A340 of Iberia skids off runway in Quito
The accident at Warsaw prompted a design change, namely to deploy spoilers partially with only one MLG on the ground to dump some lift and assist in getting the second MLG down as soon as possible to get Full Ground Spoilers and Reverse Thrust.
Relevant only to the 320 or all FBW Airbuses ?
Join Date: Dec 2007
Location: Over the clouds
Age: 65
Posts: 256
Likes: 0
Received 0 Likes
on
0 Posts
IMMO the accident begun, when IB decided to disregard the company pilots continuous recommendations to maintain the 343 and not to fly the 346 to Quito. First incident ago/07, 4 right MLG tire burst. More pilot recommendations to the company. The relationship between the company and the pilots was very hard at that moment. The company treated all that recommendations as another pilots stupid demand. Behind the scene; the typical chain of human, environment, mechanical factors and the not so typical Quito 
Tailwinds
Tailwinds
Aircraft on the ground?
Quote from atakacs:
I am still believing that a "panic stop" button (or more to the point some sort of procedure to get maximal breaking performance from the automation systems regardless of what they "think" of the pilot intentions or actual plane condition) could be a welcome enhancement.
Granted such a system would generate it's own problems (such as possible unwanted activation)…
[Unquote]
CONF iture is right to point out that, on touchdown − having belatedly recognised that the “numbers” are not right and knowing LDA to be marginal − a crew should give serious consideration to a go-around. We are speaking in general terms, of course. The pilots may be so shocked by the severity of the impact that they delay the decision. Alternatively, they may rightly decide that the aeroplane is crippled.
So the issue of whether the crew should have the ability to override the automatics with a “panic-stop” button, as atakacs proposes, is a legitimate one. His own misgivings about inappropriate (or even unintentional) deployment are evident, however. The nearest example that comes to mind is that of a VC10 in 1972, in which the (handling) captain’s attempt to perform a go-around was thwarted by the copilot’s selection of spoilers and reverse.
There might be a better solution to the Quito-type gear-damage scenario. On Airbuses, the FACs (Flight Augmentation Computers) already seem to have a good understanding of how the thing flies. [They are capable, for example, of calculating current gross weight in the air.] No conventional aeroplane, I suggest, can maintain 1-G flight at a low airspeed with low alpha (angle of attack).
The combination of the above 3 conditions for more than a moment can only mean one thing: aircraft on the ground.
I am still believing that a "panic stop" button (or more to the point some sort of procedure to get maximal breaking performance from the automation systems regardless of what they "think" of the pilot intentions or actual plane condition) could be a welcome enhancement.
Granted such a system would generate it's own problems (such as possible unwanted activation)…
[Unquote]
CONF iture is right to point out that, on touchdown − having belatedly recognised that the “numbers” are not right and knowing LDA to be marginal − a crew should give serious consideration to a go-around. We are speaking in general terms, of course. The pilots may be so shocked by the severity of the impact that they delay the decision. Alternatively, they may rightly decide that the aeroplane is crippled.
So the issue of whether the crew should have the ability to override the automatics with a “panic-stop” button, as atakacs proposes, is a legitimate one. His own misgivings about inappropriate (or even unintentional) deployment are evident, however. The nearest example that comes to mind is that of a VC10 in 1972, in which the (handling) captain’s attempt to perform a go-around was thwarted by the copilot’s selection of spoilers and reverse.
There might be a better solution to the Quito-type gear-damage scenario. On Airbuses, the FACs (Flight Augmentation Computers) already seem to have a good understanding of how the thing flies. [They are capable, for example, of calculating current gross weight in the air.] No conventional aeroplane, I suggest, can maintain 1-G flight at a low airspeed with low alpha (angle of attack).
The combination of the above 3 conditions for more than a moment can only mean one thing: aircraft on the ground.
Last edited by Chris Scott; 9th May 2008 at 00:04. Reason: Last paragraph improved
Join Date: Aug 2005
Location: Around the lamp
Posts: 66
Likes: 0
Received 0 Likes
on
0 Posts
Hi ya.
For those with spanish knowledge, the equadorian report about the A346 accident is going to be an interesting read:
http://www.dgac.gov.ec/Espa%C3%B1ol/...40-600%202.pdf
Any english translation of this report known?
Regards,
Christian
For those with spanish knowledge, the equadorian report about the A346 accident is going to be an interesting read:
http://www.dgac.gov.ec/Espa%C3%B1ol/...40-600%202.pdf
Any english translation of this report known?
