Sully's Flare on the Hudson: Airbus Phugoid Feedback

[albatross]

Probably not relevant here but
Years ago 1988 there a crash of an airbus at an airshow in France (flew into trees off the end of the runway) where the automation was called into question.

https://youtu.be/I9gELPxPG8Q

https://en.m.wikipedia.org/wiki/Air_France_Flight_296

[Amadis of Gaul]

You don't say?

[albatross]

Well I only mentioned it because as I recall the Captain maintained at one point that the aircraft automation would not let him pull the nose up any further ..

[vilas]

CONF iture

where is it that quote from Sully :
"I choose to start the APU to remain in Norma law" ?

Now that I provided NTSB statement in this regard how about you providing Sully's statement, not your opinion about

Sully did not start the APU to remain in normal law, but to secure some redundancy for electric and hydraulic, and have a chance to achieve an engine restart.?

I suggest you read about G-YMMM B777 London crash.The last moments of Hudson landing are very similar to this B777 crash where ice crystals restricted thrust in last 500ft of the approach. There speed dropped to stretch the glide. As the B777 reached the stall warning speed when the pilot pushed the control forward it landed short at VS of 1400ft/mt. As compared to this at Hudson airbus touched down at 750ft/mt. because the normal law alpha prot prevented further drop in speed as would have happened in direct law.

[vmandr]

@QuagmireAirlines

why the NTSB didn't cite ...

I take what government agencies publish with a pinch of salt.
Their Legal affairs ensure (most of the time) there is no liabilities,
conflict of interests etc. from those publications.
after all, you would not bite the hand that feeds you right?

@albatros

the aircraft automation would not let him

Yes. They train you to fly the plane and do not train you to fly ... the software flight=dynamic, software=static/canned, go figure...
(I am sure some 'tourist', will hop-in shortly to correct me )

[KayPam]

Originally Posted by albatross

Well I only mentioned it because as I recall the Captain maintained at one point that the aircraft automation would not let him pull the nose up any further ..

This is very relevant.
If one reads the complete report of the french BEA, there are clear inconsistencies.

Inconsistencies that are very clear, even for a non-specialist. I just had to put together the inconsistent part, found within the whole document.

I pointed them out in a document that I posted on a french forum.
I could provide a quick translation.
There are strong suspicions of the French authorities camouflaging a flight control law problem on the A320, in order not to compromise its commercial carreer. In which they succeeded.

Is anyone interested in looking into my findings ?


What is the matter at hand with this Sully flare ?
Was the AOA (about 14°) at touchdown inferior to what it should have been (alpha max for CONF 2, whose value I don't know by heart) ??

[vmandr]

@Κaypam

the complete report of the french BEA, there are clear inconsistencies.

thanks for meeting my point and I do not doubt your findings.
as they say 'you don't bite the hand that feeds you'
meaning, if they want they can really be precise and consistent
but in doing so, there will be other - political / commercial - consequences
regarding certification, oversight etc. So the lesser of two evils....

as for 'the matter at hand' to me is clear some are trying to discredit the man. a shame !

[QuagmireAirlines]

Originally Posted by KayPam

What is the matter at hand with this Sully flare ?
Was the AOA (about 14°) at touchdown inferior to what it should have been (alpha max for CONF 2, whose value I don't know by heart) ??

I think Sully should have been able to get that very last 3.5 degrees (stall margin) of pitch he was asking for (by holding full aft stick).
I have no problem with the concept of alpha-protection mode, yet the execution of it below 100' AGL (radar altimeter) should not have included phugoid-dampening feedback terms, the items to blame here, and is what kept Sully from pitching a tad more in flare.
Keep in mind Sully hit the water at 9.5 degrees pitch, and 11 would have been considered ideal, so he was only off by a little, and that last 2 or 3 degrees of pitch denied him would have lowered his touchdown impact (feet/min vertical speed when hitting).

