Air Canada suing Airbus re: Halifax crash landing

[Zan Vetter]

So FPA is basically the same as having to increase V/S when manually creating a CDA the old way. Thanks to Rich for the math, it made clear that we're merely descending through more height loss in cold temps, something that was previously clear when having to just fly a higher V/S. My misunderstanding was due to the misconception that FPA was somehow an improvement on that method, when rather it is just another way. 

With respect to the lawsuit, it would be unlikely to fly a wildly incorrect V/S, and would be immediately obvious (by looking at the VSI) if the aircraft wasn't performing as selected. It is far less clear if or when the aircraft is flying a bad FPA; and mental math to crosscheck inside the FAF is difficult to keep up with from a workload perspective particularly on nights like the one in question. Which was the main improvement of FPA over V/S- less workload. Only it now appears that perhaps we lost a key crosscheck insofar as having to trust that the aircraft actually followed the selected FPA.

 

[Zan Vetter]

I've been on the airbus for 8 years, and still learning every day. The other key difference with V/S is that you would, based on your crosscheck, adjust V/S inside the FAF to remain on your profile. With FPA, at least at my airline, we are prohibited from adjusting the selected FPA inside the FAF. This perhaps takes pilots a bit outside the mental space of maintaining a running groundspeed/altitude/descent rate(path) crosscheck because in blunt terms, you can't do anything about it anyway. Just fly to MDA and see or don't see the runway, and then land or go around. It's less work, perhaps less error prone, but it's also more reliant on the aircraft doing what it's been told to do.

[Kip Powick]

2 hours ago, Rich Pulman said:

Yes, it would appear you are 'missing something here

And what am I really missing ???? What are you clearing up for me?? Do you think I don't know what a stabilized approach is ?? Let's delete the common sense thing about temperature correction etc that, perhaps, you think I would  ignore in cold weather  and why not   discuss how one goes about blaming the FPV/FPA for impact with the terrain after all that is what AC via ACPA is doing..

I am saying the same thing...it really doesn't matter what angle you descend at....you DO NOT descend below MDA until visual  with the landing environment....why does everyone keep harping on what the FPA does? It may be a great asset and if flown correctly assists with rate of descent  but how could it possibly lead you into landing short...........unless you are using it AFTER MDA when you are supposed to be doing a visual approach and landing.

Kip/Defcon, we don't follow/fly the FPA below MDA.  It gets us to the MDA and that's it.  It is visual flying only from that point on....or a go-around if runway environment not in sight.

That's how we did it on the A310 as well.

Question......do you revert back to the VSI for descent when the FPA technology is not serviceable. or do you abort the approach ????

Doesn't anyone remember how a NPA was flown before the advent of FPV /FPA ?? We flew a constant rate of descent based on GS and went to MDA....if not visual we held MDA until other Nav aids, (timing/DME/bearings) "said" we were at MAP ..If you were not visual you commenced a GA.

Minimums on NPA are higher than PA for a reason, the Nav aids are not as accurate for NPA and the higher weather mins and MDA give you a bit more room to transition to a visual landing...but if not visual, one does not revert back to a descent "on" the dials in hopes of seeing the runway/landing environment...............and I know we all are supposed to  know that ..but................

[Seeker]

1 hour ago, Kip Powick said:

And what am I really missing ???? What are you clearing up for me?? Do you think I don't know what a stabilized approach is ??

 

Kip.  I think what's thrown a few people off is this thing you said; "PS.... I see nothing wrong with reaching MDA early and having to maintain MDA for a short period of time until MAP or landing."

Now, you don't have to convince me.  I believe you do know what a stabilized approach is, the difference between FPA, CDA, V/S, etc etc even though your statement above is totally old-school/non-stabilized/stick and rudder/hands and feet.  I know 'em both.  I had 10,000 hours of non-precision instrument flying - full proc turn, cross the beacon, hack the time and dive down to mins, run along at mins until you see the airport or "times up!" climb power - before getting introduced to CDA/FPA/Stable App.  I now have over 10,000 hours doing it the "new" way.