Regards,
Christian
Join Date: Aug 1999
Location: Dublin, Ireland
Posts: 1,879
Likes: 0
Received 0 Likes
on
0 Posts
The flight data recorder showed, that the airplane was touching down at a calibrated airspeed of 153 knots, an actual ground speed of 188 knots at a vertical speed of more than 1100 feet per minute resulting in a vertical acceleration of 3.09G. The airplane had been doing a 960 feet per minute sinkrate down to about 9700 feet (500 feet AGL) at a calibrated airspeed of 153 knots and 188 knots above ground. Then a pitch down command occured resulting in about 1500 feet per minute sinkrate descending through 400 feet AGL down to 200 feet AGL. At 130 feet AGL the sinkrate had reduced to 992 feet per minute, but increased again.
Although I understand that the pilots should have aborted way before I'm not too sure to understand as why they hit so hard ? When you have a 1100 ft/m sinkrate why on earth would you add a pitch down command ?
Join Date: Feb 2005
Location: flyover country USA
Age: 82
Posts: 4,579
Likes: 0
Received 0 Likes
on
0 Posts
Upon touch down the spoilers deployed automatically, the tyres 3 and 8 blew, wiring harnesses of both main gear struts broke leaving the engines in approach idle.
Join Date: May 2004
Location: 3.5 from TD
Age: 47
Posts: 1,042
Likes: 0
Received 0 Likes
on
0 Posts
Although I understand that the pilots should have aborted way before I'm not too sure to understand as why they hit so hard ? When you have a 1100 ft/m sinkrate why on earth would you add a pitch down command ?
The combination of high elevation and high ground speed means your margin of error is much reduced. And I believe that SEQU's GS does not bring you to the ground, so there you have another threat which contributed.
As far as hitting the ground so hard, you have to take into account inertia. I'm not a bus pilot, but in the 747, when you get close enough to the ground, you run out of room to correct your sink rate. So if you incorrectly push slightly down passing 200' agl to correct for being slightly high, you have pretty much sealed your fate to a hard landing. I suspect the A346 is probably similar in this regard.
This height is also increased if you are flying a steeper than 3 degree slope, which these guys probably were doing. The control inputs this guy used probably worked ok down low, but they didn't up at 9000 feet.
Join Date: Sep 2007
Location: Sweden
Age: 62
Posts: 10
Likes: 0
Received 0 Likes
on
0 Posts
maybe this is the magic button people are asking for?
Airbus reveals A380-linked pilot systems secrets
seems pretty clever to me. Still to see that it does not create new confused situations. Would be interesting to know if it handles the type of approach that was made Quito. I guess it does as the case is known after this W/O...
BR
FH
Airbus reveals A380-linked pilot systems secrets
seems pretty clever to me. Still to see that it does not create new confused situations. Would be interesting to know if it handles the type of approach that was made Quito. I guess it does as the case is known after this W/O...
BR
FH
Good system under normal ops but it still presumably relies on what the computers are thinking. I think the idea posted earlier about a panic stop button which overrides the computers and throws out everything possible to stop the plane quickly (regardless of which mode the computer thinks it is in) would be a good idea - perhaps and extra button on the Autobrake system marked "EMG".
Join Date: Apr 2008
Location: HK
Posts: 513
Likes: 0
Received 0 Likes
on
0 Posts
All this talk of "lets have a big panic button that will stop the aircraft no matter what". So what happens when the big button info gets corrupted or someone hits it by mistake?
People - the aircraft had basically "crashed" when it hit the runway! Big panic buttons would not have achieved anything.
Join Date: Jun 2009
Location: Georgia
Posts: 169
Likes: 0
Received 0 Likes
on
0 Posts
@femented herring
re: your link...
Anybody else (other than me) that thinks depending on an outside signal like GPS is a bad idea?
The two fundamental data inputs on which both systems work is the aircraft's constantly updated real-time GPS position, and a terrain database that has precise runway location and dimensions embedded in it, along with full airport details.
People - the aircraft had basically "crashed" when it hit the runway! Big panic buttons would not have achieved anything.
Join Date: Dec 2009
Location: TWIMyOyster
Posts: 14
Likes: 0
Received 0 Likes
on
0 Posts
Pennies...
Having flown into UIO quite a few times (albeit other Company and Plane-about same size, forgive me if aircraft specifics are misinterpreted), tailwind at UIO35 is not unusual. Besides, the alternative –a circling (actually visual) for 17- was no option due to the weather (vis 3000 meters).
