NO FLARE call
Thread Starter
Join Date: Mar 2017
Location: London
Posts: 170
Likes: 0
Received 0 Likes
on
0 Posts
NO FLARE call
I know Airbus say that the NO FLARE call is not required but what is the likelyhood of the aircraft not flaring? I know it's not often but anyone got a figure. 1 in a million?
Only half a speed-brake
It is a part of fail-operational system, so definitely much less than 1:10^6.
Airbus says: "If no FLARE at 30 ft and visual references not sufficient, execute a go around."
To call "No Flare" by PM is part of the trained task sharing. Unlike "No LAND (at 350')", it needs to be verbalized because PF is not scanning the instruments at that stage.
"Can no FLARE situation happen on a fail-operational system without triggering the AUTOLAND warning" - that would be a good question.
Airbus says: "If no FLARE at 30 ft and visual references not sufficient, execute a go around."
To call "No Flare" by PM is part of the trained task sharing. Unlike "No LAND (at 350')", it needs to be verbalized because PF is not scanning the instruments at that stage.
"Can no FLARE situation happen on a fail-operational system without triggering the AUTOLAND warning" - that would be a good question.
Last edited by FlightDetent; 9th Apr 2017 at 09:38.
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,406
Likes: 0
Received 0 Likes
on
0 Posts
I know Airbus say that the NO FLARE call is not required
PRO-NOR-SRP-01-70 P 12/20
Join Date: Sep 2011
Location: FL390
Posts: 238
Likes: 0
Received 0 Likes
on
0 Posts
Not necessarily fail-operational though, is it? CAT 3 single, CAT 2 etc. I've seen the "no flare" scenario in the sim so it's a plausible failure, and the only effective solution is for PM to call "go around" as soon as they realise what's happened.
Only half a speed-brake
This could be interesting ...
CAT 3 single and CAT 2 are downgrades for failures, not that the remaining system would lose the capability to light up the RED Autoland for whatever reason there is preventing the FLARE from showing up. Moot point a little, point well taken, FF. But for 3 dual - any objections to the statement?
So what is your company's call by PM if FLARE does not come up?
- "No FLARE"
- "Go-Around"
- or, not trained at all.
Eventually, vilas, we will find that FCOM does not stipulate any NO FLARE call. Unlike, for instance, "No reverser, no decel". Also, there isn't any "NO ROLLOUT".
My inital TR in TLS was "no flare", did work later for operator with "go around" call and found their explanation sensible. These days, my rogue thoughts lean to option three. It seems, the sole reason for training this is that the SIM does have the button!
CAT 3 single and CAT 2 are downgrades for failures, not that the remaining system would lose the capability to light up the RED Autoland for whatever reason there is preventing the FLARE from showing up. Moot point a little, point well taken, FF. But for 3 dual - any objections to the statement?
So what is your company's call by PM if FLARE does not come up?
- "No FLARE"
- "Go-Around"
- or, not trained at all.
Eventually, vilas, we will find that FCOM does not stipulate any NO FLARE call. Unlike, for instance, "No reverser, no decel". Also, there isn't any "NO ROLLOUT".
My inital TR in TLS was "no flare", did work later for operator with "go around" call and found their explanation sensible. These days, my rogue thoughts lean to option three. It seems, the sole reason for training this is that the SIM does have the button!
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,406
Likes: 0
Received 0 Likes
on
0 Posts
FLARE or if not No FLARE and PF trained to execute a GA without prompting seems logical. There may not be enough time to say NO FLARE GO AROUND. If PM merely calls Go Around PF may wonder why and hesitate and hard touch down can result leading to disconnection of the AP during GA.
Join Date: Jan 2014
Location: N5109.2W10.5
Posts: 720
Likes: 0
Received 0 Likes
on
0 Posts
My Ops Manual says:
"At flare height.
If FLARE does not come up on FMA, a go-around must be performed.
If visual references are sufficient and a manual landing is possible, the PF may decide to complete the landing."
Q. How does PM know that PF has sufficient visual reference to complete the landing or not?
PM should normally hear, "50, Flare, 30, 20, 10, Retard".
If PF only hears "50 ... 30, 20" etc. then that should be sufficient to realise PM has not called "Flare" and then take the appropriate action.
"At flare height.
If FLARE does not come up on FMA, a go-around must be performed.
If visual references are sufficient and a manual landing is possible, the PF may decide to complete the landing."
Q. How does PM know that PF has sufficient visual reference to complete the landing or not?
PM should normally hear, "50, Flare, 30, 20, 10, Retard".
If PF only hears "50 ... 30, 20" etc. then that should be sufficient to realise PM has not called "Flare" and then take the appropriate action.
