LH A320 Rough Landing @ Hamburg

[Dani]

Well, another Airbus-FBW-conspiracy theory. Of course the aircraft has other characteristics in the ground mode. But if you use the wrong landing technique (and wrong runway), you get at your limits, in any aircraft.

As soon as you apply TOGA, the aircraft knows in which mode it is, problem solved.

Dani

[TvB]

It has nothing to do with "conspiracy", dear Dani and others. It is simple that AI's design philosophy has sometimes a few "hidden" (from the pilot and others) features incorporated that make sense under normal operation, but may lead into problems in others.

The remarkable thing is that AI keeps these things as a secret to pilots. I always recall a senior member of the management, telling me in regards to their reduced V1 calculation: "We (AI) did not make any publicity about this fact as we are safety minded and we would like the pilots to ignore that they have a more safety margin."

Well, keep the pilots in the dark is not a good idea, as this case proves again.


So in regards to design of flight controls and protection features I prefer the conventional approach, where I am and I stay in command and the aircraft doesn't "think" for itself. Aircrafts and specially the ones from AI are not so very good thinkers and what is even more a problem: those who designed certain features do not see the entire picture of the operation under sometimes very special conditions (gusts and x -wind just being one). It is not okay, and has never been, when I as PiC have to start thinking in such a special situation how I could get around the computers. There is no time for that, specially not in turbulence or x-wind conditions with gusts.

So I wonder if AI birds now will get a more stringent gust and x-wind limitation. Probably not, and as usual they may shift the responsibility to the pilots...

[CaptainProp]

Well spoken Dani.

5 years of flying airbus and never had any system "hick-ups". Whenever I, or my colleagues at the time, had some problems I always found out in the end that the problem was myself and my inability to program the FMS, manage the FCU etc etc.

Looks to me that the almighty LH pilots/flt department are having some responsibility issues here and are desperately reaching for something to blame....

CP

[falconer1]

surprise, surprise...

WOW circuits can aggravate tight situations very close to the ground in 50+ KTS gusting winds...

"Such a behavior of the machine was not documented in any manual," criticized a pilot to the Spiegel - and it is still not in there, because Airbus keeps things close to their chest until the completion of the investigation. This landing could have ended in a terrible catastrophe.

how about hurricane mods in all new aircraft??????

hope that pilot who gave that interview with "Der Spiegel" is not entrusted in a command position to fly SLFs...

can it get more ridiculous than that????

[CaptainProp]

TvB - I respect your opinion and you preferring "the conventional approach". Having said that, all systems and how they operate are explained in great detail in the FCOM. Problem is, not saying that this applies to you in person, that a lot of pilots are simply too lazy to really dig in to the manuals and really, I mean really, learning how the systems work... I don't know how many times I have heard from colleagues that "ah, that's not something that's important to fly the airplane, just a bunch of engineering BS".

CP

[CaptainProp]

falconer1 - could not agree more.

CP

[Dani]

isn't it amazing that it's always the non-pilots and non-Airbus-pilots blame the craft? Is there an actual rated Airbus pilot out there who has similar allegiations against the system?

Except the LH captain in the incident maybe...

[Cacophonix]

One has to ask how many Airbus aircraft of this type, using this system have landed at, near or over their cross wind limits. Out of these landings how many resulted in one of the following outcomes: had an incident, crashed, landed safely.

I suspect you will find that the negative outcomes constitute an extremely small percentage of the total possible outcomes.

It appears that this landing attempt was made in unfavourable crosswind conditons in an unfavourable runway direction. What is more likey to have caused this incident, some unique Airbus flaw or the other factors pertaining on that day?

Occam might be a good guide here.

[bsieker]

Originally Posted by TvB

It has nothing to do with "conspiracy"

[...] where I am and I stay in command and the aircraft doesn't "think" for itself.

Saying that the aircraft "thinks for itself" is the conspiracy theory.

It doesn't.

Let's looks at some stuff in the FCOM: 3 seconds after main gear touchdown the control configuration in Normal Law blends into ground mode, which is a direct stick-to-surface relation. (Flare mode is standard roll-rate demand.)

So, if anything, during the period in which the Spiegel says "the computer was in command", the pilots were actually in more direct control than before, directly commanding aileron/roll spoiler surface deflection.

I have also looked through the FDR graphs from the prelim report (it is very well hidden on the BFU website, in one of the bulletins: not even its search function turns it up, but google is your friend) and there are only two transient transitions to ground mode, during which aileron deflection does not reduce.

[20milesout]

Final report (English)

Abschlussbericht der BFU (Deutsch)

[XPMorten]

Right, no gust, it was yaw induced roll - as expected

XPM

[Starbear]

20milesout: Thanks for posting the links, a very interesting report.