Originally Posted by albatross

Probably not relevant here but
Years ago 1988 there a crash of an airbus at an airshow in France (flew into trees off the end of the runway) where the automation was called into question.
https://youtu.be/I9gELPxPG8Q ..https://en.m.wikipedia.org/wiki/Air_France_Flight_296

I thought of that one too this morning. I'd almost forgotten about it. I can remember the old 1988 video of it at the Paris Air Show. That should have been a warning to Airbus to examine what was going on at very very low altitudes (100 feet and below) to allow the pilot to really use every last stall margin possible.

Originally Posted by misd-agin

He didn't read the airspeed "just fine." He misread it by more than 15-19 kts ("safely above Vls" vs 15-19 kts below Vls) and didn't process that his speed was barely above the red band which is an image we never see in normal approaches. That is what extreme stress will do to you and is understandable. The NTSB reports covers the stress impact on human performance.

OK, agreeing he was too low on airspeed during the descent. I'm a flight control engineer, which makes me conclude I'm just glad he was flying and not stalling! Sure an optimal speed which happens to give max(L/D) at whatever weight they had would be great, giving more pitch room to flare too. Yet, getting airspeed as low as possible is laudable, which he did, and he did have 3.5 degrees of actual stall margin left, so did he really mess it up that bad????

Originally Posted by misd-agin

Without the AOA protection AB built into the FBW we can only speculate what would have happened without AOA protection. At a high AOA, reached at 150', he then attempted to raise the nose at 100'. That attempt, with only a couple of knots above stall speed, might have lead to a situation where the vertical descent rate would have been even higher at water entry. That's if he realized it and didn't stall the a/c. If he had stalled a non FBW protected a/c the outcome would have been worse.

As I've said, alpha-protection is a good thing above 100 feet, just not at 100' AGL and below when the phugoid damping feedback was preventing full use of max lift from flare to impact.

Originally Posted by KayPam

This is very relevant.
I pointed them out in a document that I posted on a french forum.
I could provide a quick translation.
Is anyone interested in looking into my findings ?

I'd look into it, from a flight control engineer's point of view, not a pilot. Although I've flown gliders, and, coincidentally, that time Sully was a glider pilot too...
Can you post an English translation in a pdf file or something?

[KayPam]

Having had a look at the flight data plots, it is quite obvious Sully wanted to flare more.

However, the flight control laws having obviously not designed to accomodate an engine-out, flaps 2, gear-up water landing, they were not optimized for this case.

So it is very possible that direct law in this kind of context would allow for better landing than normal law - that is if the pilot is competent, of course.

I don't see how you would discredit Sully. What more should he have done ? A smoother landing ? Let us laugh.

But hey guess what, it so happens that the Habsheim pilots who crashed their airplane in a forest were in conf 2 as well ! So we can just take the alpha max value from the report : 17.5°
But alpha in our case (Sully) was about 14°, whereas full back stick order had been maintained for 2 seconds.
Just as a reminder, 14.5° is alpha prot (still from the report) and with stick neutral in alpha protection, alpha prot is targeted.
So basically, Sully had been pulling for 8 seconds but that did not give him any more than alpha prot. An AOA that he had at the beginning of his nose up order.
An aircraft not responding to your flare order, even if it is at a low speed, is obviously a problem.*
I will investigate this a bit more, later.

However, it should be noted that had the aircraft responded to the nose up order, it would have impacted the water with a higher pitch up angle and a lower vertical speed. Nobody can say for sure if the damages would have been worse or not.

[Winnerhofer]

Bilbao: With No Flare For It!

As the alpha-protection was triggered during this event, the system commanded a nose down signal, which was performed, even though both pilots had their sticks full backward, commanding a “climb”.

[QuagmireAirlines]

Originally Posted by KayPam

However, it should be noted that had the aircraft responded to the nose up order, it would have impacted the water with a higher pitch up angle and a lower vertical speed. Nobody can say for sure if the damages would have been worse or not.

There was an A320 water ditching study done (I lost the link for it, I'll see if I can find it again later) that said 10 to 12 degrees pitch angle is OK. (Sully hit at 9.5 deg pitch, so not bad, only a tad low.) Airbus also recommended strongly that you hit the water with a pitch angle of 11 degrees. This has been looked at, and I think you essentially want to do a 3-point landing on the 2 engine nacelles and tip of the tail at the same time.