To call "No Flare" by PM is part of the trained task sharing. Unlike "No LAND (at 350')", it needs to be verbalized because PF is not scanning the instruments at that stage.
Join Date: Sep 2002
Location: ex-DXB
Posts: 927
Likes: 0
Received 0 Likes
on
0 Posts
I know Airbus say that the NO FLARE call is not required but what is the likelyhood of the aircraft not flaring? I know it's not often but anyone got a figure. 1 in a million?
We would have heard about it if so..!
Since Airbus has been making aircraft for decades and with the multi sectors that these aircraft fly worldwide each day, it must be considerably more than 1 in a million.
The requirements re failure to flare are in CS-AWO and CS-25 (.1329?), but these are not presented as a simple probability.
Basically any failure the overall system should not result in a fatality with a probability something like 10^-9.
These are only numbers, theory, whereas in practice the auto technical reliability is very good, and much better than the human for the majority of landing tasks in the same conditions.
Flare is usually a sub mode of the flight guidance system. In modern high integrity systems lack of mode change or sensor failure can be detected and used to disengage the autopilot or revert to a lower level of operation; the latter being somewhat irrelevant at low altitude - fail op reverts to fail passive, but the aircraft still lands.
In addition many systems use a trim up function which biases the aircraft to pitch nose up if the autopilot disengages at low altitude. There are many reasons for such a system, control torque, minimising GA height loss, and accident severity - probability of fatality.
The need for a 'no flare' call may have been carried over from older systems - simplex monitored or some duplex autopilots with a 'bolt-on' flare mode. Modern systems should not need a call. How would a failure be detected and how could the pilot intervene in time. Older systems suffered more 'failures' due to unwarranted human intervention than there were actual failures.
Those operators who still require a call might consider the type of system they are using, the expectations of crews ability to monitor, and how they might react. If procedures require a call for a mode change, what happens when it doesn't happen. Is the mode change proportional to rad alt, if so why not call altitude - auto call outs do that any way.
And those who use HUD might wish to make a similar evaluation and be prepared to argue the differences between autoflight failure to flare vs human performance in the same conditions; why would one be better than the other.....
The industry should focus more on failures like the AMS 737 accident.
Accident classifications range from CFIT to human error, but it was a component of the auto system which had the 'failure'.
Although this is not a feature of modern designs, excepting for 'grandfather rights' older 737s, the industry still has to be vigilant for those problems not yet identified, operationally or in certification (CRJ accident Sweden, enhanced vision incident Australia).
Basically any failure the overall system should not result in a fatality with a probability something like 10^-9.
These are only numbers, theory, whereas in practice the auto technical reliability is very good, and much better than the human for the majority of landing tasks in the same conditions.
Flare is usually a sub mode of the flight guidance system. In modern high integrity systems lack of mode change or sensor failure can be detected and used to disengage the autopilot or revert to a lower level of operation; the latter being somewhat irrelevant at low altitude - fail op reverts to fail passive, but the aircraft still lands.
In addition many systems use a trim up function which biases the aircraft to pitch nose up if the autopilot disengages at low altitude. There are many reasons for such a system, control torque, minimising GA height loss, and accident severity - probability of fatality.
The need for a 'no flare' call may have been carried over from older systems - simplex monitored or some duplex autopilots with a 'bolt-on' flare mode. Modern systems should not need a call. How would a failure be detected and how could the pilot intervene in time. Older systems suffered more 'failures' due to unwarranted human intervention than there were actual failures.
Those operators who still require a call might consider the type of system they are using, the expectations of crews ability to monitor, and how they might react. If procedures require a call for a mode change, what happens when it doesn't happen. Is the mode change proportional to rad alt, if so why not call altitude - auto call outs do that any way.
And those who use HUD might wish to make a similar evaluation and be prepared to argue the differences between autoflight failure to flare vs human performance in the same conditions; why would one be better than the other.....
The industry should focus more on failures like the AMS 737 accident.
Accident classifications range from CFIT to human error, but it was a component of the auto system which had the 'failure'.
Although this is not a feature of modern designs, excepting for 'grandfather rights' older 737s, the industry still has to be vigilant for those problems not yet identified, operationally or in certification (CRJ accident Sweden, enhanced vision incident Australia).
Join Date: Jan 2014
Location: N5109.2W10.5
Posts: 720
Likes: 0
Received 0 Likes
on
0 Posts
The need for a 'no flare' call may have been carried over from older systems - simplex monitored or some duplex autopilots with a 'bolt-on' flare mode.
We still have a requirement to check the ILS QDM is correct at 350 feet - yet the data is frozen at 700 feet (data lock). So why delay the check so late until it is impossible to correct?