One of the things that might have been considered in the pre-landing phase is the braking action at UIO when experiencing wet runway conditions. (A.o. due to the high GS at UIO) a lot of rubber deposit can be found in the touchdown zone’s; extended touchdown zone’s due to the normal & displaced threshold 35, and the frequently inaccurate touchdown positions (both 17 and 35). When was the last time the RWY was grooved (report does not say)? If not recent the effective braking surface will probably not be very long (a few hundred meters in the middle). Combined with possibly remaining oil deposits on the RWY this can be quite hairy at UIO. Since friction meters in cars are not reliable under these conditions, the only thing you have is pilot reports and experience. Braking action is hardly an exact science, however in this case .. ‘poor’ (mu≈0.2 or lower) is far lower than normal mu≈0.4 (ba=’good’) for ‘normal’ wet runways (as used in the landing distance calculations). Reported ba ‘poor’ might have to generate even greater warning flags with the cockpit crew than it did, especially if you have been there a few times.
The proper -2020 hindsight- alternative for the crew in the pre-landing phase would have been to divert. However, that is not a very popular concept–company, passengers, attendants- if the data indicates a safe landing can be made and pilot reports are often inaccurate. I understand the decision to start the approach for UIO-35.
The UIO ILS/Visual 35 approach itself is challenging, especially with vis 3000 meters. Due to the terrain (hill) in front of the RWY, the GS is (500 meters) displaced leaving insufficient LDA for large aircraft. Therefore a PAPI is placed at the normal landing position, which flightpath you will pick up after passing the hill (with a VOR-QMS conveniently placed on top). With 3000 meters of visibility you will have the Runway on the ILS at approx DA. Normally –in good conditions- you will leave the ILS-GP to intercept the visual GP at or around QMS (about 4-4.5 NM out), now as I understand from the report –due vis- at DA (about 1.6 NM out). This, in order to pick up the PAPI from this position, requires an impressive sinkrate (more than 1500’/min), while being stabilized at 500’ GND has become impossible (more likely at 200’ at the earliest). Quite likely some interesting sinkrate warnings have added to the party. With that one may question this DA for larger aircraft like the A346; fine for the small aircraft, but far too low for these aircraft/procedure. No company advise had apparently been given on this situation.
UIO is famous for its hard landings: high groundspeed; challenging ILS/Visual approach procedure; high elevation (flare slightly higher); and last but not least the upslope of RWY 35. Even in perfect weather conditions a challenge. So, from this fairly unstabilized position in –for this airport- very limited visibilities it is now time to flare. This pilot knew he only had very few (136m as they later determined) meters of ‘extra’ runway on top of the certified touchdown point; not a lot and far less than the normal touchdown-zone. So, he minimized the flare. UIO RWY 35 unfortunately has a significant upslope in the touchdown zone; total upslope of RWY 35 may be limited (0.4%) but due to the “bump” in the runway actual upslope may be as much as 0.8% (if I recall correctly) in the touchdown zone. Add that to the 1100’/min VS (not abnormal for the 3 deg final app itself) upon touchdown, and one might give the A346 a bit of credit for not breaking up altogether. But then again, should an aircraft brake down from a “unflared” landing from a 3 degrees flightpath? Did you never “forget” to flare or flared “too late” (mostly aggravating the situation since the main gear is aft of the CG). I did .. a few times .. fortunately not at UIO (knock on wood).
Normally the bulk of the braking action comes from the wheelbrakes, however with braking action ‘poor’ (contaminated) the speedbrakes/liftdumpers (WOW) and reversers become ever more important. Autobrakes indicated defective upon the hard landing causing wheelbraking to be delayed by 3 seconds and eating up valuable runway before manual braking started (especially at UIO- groundspeeds). Upon landing deployment of the reversers was blocked by computer logic (air-ground sensors); there goes a major portion (up to 30-35%) of your total braking action, especially at high speeds. Add to this not only ground idle forward thrust, but approach idle and your braking action is further reduced. From the moment these systems failed this aircraft was doomed. Why did these systems fail; what happens with an RTO due to a tire failure during take off? Airbus computers help right upto the point that you really need them? Food for thought.
Yes, a lousy landing, but that's about it.
My two pennies worth...
One of the things that might have been considered in the pre-landing phase is the braking action at UIO when experiencing wet runway conditions. (A.o. due to the high GS at UIO) a lot of rubber deposit can be found in the touchdown zone’s; extended touchdown zone’s due to the normal & displaced threshold 35, and the frequently inaccurate touchdown positions (both 17 and 35). When was the last time the RWY was grooved (report does not say)? If not recent the effective braking surface will probably not be very long (a few hundred meters in the middle). Combined with possibly remaining oil deposits on the RWY this can be quite hairy at UIO. Since friction meters in cars are not reliable under these conditions, the only thing you have is pilot reports and experience. Braking action is hardly an exact science, however in this case .. ‘poor’ (mu≈0.2 or lower) is far lower than normal mu≈0.4 (ba=’good’) for ‘normal’ wet runways (as used in the landing distance calculations). Reported ba ‘poor’ might have to generate even greater warning flags with the cockpit crew than it did, especially if you have been there a few times.