I am always amazed at what we each seem to take from such reports. XP Morten says for example:

Right, no gust, it was yaw induced roll - as expected

Whereas I got: Flight to Ground Law changes with just one gear on ground; Ground Law limits surface deflection by about 50%; Crosswind limits imprecise; Contradictions in both Airbus manuals and also Company manuals; Airbus write a more precise technique for strong crosswind and much much more

It is probably correct that the approach should have been discontinued earlier (but not mandatory) but that aside this low exeperinced but able young co-pilot did a very good job but used a technique (published) which guaranteed a drift downwind and when it went awry there was insuficient aerodynamic control via ailerons/spoilers (and rudder) to correct the situation quickly.

I hope this post is elevated back to the main rumour/news page as its worthy of a read by all Airbus pilots, especially trainers.

SB

[XPMorten]

Whereas I got: Flight to Ground Law changes with just one gear on ground;

Yaw Induced Roll got them INTO the situation;

"Rudder inputs were in the case at hand of importance because the aircraft manufacturer had determined
during their analysis that the rudder inputs had been too fierce and abrupt."

"The aircraft yawed towards the left, thereby increasing the lift from the right wing and decreasing that
from the left wing. In spite of the co-pilot's right sidestick correction, this resulted in unintended contact
between the downwind main landing gear and the runway."

"After touchdown the aircraft yawed a further 5° to the left."

"It is in the opinion of the aircraft manufacturer, that contact between the left wingtip and runway could
have been avoided if there had been an earlier, coordinated sidestick roll input, combined with less left
rudder deflection"

In other words, Airbus says the pilots created a yaw induced roll by
applying too much rudder and too little aileron BEFORE touchdown
(which will give you a Yaw Induced Roll)

The flightlaw change prevented them from getting OUT of the situation sooner,
but it was not the cause of the incident.

The BETA angle was much higher than it should have been
after touchdown which requires ALOT of rollpower.

XPM

[tildeleb]

Many thanks to 20milesout for posting this. I've been waiting for the final on LH44.

Starbear: I too am amazed at the different take aways from the report.

Section 2.1.1 of the BFU Final Report

"The exemplary account of the aircraft manufacturer that, if the co-pilot's sidestick input had been delayed by 1.5 seconds the result would have been a maximum of 10 deg left wing down -- and would thus have avoided wingtip contact with the ground -- is theoretically understandable. However from a practical point of view of a pilot, this assessment can be queried [questioned]. With the aircraft under manual control, a time period of 1.5 seconds would have been difficult to coordinate precisely. Further it should be noted that in this specific case a reduction of the wind-down angle to a maximum of 10 deg would have prevented wingtip contact with the ground but also that pilots are enjoined to avoid unintended changes in the attitude just prior to touchdown."

From the OM/B and 1.18.2.2 of the BFU Final Report

"In strong crosswind conditions, small amounts of lateral control may be used to maintain the wings level. This lateral stick input must be reduced to zero at first main landing gear touchdown. Even during an approach of considerable turbulence, the control system resists the disturbances quite well without pilot inputs. In fact, the pilot should try to limit his control inputs to those necessary to correct he flight trajectory path and leave the task of countering air disturbances to the flight control system."

From 2.2.1 of the BFU Final Report

"Up until 20 ft above ground the approach was stabil[e] and exhibited aeronautically no reason for a go-around procedure."

From 2.2.3.1 of the BFU Final Report

"The BFU is of the opinion that, the captain as pilot-in-command did not reach his decision [about landing] using thus reasoning, because he did not regard the value maximum crosswind demonstrated for landing as an operational limit for the aircraft."
...
"Given this unusual weather situation, there was no company operational management support for the crew in their decision taking [making] process prior to or during the flight"

"Recommendation 01/2010 of the BFU Report

The aircraft manufacturer should initiate an assessment for the Aircraft Type A320 and Series with the aim that the transition logic Flight Mode/Ground Mode of the lateral control system switches into Ground Mode only if the aircraft is indeed on the ground." ...

My takeaways:

1. Given the wind conditions it isn't clear that anyone should have been taking off or landing that day. A 50% GA rate over the last 10 minutes is very high. If you are going to try to land in a 30 kt crosswind with gusts of 45-55 kt, you are hoping you don't get a black swan. LH management and Hamburg airport management let planes take off and land in conditions which were highly questionable.

There is kind a peer pressure situation where it is hard for a captain to say we are not going to fly when everyone else is. This is why the company needs to step in. The basic approach (no pun intended) seems to be lets fly to the destination, try for landing, and see how it goes.

2. This was complex worst case scenario, that happens only when there are hurricanes or other severe weather depressions present in the area. Wind gusts this high happen only 7 times/yr at this airport on average. In addition the specific wind gusts for this landing were worst case.