If Sully could have used that last bit of 3.5 deg stall margin (pitch up) that the Airbus control laws denied him during the last 50 feet or so, he would have hit at about 12 degrees, with a lower vertical speed, all better.

To me the perfect landing would have been 11 degrees pitch, 1 degree stall margin remaining, and the slow airspeed that corresponds to, while hitting the water at 500 ft/min. Let's call perfect an "11/1/500". Instead, Sully did it as an "9.5/3.5/750" and I blame that on some unwanted feedback terms in alpha-protect below 50' AGL.

[QuagmireAirlines]

Originally Posted by albatross

Well I only mentioned it because as I recall the Captain maintained at one point that the aircraft automation would not let him pull the nose up any further ..

About that 1988 Paris Air Show accident, did the pilot have any stall margin remaining? I haven't read anything about that accident in a long while. I was in the middle of control law development for some MD-11 modes at the time it happened.

[CONF iture]

Originally Posted by vilas

Now that I provided NTSB statement in this regard how about you providing Sully's statement, not your opinion about

My opinion is not different from any pilot who lose one or both engines on a twin, starting the APU to ensure a primary source for electrical power - Not only it is a request from the checklist, but it is just common sense. Don't tell me now that a guy a 737 does it to maintain the flight envelope protections ...

[vilas]

CONF iture
You Herd the NTSB statement so it's not a joke. It was NTSB"s opinion based on surely some corroborating evidence. I didn't say that was the only purpose. Your opinions generally contain ample amount of visceral hatred of some aircraft or equipment so they are not necessarily like any other pilot. So we leave it at that. Now tell me your opinion about the second part, the comparison of dropping speed in the London B777 accident without protection and the result.

[KayPam]

Originally Posted by QuagmireAirlines

There was an A320 water ditching study done (I lost the link for it, I'll see if I can find it again later) that said 10 to 12 degrees pitch angle is OK. (Sully hit at 9.5 deg pitch, so not bad, only a tad low.) Airbus also recommended strongly that you hit the water with a pitch angle of 11 degrees. This has been looked at, and I think you essentially want to do a 3-point landing on the 2 engine nacelles and tip of the tail at the same time.

If Sully could have used that last bit of 3.5 deg stall margin (pitch up) that the Airbus control laws denied him during the last 50 feet or so, he would have hit at about 12 degrees, with a lower vertical speed, all better.

To me the perfect landing would have been 11 degrees pitch, 1 degree stall margin remaining, and the slow airspeed that corresponds to, while hitting the water at 500 ft/min. Let's call perfect an "11/1/500". Instead, Sully did it as an "9.5/3.5/750" and I blame that on some unwanted feedback terms in alpha-protect below 50' AGL.

Okay there has been some studying and simulation after the cactus 1549 accident-landing.

(But at the time of this water landing, these studies had not been done)

Based on your info that I was not aware of :
Alpha max is 17.5°, Sully touched with 9.5° of pitch and 750fpm which gives -3.5° of flight path angle (this gives an AOA of 13° instead of 14°, so this calculation is precise within 1°)
If there had been an AoA increase as demanded by Sully : full back stick and 17.5° of AoA : we would have had something between -2 and 0° of flight path angle, so pitch angle would have been somewhere between 15.5 and 17.5° : way too much (even with the 1° error margin, 14.5 is still 2.5° above the 12° limit you quoted)

But on the other hand, maybe Sully was instinctively targeting 11-12° of pitch, and was pulling full back only because the aircraft was not responding.

So you can see that the situation is not very straightforward and yes, the optimum pitch angle for ditching has been studied but it's very hard to say what the situation would have been with a direct law instead of a normal law.

Originally Posted by QuagmireAirlines

About that 1988 Paris Air Show accident, did the pilot have any stall margin remaining? I haven't read anything about that accident in a long while. I was in the middle of control law development for some MD-11 modes at the time it happened.

Yes, that's an inconsistency of the report itself !
Alpha max is quoted at -17.5°. The report itself says that if the pilot pulls full back then they will get alpha max. The report says that the pilot pulled full back.
But the report also shows, in the flight data tables, that aoa never went above 15°.

