A320 dual engine failure scenario
Join Date: Jan 2014
Location: N5109.2W10.5
Posts: 720
Likes: 0
Received 0 Likes
on
0 Posts
APU didn't save the day for Sully because at no stage Sully had dual engine flame out.
From ntsb.gov/investigations/AccidentReports/Reports/AAR1003.pdf
page 104,
" Even though the engines did not experience a total loss of thrust, the Engine Dual Failure checklist was the most applicable checklist contained in the US Airways QRH, which was developed in accordance with the Airbus QRH, to address the accident event because it was the only checklist that contained guidance to follow if an engine restart was not possible and if a forced landing or ditching was anticipated (starting from 3,000 feet). However, according to postaccident interviews and CVR data, the flight crew did not complete the Engine Dual Failure checklist, which had 3 parts and was 3 pages long. Although the flight crewmembers were able to complete most of part 1 of the checklist, they were not able to start parts 2 and 3 of the checklist because of the airplane’s low altitude and the limited time available."
page 105,
"Although the flight crew was only able to complete about one-third of the Engine Dual Failure checklist, immediately after the bird strike, the captain did accomplish one critical item that the flight crew did not reach in the checklist: starting the APU. Starting the APU early in the accident sequence proved to be critical because it improved the outcome of the ditching by ensuring that electrical power was available to the airplane. Further, if the captain had not started the APU, the airplane would not have remained in normal law mode. This critical step would not have been completed if the flight crew had simply followed the order of the items in the checklist."
Each engine generator would have dropped off line when N2<55% or when the engine master switch is set to off. The windmilling engines would still develop some hydraulic pressure. See Page 55.
"FDR data indicated that both thrust levers were set to the idle position at 1528:01, about 50 seconds after the bird encounter. The N1 and N2 speeds for the left engine both decreased while the speeds for the right engine did not respond. About 30 seconds later, the right engine master switch was moved to the OFF position.95 According to the Airbus FCOM, for an automatic start sequence, when the engine master switch is in the OFF position and the throttle is set at idle, the fuel valve will only open when the N2 speed is more than 15 percent when in flight. When the first officer attempted to move the right engine master switch to the ON position, the N2 speed was less than 15 percent. At 1529:27, the left engine master switch was moved to the OFF position, at which time the N2 speed was about 83 percent. The left engine master switch was moved to the ON position about 10 seconds later, at which time the N2 speed was about 39 percent."
Last edited by Goldenrivett; 27th Apr 2017 at 16:56. Reason: extra text
'What ifs' are to be encouraged, but their weakness is often the lack of context, or safety relevance (risk probability) and thus training value. There are also issues of simulation (training or PC), instructor knowledge, and use of the results.
If the OP question is a technicality, then that aspect can be discussed in isolation, however if it's within an hypothetical scenario then it is vital to consider the preceding events and the overall context of the situation.
The OP assumes that the failures have been clearly identified, without ambiguity, and with low stress; yet in many less severe real aircraft situations, crews suffered startle and surprise with degraded mental ability. With hindsight human activity can appear to be irrational, yet it probably made sense to the crew at the time. Thus it is very important to consider their viewpoint, not just our view in remote isolation from an ambiguous evolving event and very limited timescale.
The OP implied that the crew had sufficient mental model of the local area; this could be a gross assumption. More often the situation model has to be constructed - recognising and linking previous aspects, and only then might options and desired outcome be considered.
As per #4, fly the aircraft. Navigate to a suitable clear space; if this is a runway great, if not manage the approach as best possible to avoid obstacles. Stopping will be the least of your concerns.
Aim for normality where possible, don't think that you can make up procedures on the 'fly'. Aiming for a short low speed landing probably involves greater risk than an overrun.
Thinking for rationalising the situation and planning the outcome, and justifying the course of action will require whatever spare mental is available. The task is to avoid the crash, mitigate the outcome of a non ideal landing. At best we might be able to follow the advice in the checklist, at worst in attempting to rewrite the drills or consider low priority issues we may fail to achieve the primary safety aim.
PS. Crews may not have any more time that that taken to write and check this post; stopping, no more than the time to read the PS.
If the OP question is a technicality, then that aspect can be discussed in isolation, however if it's within an hypothetical scenario then it is vital to consider the preceding events and the overall context of the situation.
The OP assumes that the failures have been clearly identified, without ambiguity, and with low stress; yet in many less severe real aircraft situations, crews suffered startle and surprise with degraded mental ability. With hindsight human activity can appear to be irrational, yet it probably made sense to the crew at the time. Thus it is very important to consider their viewpoint, not just our view in remote isolation from an ambiguous evolving event and very limited timescale.
The OP implied that the crew had sufficient mental model of the local area; this could be a gross assumption. More often the situation model has to be constructed - recognising and linking previous aspects, and only then might options and desired outcome be considered.
As per #4, fly the aircraft. Navigate to a suitable clear space; if this is a runway great, if not manage the approach as best possible to avoid obstacles. Stopping will be the least of your concerns.
