOQA. No amount of explaining as to the stupidity of waiving off such a landing when all things under safe control will satisfy big ego FOQA hawks who just want to see you punished. They need " to be seen doing a great job " of taking others a notch or two to satisfy a target quota.
However, anal retentive powers that be do not allow this prudent de isin making anymore. We must " follow the book ". Otherwise you have arm chair experts gunning for your six.
An f/o as PF touching down at 3001 feet beyond the threshold on a 14000ft LDA runway at well below MLDW will get the captain in big doo-doo, thanks to FOQA. No amount of explaining as to the stupidity of waiving off such a landing when all things under safe control will satisfy big ego FOQA hawks who just want to see you punished. They need " to be seen doing a great job " of taking others a notch or two to satisfy a target quota.
Join Date: Jan 2008
Location: US/EU
Posts: 694
Likes: 0
Received 0 Likes
on
0 Posts
Crew of Crashed Taiwan TransAsia Plane Shut Off Working Engine
TAIPEI — The crew of a TransAsia Airways ATR plane that crashed in Taiwan in February, killing 43 people on board, had shut off the working engine after the other lost power, a source with direct knowledge of the matter said on Wednesday.
http://www.nytimes.com/reuters/2015/...-airplane.html
http://www.nytimes.com/reuters/2015/...-airplane.html