So there was a +2.5° margin relative to alpha max !
That's a problem in itself.

I believe that had the aircraft responded correctly to the full back stick input, maybe the gears would have caught some leaves but the aircraft would have had a much better chance of going around "correctly" and remain in flight.
Because there was a slight reserve of speed but that could have been enough to go above the trees, and once the aircraft was above the trees, the engines had spooled up.

Any one with a bit of physics knowledge could go into the flight data tables and read the values. Then :
- compute V alpha max
- compare V alpha max with the speed before impact
- conclude if this speed difference converted into height could have been enough to go around the trees. (based on approximate estimation of the fuselage's height under the tree line)

These reports are very carefully written, toughly proofread and thoroughly revised. So any inconsistency can be regarded as highly suspicious.
Best case scenario explanation : it was written and proofread in a hurry.

[Amadis of Gaul]

I'm starting to get the sense that the whole purpose of this thread is to make sure that as many people as possible are aware that the OP knows what a phugoid is.

[PDR1]

...or perhaps doesn't.

But in relating the two accidents I'd suggest people are conflating two unrelated issues.

In the Sully case we're talking about a flare - converting airspeed into a brief burst of lift to reduce the vertical and horizontal speed on "touchdown". It may well be true that taking a further couple of degrees of AoA could have produced a different combination of horizontal/vertical velocity and attitude at touchdown. It is by no means certain that this combination would have been any better or worse than the combination that he actually used. Given that the combination used resulted in everyone getting off the aeroplane it's hard to make a case that a higher AoA would have been "better". It's quite possible that the higher AoA would have dragged the tail in the water and ruptured the fuselage, allowing it to flood MUCH quicker. But It digress - that's the Sully case.

In the OTHER case the aeroplane had entered the manoeuvre with excess speed and then slowed by retarding the power, so when the alpha-floor-limit manoeuvre was initiated the engines were running at a much lower power level, and so took more time to spool up than was available. The insufficient climb rate was due to a lack of THRUST, not a lack of LIFT. Further increases in AoA would almost certainly have had no effect on the rate of climb because the associated increase in drag would soak up the increasing thrust and just delay the achievement of PRoC. It might change the fuselage attitude, but it would be unlikely to change the flightpath.

The two situations are completely different.

[KayPam]

The common point between those two accidents is that alpha max was not reached, even with full back stick, despite airbus claiming that alpha max will be reached if full back stick is applied.

The airplane should react like it is supposed to.

So that aircrews can know what to expect.

[CONF iture]

Originally Posted by vilas

You Herd the NTSB statement so it's not a joke. It was NTSB"s opinion based on surely some corroborating evidence.

Where is that corroborating evidence then ... ?
There is a hype put on those protections in this case that is simply not necessary.
And what allows you to state that Sully would have stalled without protection ... ?
Truth is that Sully would have most probably obtained a better touchdown with direct law.

The threshold at LHR is not much moveable, on the Hudson Sully could place it at his willing which helps a lot when it's time to flare, as long as the protections don't interfere ...

[misd-agin]

It's unknown how Sully would have performed without FBW protection. He significantly misread his airspeed. Why are we to believe that if he wasn't in a FBW a/c that he'd have read his airspeed better? If he'd flown a non FBW a/c, that didn't have AOA protection, he'd have triggered he stick shaker, and perhaps pusher, unless he'd lowered he nose and rudeced the AOA. At 150', or less, that would have generated an increased rate of descent that we can only speculate as to if he would have had enough airspeed, and AOA available, to reduce the increased sink rate prior to touchdown. IMO it would have been a much more difficult situation to handle.

Touchdown was at less than the desired pitch attitude of 11 degrees. But Airbus also didn't recommend reaching the pitch limited AOA at 150' either. If normal speeds had been maintained during the final segment of the glide there would have been enough airspeed to raise the nose to achieve more than 9.5% NU.

Learning point - slight excess airspeed might be a goal for a water ditching. It gives you the opportunity to choose to touchdown point with a slight degree of variability due to wave patterns or sea state. Wiith low airspeed/high AOA you basically have only one touchdown point.