3. Airbus had to work very hard to find a way that the landing could be accomplished. Even though they found it, it is completely unclear to me that most pilots could fly it without practicing it first. Airbus doesn't want to have to make another update to 20 years old FCS for the A320 series.

4. If the FCS hadn't transitioned into Ground Law on a 1 wheel touchdown, the max right sidestick inputs that both pilots gave probably would have prevented a wingtip touchdown. In my opinion a fix is called for.

5. Documentation issues with the FM with respect to limits for cross wind landing limits and use of sidestick when landing.

On a personal note. In my opinion I believe the young FO did a great job, given her training and experience level. I believe she was criticized in the German press and the Captain was praised as a hero, which I think was very unfortunate. The Captain made the final decision to land, picked the runway, and declined runway 33, not the FO. I look forward to shaking her hand some day.

Questions:

A. How come the approximately 10 kt gust between just before wing touch which is clearly visible in diag 16 page 41/46 (the vector just before the red box for "wing touch") wasn't mentioned in the report? Didn't this contribute to the roll and wing touch?

B. I would love to know if this approach will be added to the simulator training for all LH/A320 pilots?

C. I would have liked to seen a CVR transcript?

FYI I am not a pilot but I am an engineer.

[rak64]

gusts in storms

1 remark,

usually gusts are in additions to the regular wind, because of its convectional nature.
But in heavy wind condition is normaly no convection available, because the energy getting distributed.
In fact that means wind a number and gusts are negative, pehaps wind 60 knots in gusts 40 knots.
I wasw flying that night, i did a go around a uneventfull landing, than i was watching closly the wind indications.
That was what i discovered.

[tildeleb]

But in heavy wind condition is normaly no convection available, because the energy getting distributed. In fact that means wind a number and gusts are negative, pehaps wind 60 knots in gusts 40 knots.

It's an interesting comment and I appreciate it. A negative gust can really have the same effect as a positive one, there is really no difference. Thanks.

[Meikleour]

Whilst all the issues about control laws etc. have been most interesting - after my 39 years in the airline business why has not more been made of the human factors issues with this incident?

Clearly the captain has at all times a `duty of care` to the passengers which requires a matching of the best skills available on the flight-deck to handling of the more critical situations. Thus, given the experience disparity within the crew, why did the captain even think it was a good idea to have the F/O flying what was obviously a very critical approach. Leave aside company policies of encouraging skill development for F/O `s this has to be seen as a severe lapse in judgement or at worst an overconfident belief in his personal skill to `retrieve` situations?

My second concern was why did the crew not consider the more into wind non-precision runway from the outset? I believe the weather allowed successful approaches to both runways. This is especially relevant given that any gusts would have veered and increased the crosswind component on the ILS runway. Was this point really not considered by the captain?

All in all, to be discussing obscure aspects of the control laws when the incident could so easily have been avoided by the application of common airmanship is premature.

[A-3TWENTY]

Actually , there are two discrepancies in the airbus manuals , and one of them was never mencioned in the final report.

In the FCOM 3.01 -Limitations , it states that the max.demonstrated xwind is 38Kts. Gusts included. This is the value mentioned in the report.

But ..... I f you go to FCOM 2.04.10.Pag11 it says that thE max. x wind component for landing is 33 kts for landing in dry ,damp or wet rwy and also says that Airbus does not recommend to take off or land above this limits.

Some of you will say that this is in the contaminated runway chapter , but since I don`t know a dry contaminated rwy or a wet(less than 3mm) contaminated runway , this is a limit for me.

One other point is:

When Airbus states "max demonstrated" ,they don`t explain why it was the max. value.
Maybe it could be because either they never found an xwind higher than 38 kts, or the test pilot tought wouldn`t be safe beyond that value,or,or, etc. I don`t know and they don`t tell me.

So , my questios is:

Should I make experiences with an airplane full os passengers without the expertise of a test pilot ?
Should I push up an airplane full of pax to the unknown? I think that , as a Captain , I am just a Captain , (not a test pilot), our pax are not test paxs and my work is to take them A to B as safe as posible.

I also think that the companies should state clear limits ,since airbus don`t.

My company has no limits as well , so my limits are the ones in the FCOM 2.04.10 Pag11

A-3TWENTY

[Starbear]

A-3twenty

Actually , there are two discrepancies in the airbus manuals , and one of them was never mencioned in the final report.

In the FCOM 3.01 -Limitations , it states that the max.demonstrated xwind is 38Kts. Gusts included. This is the value mentioned in the report.

I think you will find that this is an amendment, subsequent to and almost certainly because of the Hamburg incident, which occurred two years ago when the other figures applied.