I suspect it is an old rule carried over from the days when we set the ILS QDM on an OBS next to the heading and speed knobs. It was easy to accidentally change the OBS, and the latest we could correct the error was 350 feet agl.
I don't have time to find the reference, but Airbus say somewhere that if "LAND" is displayed in green on the FMA below 400' RA, there is no failure that will prevent the FLARE mode from operating.
Not sure if that is true - they might have been referring to probabilities of failures. This is a tricky area: PF should be eyes outside by then, so s/he doesn't know when the RAD ALT is at approx 40' (when FLARE should normally be displayed), so how does s/he know if FLARE is missing until it is too late ?
At a recent CRM course, a Senior TRE seemed to be telling us that Airbus have not banned a 'NO FLARE' call.
However, as Fursty says; if we are not supposed to say 'NO FLARE', then I think the most sensible call in the absence of FLARE would be to say 'GO-AROUND', (which should be reacted to immediately), There isn't time for much else before hitting the ground.
Not sure if that is true - they might have been referring to probabilities of failures. This is a tricky area: PF should be eyes outside by then, so s/he doesn't know when the RAD ALT is at approx 40' (when FLARE should normally be displayed), so how does s/he know if FLARE is missing until it is too late ?
At a recent CRM course, a Senior TRE seemed to be telling us that Airbus have not banned a 'NO FLARE' call.
However, as Fursty says; if we are not supposed to say 'NO FLARE', then I think the most sensible call in the absence of FLARE would be to say 'GO-AROUND', (which should be reacted to immediately), There isn't time for much else before hitting the ground.
Only half a speed-brake
I'm fine with "NO FLARE" in case it would happen. For CAT II and IIIa the PF is looking out. But can it - without a red AUTOLAND?
Join Date: May 2000
Location: Seattle
Posts: 3,195
Likes: 0
Received 0 Likes
on
0 Posts
It certainly depends on the airplane. The 744 and 748 can now be landed without the FLARE mode being armed. This has evolved through the life of the airplane, from Go-around required, to an advisory "No FLARE" call by the PM, to doing nothing today. Apparently the sink rate is within the design specs for an occasional landing...
@FlightDetent; All Airbus FCOM says about the red Autoland light in relation to the flare, is that it will illuminate if the FMGS detects a "long flare":
FCOM PRO-NOR-SRP-01-70 Warnings for ILS/GLS Approach.
No definition is given about how it decides what a 'long flare' actually is, but it does NOT say that it will illuminate if there is no flare.
FCOM PRO-NOR-SRP-01-70 Warnings for ILS/GLS Approach.
No definition is given about how it decides what a 'long flare' actually is, but it does NOT say that it will illuminate if there is no flare.
Thread Starter
Join Date: Mar 2017
Location: London
Posts: 170
Likes: 0
Received 0 Likes
on
0 Posts
The red autoland light does not illuminate for a no FLARE event.
I'm sure Airbus had a number floating about.
As someone said above, I guess it's never happened and probably never will so that's why there isn't a formal "no FLARE" like there is for "no spoilers, no reverse etc"
I'm sure Airbus had a number floating about.
As someone said above, I guess it's never happened and probably never will so that's why there isn't a formal "no FLARE" like there is for "no spoilers, no reverse etc"
I agree that the risk is tiny if not negligible and most probably the bigger risk will be keen f/os calling no flare just as flare annunciates leading to unnecessary go-arounds.
On B744 only thing you can do (if you catch it in time) is to announce 'no flare'. No action is required as there is no time do anything. Landing is supposed to be a bit firmer but within limits.
At least this is what they told us during training. Never had this in real life.
At least this is what they told us during training. Never had this in real life.
Only half a speed-brake
My connection between no FLARE and red AUTOLAND was a little different, sorry for implying otherwise.
FLARE is a logical state in the FBW computers, a branch in the algorithm. There is no clutch or similar, and the software is deterministic even if our bar stories may suggest different.
Hence my speculation that if there were to be an external problem or internal failure preventing the FLARE from engaging, that abnormality would be discovered by the monitoring part and red AUTOLAND flash straight away.
It is a speculation, but goes the other way: as long as you did not get red AUTOLAND, the FLARE will always engage.
Further to the comments above, we do not train for no ROLLOUT and no THR IDLE either.
FLARE is a logical state in the FBW computers, a branch in the algorithm. There is no clutch or similar, and the software is deterministic even if our bar stories may suggest different.
Hence my speculation that if there were to be an external problem or internal failure preventing the FLARE from engaging, that abnormality would be discovered by the monitoring part and red AUTOLAND flash straight away.
It is a speculation, but goes the other way: as long as you did not get red AUTOLAND, the FLARE will always engage.
Further to the comments above, we do not train for no ROLLOUT and no THR IDLE either.