Anyway, like I said, I think you know what you're, and we're, talking about but the statement above kinda says the opposite because these approaches, as we fly them, have no MAP.

[Homerun]

48 minutes ago, Kip Powick said:

 

Question......do you revert back to the VSI for descent when the FPA technology is not serviceable. or do you abort the approach ????

Doesn't anyone remember how a NPA was flown before the advent of FPV /FPA ?? We flew a constant rate of descent based on GS and went to MDA....if not visual we held MDA until other Nav aids, (timing/DME/bearings) "said" we were at MAP ..If you were not visual you commenced a GA.

Minimums on NPA are higher than PA for a reason, the Nav aids are not as accurate for NPA and the higher weather mins and MDA give you a bit more room to transition to a visual landing...but if not visual, one does not revert back to a descent "on" the dials in hopes of seeing the runway/landing environment...............and I know we all are supposed to  know that ..but................

No, we are prohibited from flying a NPA with v/s on the Airbus ( AC ). Its FPA or no approach.

The minimums for the approach in question in YHZ were 275 agl at the time of the accident...not much different than a CAT 1 ILS, except there is no guarantee when flying an NPA with FPA that you will be in the "slot" at MDA.  So issues could arise if you arrive at MDA at a distance from the threshold (closer or further) that is different than contemplated by the approach design. IMO, being at an incorrect distance from the threshold at MDA could potentially cause all sorts of visual illusions given rolling terrain around an airport, poor lighting, low visibility etc.

[Kip Powick]

19 minutes ago, Homerun said:

No, we are prohibited from flying a NPA with v/s on the Airbus ( AC ). Its FPA or no approach.

Thank you for that info

[Vsplat]

4 pages worth.  Hmmm. 

Maybe we can establish/reset some basics.

In the old days, Non precision descents used a target vertical speed tied to an expected groundspeed.  If groundspeed was different, the vertical profile would be different than planned and the aircraft would arrive at MDA too early or too late.

SELECTED FPA on the narrow body Airbus is intended to do nothing more than augment vertical speed with accurate compensation for groundspeed, so fly a consistent, expected vertical profile regardless of groundspeed. 

Seemed to be a good fix for 'the old way".  But, while those using vertical speed intuitively understood that it was not anchored anywhere, that understanding did not seem to transfer to selected FPA. 

I believe the biggest reason for this confusion is that, on the Airbus, there is both SELECTED FPA, described above, and MANAGED FPA.   MANAGED FPA IS anchored.  It will join two points at the indicated altitudes stated in the FMS.  MANAGED FPA is known to be dependent on barometric values (HDG/VS), so when it is cold and FAF crossing altitude in the box is wrong,  managed can't be used.

For a time, Airbus advertised that, while MANAGED FPA used HDG/VS mode and so was barometric, SELECTED FPA used TRK/FPA mode, at the time considered to include only inertial data and so would keep the aircraft on the selected path no matter the termperature, wind, whatever.  The path was still unanchored, so where that path ended was entirely dependent on when it started, but at least it was going to be predictable.  Well, that got challenged right around 2000 and the discussion of barometric influence on selected FPA started to emerge.  What has happened since has been a confusing debate between operators of countries that correct altimeters for cold, those that don;t, and Airbus.  The thread here is a good reflection of most of that debate.

Some of us on this forum fly the bus, and you can see the many ways this is understood.  Does it need to be so complex?  Maybe the answer to that question will come out of all this.

FWIW

Vs

 

[Kip Powick]

36 minutes ago, seeker said:

Kip.  I think what's thrown a few people off is this thing you said; "PS.... I see nothing wrong with reaching MDA early and having to maintain MDA for a short period of time until MAP or landing."

Ok..Thank you for the clarification........was not really fully aware that one was not permitted to do that.