Aim for normality where possible, don't think that you can make up procedures on the 'fly'. Aiming for a short low speed landing probably involves greater risk than an overrun.
Thinking for rationalising the situation and planning the outcome, and justifying the course of action will require whatever spare mental is available. The task is to avoid the crash, mitigate the outcome of a non ideal landing. At best we might be able to follow the advice in the checklist, at worst in attempting to rewrite the drills or consider low priority issues we may fail to achieve the primary safety aim.
PS. Crews may not have any more time that that taken to write and check this post; stopping, no more than the time to read the PS.
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,407
Likes: 0
Received 0 Likes
on
0 Posts
From airbus submission to NTSB from DFDR/CVR data. Despite suffering bird(s) ingestion in both engines, the engines did not suffer any uncontained failure but continued to deliver hydraulic and electrical power (engine N°1 N2 was high enough, up to the re-start attempt, for its IDG to remain on-line).
It also depends on speed at this altitude.
You do not need to perform a 360° (In total) turn. Actually a 180° and a bit more will be enough.
The optimal bank angle for that turn is 45° theoretically, but 37° would be more sensible given considerations about getting the correct speed (higher at 45° bank).
Given a 2000fpm rate of descent during the turn and 1700 after, at 37° the half turn should take a bit less than one minute, so you would end up at around 1500ft in front of the runway. Getting able to get there will depend on energy management, airplane performance during the climb, and wind.
Assuming no wind, an average of 10% (6° climb), you would be at 5-6nm at the beginning and end of the turn. At 3° descent rate you would be well aligned for the runway.
However it may be very difficult to guess at the first time when to get the gear and flaps out.
My guess would be flaps 1 as soon as possible, flaps 2 if landing is assured, or delay the flaps 2 if a bit low.
Gear down would be very difficult to time as well, because it may be the emergency gravity procedure. But my guess would be 30s before landing especially if a bit low on energy. Then, using F3/Full to convert speed into altitude if required. May also be used to reduce vertical descent.
Then, if flaring from 1700fpm, flare should begin at approx 100ft and at least 10-15kt higher than the VLS.
With the APU started it may be much easier to do (availability of hydraulics mainly, but also some computers)
Feasible, but on a complex aircraft like this I would be a bit worried about the likeliness of success..
On a light aircraft I've done this procedure without a problem in real conditions. The only problem is to convince the controller to do such an unconventional exercise.
The best option to manage such an approach would be to be very high on the profile. Use F1/gear down to aim for the runway an even speedbrakes if required. Being able to retract the speedbrakes gives the possibility to go back up on the descent profile.
Then, the most important thing about flaring the aircraft is to understand that for same flare intensity, if the vertical descent rate is x times higher, the flare height should be x^2 times higher. I use this technique successfully every time I perform a landing on a 4° slope. Flare height is almost double. Being this early also ensures more time to notice unsatisfactory flare and use higher than normal intensity if required. However for very large vertical speed differences, speed decay during flare becomes sensible and must be taken into account. Hence the 10-15kt which is a guess on my part for 1700fpm.
There is a physics computation that I did, for 1kt of speed trend at 150kt, you will have 75fpm. So to cancel 1700fpm, you will see a 22kt speed trend on the PFD. Flare will last for about 6 seconds so 13 kt lost. My guess was not too far off !
Join Date: Jun 2006
Location: B.F.E.
Posts: 228
Likes: 0
Received 0 Likes
on
0 Posts
Profile wise, clean at green dot fly the circuit at about twice the height you would usually be for a given position. For example if you would typically be 1500 feet at 5 miles (300ft / nm), be at 3000 instead and so on. Save the configuration (at least beyond config 1) until you absolutely have it made, and start to feel just a little high. Give each configuration time to settle before selecting the next (you will need to lower the nose more than you think to maintain a safe speed as the drag builds up). The QRH gives a 1.5 : 1 glide ratio fully configured (A321). Give about 1000 feet to crank the gear down. If you are a little high on final, s-turns or whatever it takes to get down. Better a little high than a little low! Find your nonmoving aim point out the window. That does not lie.
The QRH gives a 1.5 : 1 glide ratio fully configured (A321)
The actual glide ratio should be around 4 to 6.
1.5 is worse than a parachute.
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,407
Likes: 0
Received 0 Likes
on
0 Posts
I'm not so sure about that, however quick decision is key.
It also depends on speed at this altitude.
You do not need to perform a 360° (In total) turn. Actually a 180° and a bit more will be enough.
The optimal bank angle for that turn is 45° theoretically, but 37° would be more sensible given considerations about getting the correct speed (higher at 45° bank).
Given a 2000fpm rate of descent during the turn and 1700 after, at 37° the half turn should take a bit less than one minute, so you would end up at around 1500ft in front of the runway. Getting able to get there will depend on energy management, airplane performance during the climb, and wind.
Assuming no wind, an average of 10% (6° climb), you would be at 5-6nm at the beginning and end of the turn. At 3° descent rate you would be well aligned for the runway.
However it may be very difficult to guess at the first time when to get the gear and flaps out.