The proper -2020 hindsight- alternative for the crew in the pre-landing phase would have been to divert. However, that is not a very popular concept–company, passengers, attendants- if the data indicates a safe landing can be made and pilot reports are often inaccurate. I understand the decision to start the approach for UIO-35.
The UIO ILS/Visual 35 approach itself is challenging, especially with vis 3000 meters. Due to the terrain (hill) in front of the RWY, the GS is (500 meters) displaced leaving insufficient LDA for large aircraft. Therefore a PAPI is placed at the normal landing position, which flightpath you will pick up after passing the hill (with a VOR-QMS conveniently placed on top). With 3000 meters of visibility you will have the Runway on the ILS at approx DA. Normally –in good conditions- you will leave the ILS-GP to intercept the visual GP at or around QMS (about 4-4.5 NM out), now as I understand from the report –due vis- at DA (about 1.6 NM out). This, in order to pick up the PAPI from this position, requires an impressive sinkrate (more than 1500’/min), while being stabilized at 500’ GND has become impossible (more likely at 200’ at the earliest). Quite likely some interesting sinkrate warnings have added to the party. With that one may question this DA for larger aircraft like the A346; fine for the small aircraft, but far too low for these aircraft/procedure. No company advise had apparently been given on this situation.
UIO is famous for its hard landings: high groundspeed; challenging ILS/Visual approach procedure; high elevation (flare slightly higher); and last but not least the upslope of RWY 35. Even in perfect weather conditions a challenge. So, from this fairly unstabilized position in –for this airport- very limited visibilities it is now time to flare. This pilot knew he only had very few (136m as they later determined) meters of ‘extra’ runway on top of the certified touchdown point; not a lot and far less than the normal touchdown-zone. So, he minimized the flare. UIO RWY 35 unfortunately has a significant upslope in the touchdown zone; total upslope of RWY 35 may be limited (0.4%) but due to the “bump” in the runway actual upslope may be as much as 0.8% (if I recall correctly) in the touchdown zone. Add that to the 1100’/min VS (not abnormal for the 3 deg final app itself) upon touchdown, and one might give the A346 a bit of credit for not breaking up altogether. But then again, should an aircraft brake down from a “unflared” landing from a 3 degrees flightpath? Did you never “forget” to flare or flared “too late” (mostly aggravating the situation since the main gear is aft of the CG). I did .. a few times .. fortunately not at UIO (knock on wood).
Normally the bulk of the braking action comes from the wheelbrakes, however with braking action ‘poor’ (contaminated) the speedbrakes/liftdumpers (WOW) and reversers become ever more important. Autobrakes indicated defective upon the hard landing causing wheelbraking to be delayed by 3 seconds and eating up valuable runway before manual braking started (especially at UIO- groundspeeds). Upon landing deployment of the reversers was blocked by computer logic (air-ground sensors); there goes a major portion (up to 30-35%) of your total braking action, especially at high speeds. Add to this not only ground idle forward thrust, but approach idle and your braking action is further reduced. From the moment these systems failed this aircraft was doomed. Why did these systems fail; what happens with an RTO due to a tire failure during take off? Airbus computers help right upto the point that you really need them? Food for thought.
Yes, a lousy landing, but that's about it.
My two pennies worth...
Join Date: Dec 2009
Location: TWIMyOyster
Posts: 14
Likes: 0
Received 0 Likes
on
0 Posts
Re-word (attempt)
I will try again on the important issues ... forgive me for not being fluent in English
1. Wet RWY is often extremely slippery at UIO, worse than the Landing distance calculations assume (mu-story), due heavy and extensive rubber deposits (often in combination with oil after a period of dry weather). Some companies have refused to fly into UIO unless the RWY was going to be grooved (after experiencing a few scary moments; it is no pleasure having your cockpit hanging over the end lights after landing). When was the last time the RWY was grooved (report does not say)? Was a “slippery when wet” NOTAM issued/warranted (and what is the company advise i.c.o.)? IMHO normal calculations for wet runways at UIO tend to be overly optimistic and incorrectly pretend it is an exact science (hence the “136 meters” in the report); it is not.