[alf5071h]

Re #509, I would agree that HF is not the main theme of the report, but re-reading it, there is a reasonable balance between the ‘technical’ and human issues. Perhaps the BFU sensed (correctly) that HF dominant reports don’t get much action, particularly in EASA.

The technical issues are relatively insignificant – they represent ‘how the aircraft is’ and that pilots are expected to have the necessary skills to fly it. How and where they get these skills is another issue.
The comments on manufacturer’s manuals may be pedantic; manuals can always be criticized as they represent an author’s viewpoint at the time of writing and may only consider the foreseeable circumstances.
If an item is not in the AFM limitations section then it is not a limit, only advice, but by using the FCOM in lieu of an AFM (EU-OPS) the point is overlooked, – a problem of regulation.

Many aspects of piloting skill have to be assumed, thus they are open to interpretation, judgement, etc; even after a successful check we can still have a bad day. Unfortunately, in the current litigious climate, making assumptions is increasingly unacceptable and most issues to be written down as SOPs etc.

In general, the report avoids judging the Captain’s decision. We must beware of hindsight, there’s often a better way to do something when all of the ‘facts’ are known.
Decisions taken at the time are ‘correct’; they are made on the information available. Errors can only be allocated after the event – the outcome was not as expected. The pilot’s task is to avoid an unexpected outcome by thinking ahead.
In this incident the decison outcome may have resulted from either a poor situation assessment or the choice of inappropriate action, most of which is driven by the depth of knowledge and controlled thinking – the use of knowledge.
The report discusses ‘thinking’ (decision making) and judges the processes as adequate in circumstances. The weakness appears to be in the information which supported the situational assessment, both the physical situation and the background knowledge.

A major problem for the industry is that crosswind operations are poorly defined.
The discussion on maximum demonstrated crosswind (MDCw) perhaps misses some points.
The method of wind measurement for the tests may not be the same as for operations, although in this instance there were. However, when using ATC averaged wind data (over the preceding 2 min and ‘gusts’ recorded as a maximum value in the preceding 10 min), the MDCw value could actually be achieved in relatively benign conditions yet reported as more severe. Landings in each direction and multiple landings should reduce the error.

Also in operations, the reported wind may not represent the actual conditions (most times). Furthermore, gusts are not predictable so perhaps pilots need to allow for this; but the regulations don’t provide any margin.
Thus, perhaps the operator’s limit should include gusts and have a margin for reporting error (10%); so a 38 kt MDCw would be a 34 kt operating limit including gusts.
The incident crew appears to have lacked guidance in these issues, and particularly having made a well considered judgement, the attempted landing was justified.

Whether the First Officer or the Captain should have flown the landing is debatable.
With hindsight (the report), if the Captain has more experience in the conditions and in the aircraft then the safest option would be with him, but then how do FOs ever gain experience.
It could be feasible that a new Captain had never experienced high crosswinds, thus this Captain has to ‘learn’ during the first crosswind encounter – the same situation that the FO was in, but in this instance with overseeing support.
This point is similar to the problems arising from assumptions about skill and piloting capability, and may also argue for a safety margin in the crosswind limit.
Perhaps a lower crosswind limit would have triggered the Captain to consider the other runway, but then if there was a similar result, hindsight bias could have criticized poor CRM, lack of ALAR awareness, choice of NPA, etc.
No plan (or decision) survives contact with the enemy.

The manufacturer can always be cited for mis-judging the ‘average skill’ requirements when certificating crosswind operations.
How is average skill defined? CS 25 defines aircraft design standards, where (how) are the equivalent human standards defined. There is little or no harmonization in worldwide skills training or even in daily operation – we do have bad days; and then we probably all judge ourselves as above average.

I agree that airmanship is critical in these situations, but my definition of airmanship includes knowledge, which in this instance requires understanding of how MDCw is obtained, the lack of margin in MDCw values, the meteorological variability of wind / gusts, and the range of error in reporting wind speed and direction. If it is unrealistic to expect retention / recall of this knowledge then the crew need a safety margin, either self applied or a hard limit. The latter is more practical.

Perhaps this incident is a reflection of the modern industry with increasing litigious and commercially pressurized operations. The more we are driven by these aspects, the more difficult it becomes to provide an equivalent level of safety and to use beneficial human attributes.
As current operations increasingly require airmanship intervention, the more difficult it becomes to gain and apply airmanship due to so-called safety interventions and regulations.
Does our future lay with an automaton, constrained by ‘impossible’ laws and inadequate SOPs, striving to prevent reoccurrence of a minor scrape; or are we to maximize our use of human flexibility and adaptability, both for safety (avoiding the big dings) and maintaining commercial advantage?
I suspect that reality is somewhere in between, but who defines the compromise?