My guess would be flaps 1 as soon as possible, flaps 2 if landing is assured, or delay the flaps 2 if a bit low.
Gear down would be very difficult to time as well, because it may be the emergency gravity procedure. But my guess would be 30s before landing especially if a bit low on energy. Then, using F3/Full to convert speed into altitude if required. May also be used to reduce vertical descent.
Then, if flaring from 1700fpm, flare should begin at approx 100ft and at least 10-15kt higher than the VLS.
With the APU started it may be much easier to do (availability of hydraulics mainly, but also some computers)
Feasible, but on a complex aircraft like this I would be a bit worried about the likeliness of success..
On a light aircraft I've done this procedure without a problem in real conditions. The only problem is to convince the controller to do such an unconventional exercise.
The best option to manage such an approach would be to be very high on the profile. Use F1/gear down to aim for the runway an even speedbrakes if required. Being able to retract the speedbrakes gives the possibility to go back up on the descent profile.
Then, the most important thing about flaring the aircraft is to understand that for same flare intensity, if the vertical descent rate is x times higher, the flare height should be x^2 times higher. I use this technique successfully every time I perform a landing on a 4° slope. Flare height is almost double. Being this early also ensures more time to notice unsatisfactory flare and use higher than normal intensity if required. However for very large vertical speed differences, speed decay during flare becomes sensible and must be taken into account. Hence the 10-15kt which is a guess on my part for 1700fpm.
There is a physics computation that I did, for 1kt of speed trend at 150kt, you will have 75fpm. So to cancel 1700fpm, you will see a 22kt speed trend on the PFD. Flare will last for about 6 seconds so 13 kt lost. My guess was not too far off !
It also depends on speed at this altitude.
You do not need to perform a 360° (In total) turn. Actually a 180° and a bit more will be enough.
The optimal bank angle for that turn is 45° theoretically, but 37° would be more sensible given considerations about getting the correct speed (higher at 45° bank).
Given a 2000fpm rate of descent during the turn and 1700 after, at 37° the half turn should take a bit less than one minute, so you would end up at around 1500ft in front of the runway. Getting able to get there will depend on energy management, airplane performance during the climb, and wind.
Assuming no wind, an average of 10% (6° climb), you would be at 5-6nm at the beginning and end of the turn. At 3° descent rate you would be well aligned for the runway.
However it may be very difficult to guess at the first time when to get the gear and flaps out.
My guess would be flaps 1 as soon as possible, flaps 2 if landing is assured, or delay the flaps 2 if a bit low.
Gear down would be very difficult to time as well, because it may be the emergency gravity procedure. But my guess would be 30s before landing especially if a bit low on energy. Then, using F3/Full to convert speed into altitude if required. May also be used to reduce vertical descent.
Then, if flaring from 1700fpm, flare should begin at approx 100ft and at least 10-15kt higher than the VLS.
With the APU started it may be much easier to do (availability of hydraulics mainly, but also some computers)
Feasible, but on a complex aircraft like this I would be a bit worried about the likeliness of success..
On a light aircraft I've done this procedure without a problem in real conditions. The only problem is to convince the controller to do such an unconventional exercise.
The best option to manage such an approach would be to be very high on the profile. Use F1/gear down to aim for the runway an even speedbrakes if required. Being able to retract the speedbrakes gives the possibility to go back up on the descent profile.
Then, the most important thing about flaring the aircraft is to understand that for same flare intensity, if the vertical descent rate is x times higher, the flare height should be x^2 times higher. I use this technique successfully every time I perform a landing on a 4° slope. Flare height is almost double. Being this early also ensures more time to notice unsatisfactory flare and use higher than normal intensity if required. However for very large vertical speed differences, speed decay during flare becomes sensible and must be taken into account. Hence the 10-15kt which is a guess on my part for 1700fpm.
There is a physics computation that I did, for 1kt of speed trend at 150kt, you will have 75fpm. So to cancel 1700fpm, you will see a 22kt speed trend on the PFD. Flare will last for about 6 seconds so 13 kt lost. My guess was not too far off !
Probability and Providence
"It heavily depends on the circumstances, i.e., when and where the dual failure occurred.
The situation will be very different if the engines quit during daylight, in good visibility conditions, at an altitude and distance such that the aircraft can glide to an airport and end up on a runway, or if they fail at low altitude in the middle of nowhere, amid mountains, above the sea, or above a large city.
So the final outcome will heavily depend on ‘providence,’ as well as on the pilots’ gliding skills and ability to make the right decisions."
" Lessons from the Hudson "
https://www.researchgate.net/profile...ication_detail
The situation will be very different if the engines quit during daylight, in good visibility conditions, at an altitude and distance such that the aircraft can glide to an airport and end up on a runway, or if they fail at low altitude in the middle of nowhere, amid mountains, above the sea, or above a large city.
So the final outcome will heavily depend on ‘providence,’ as well as on the pilots’ gliding skills and ability to make the right decisions."
" Lessons from the Hudson "
https://www.researchgate.net/profile...ication_detail