2. The published Decision Altitude is not realistic for aircraft like the A346; due to the construction of the approach (forced and necessary “duck under” to pick up the PAPI due to the displaced glide slope antenna – see approach plate) it is practically impossible to arrive over the threshold in a stabilized way from your DA. The captain managed -where many would probably have not– however resulting in an unstabilized approach. Why does the state allow unrealistic minima for Cat-C/D aircraft; why did the company not impose higher limits or at least put out a warning (even after several warnings of its crews, other companies did). The crew may be to blame, but where do other parties fit in?
3. Although the aircraft landed hard and some tires did burst, the aircraft was basically still intact. The problem was that the air-ground logic malfunctioned and from that moment on the airplane was going to crash – no escape possible. Approach (!) idle forward thrust i.s.o. reverse thrust (could not get it into reverse due failed air-ground logic) does not help the already poor braking action of the wheelbrakes (incl. the burst tires). Basically the increased forward thrust negated most of the braking of the wheels. Failure of the air-ground logic was a major contributing factor to the crash. Why does air-ground logic fail after a hard landing/tire burst? Should this not have been designed with a backup mode? What in case of an RTO due to a tire failure? Airbus/Certification authorities?
4. Only after 3 seconds the crew started manual braking, traveling over 300 meters before any braking occurred. Assuming the hard landing “warranted” an autobrake failure, how many seconds does it take for the crew to react? I don’t think 3 seconds is very long (especially after a hard landing in the given conditions), but where does it fit into your calculations? I have seen autobrakes fail upon touchdown for far less a reason, but is using a –potentially failing- autobrake system in such a critical situation the best option (given the extra time and landing distance i.c.o. failure)? Should extra time be added for using the autobrake system: for the failure to be displayed, for the crew to recognize this failure and for the crew to react (a bit like the V1-RTO times). 136 meters gives you only 1.3 seconds, not 3 seconds; 1.3 seconds is probably not even enough for the warning to appear. Certification authorities, (company) safety margins?
The report does not address these (political/sensitive) issues.
Hope I made myself clearer.
1. Wet RWY is often extremely slippery at UIO, worse than the Landing distance calculations assume (mu-story), due heavy and extensive rubber deposits (often in combination with oil after a period of dry weather). Some companies have refused to fly into UIO unless the RWY was going to be grooved (after experiencing a few scary moments; it is no pleasure having your cockpit hanging over the end lights after landing). When was the last time the RWY was grooved (report does not say)? Was a “slippery when wet” NOTAM issued/warranted (and what is the company advise i.c.o.)? IMHO normal calculations for wet runways at UIO tend to be overly optimistic and incorrectly pretend it is an exact science (hence the “136 meters” in the report); it is not.
2. The published Decision Altitude is not realistic for aircraft like the A346; due to the construction of the approach (forced and necessary “duck under” to pick up the PAPI due to the displaced glide slope antenna – see approach plate) it is practically impossible to arrive over the threshold in a stabilized way from your DA. The captain managed -where many would probably have not– however resulting in an unstabilized approach. Why does the state allow unrealistic minima for Cat-C/D aircraft; why did the company not impose higher limits or at least put out a warning (even after several warnings of its crews, other companies did). The crew may be to blame, but where do other parties fit in?
3. Although the aircraft landed hard and some tires did burst, the aircraft was basically still intact. The problem was that the air-ground logic malfunctioned and from that moment on the airplane was going to crash – no escape possible. Approach (!) idle forward thrust i.s.o. reverse thrust (could not get it into reverse due failed air-ground logic) does not help the already poor braking action of the wheelbrakes (incl. the burst tires). Basically the increased forward thrust negated most of the braking of the wheels. Failure of the air-ground logic was a major contributing factor to the crash. Why does air-ground logic fail after a hard landing/tire burst? Should this not have been designed with a backup mode? What in case of an RTO due to a tire failure? Airbus/Certification authorities?
4. Only after 3 seconds the crew started manual braking, traveling over 300 meters before any braking occurred. Assuming the hard landing “warranted” an autobrake failure, how many seconds does it take for the crew to react? I don’t think 3 seconds is very long (especially after a hard landing in the given conditions), but where does it fit into your calculations? I have seen autobrakes fail upon touchdown for far less a reason, but is using a –potentially failing- autobrake system in such a critical situation the best option (given the extra time and landing distance i.c.o. failure)? Should extra time be added for using the autobrake system: for the failure to be displayed, for the crew to recognize this failure and for the crew to react (a bit like the V1-RTO times). 136 meters gives you only 1.3 seconds, not 3 seconds; 1.3 seconds is probably not even enough for the warning to appear. Certification authorities, (company) safety margins?
The report does not address these (political/sensitive) issues.
Hope I made myself clearer.