[DEFCON]

Thanks to Seeker and Vsplat for their explanations; I now have a much better appreciation of the Airbus flight guidance system.

Kip wondered how we got into the debate on FPA in the first place as the aircraft contacted the ground prior to the runway. Seeing that AC launched a suit in respect of some sort of failure of the technology, questions related to same seem appropriate.

Does the Airbus system provide flight guidance below MDA and if so, is the information presented on the display a continuing projection of the computers previous calculations of the aircraft's position & trajectory that created the FPA in the first place, or do inputs from sensors keep updating the information displayed until landing?

 

[Homerun]

3 hours ago, DEFCON said:

Thanks to Seeker and Vsplat for their explanations; I now have a much better appreciation of the Airbus flight guidance system.

Kip wondered how we got into the debate on FPA in the first place as the aircraft contacted the ground prior to the runway. Seeing that AC launched a suit in respect of some sort of failure of the technology, questions related to same seem appropriate.

Does the Airbus system provide flight guidance below MDA and if so, is the information presented on the display a continuing projection of the computers previous calculations of the aircraft's position & trajectory that created the FPA in the first place, or do inputs from sensors keep updating the information displayed until landing?

 

You are flying visually below MDA.

[Boney]

Still awaiting the result from TC, but I still believe that "wind", aka windshear, was a factor.

[Moon The Loon]

On 4/5/2017 at 3:26 PM, Boney said:

Still awaiting the result from TC, but I still believe that "wind", aka windshear, was a factor.

"TC?"

Dontcha mean the TSB?

They are two separate, isolated, independent organizations.

Theoretically.

[Boney]

Correct, TSB. 

Thanks. 

[DEFCON]

Thanks Homerun.

Even if the wrong barometric pressure was applied, wouldn't the rad alt tell the crew there was a fundamental flaw in the numbers during the approach from the FAF inbound?

  

[thor]

I don't think that Air Canada had upgraded this aircraft with EGPWS or GPS. Radar altimeter would be reading the ground directly below the aircraft and as the RWY is elevated above the surrounding ground, the altitude would not be used.

[DEFCON]

Wouldn't a cross check of the aircraft altitude at the FAF confirm the ac's 3 dimensional position before the crew continued the descent?

 

 

[thor]

Yes, barometric altimeter not radar altimeter.

[Vsplat]

Bear in mind that radio altitude is not provided on charts except for CAT III and CAT II that uses RA for Decision Height.  Not available for non precision approaches or CAT I ILS.   Interpreting Radio Altitude in a dynamic environment would be confusing, especially for non-localizer approaches, as even minor deviations and bank angles would mess with it,

Barometric cross checks are the norm of course, but even for enhanced charts like some of the Jeppesen products, the cross check values inside the FAF are ISA.  Not much use once cold corrections are applied unless the crew corrects the whole strip.  I don't know of an operator that does that.

Then there is the FPA correction for cold.  From what I can see the result of this correction is not a consistent flight path, but one whose arc is intended to intersect with MDA at about the right spot.  The exact shape of that arc changes for every approach.  Cross checks would be approximate.  Maybe effective mitigation, too early to know for sure.

So many moving parts to this, it will really take the whole report to understand what got a hold of this flight.

Vs

[Specs]

On Thursday, March 30, 2017 at 0:38 PM, dagger said:

http://www.ctvnews.ca/business/air-canada-lawsuit-accuses-airbus-of-negligence-in-halifax-crash-landing-1.3347461

 

HALIFAX - Air Canada is claiming a French aircraft manufacturer's negligence contributed to a crash landing at Halifax Stanfield International Airport two years ago.

The Canadian airline is suing Airbus SAS, saying the company failed to identify shortcomings of the Airbus 320.

It's not just a delay tactic in hopes of settling the Class action sooner?  Maybe an Insurance policy thing then?

[Moon The Loon]

I still can't get my head around the most important part of ANY approach is the crew's awareness of the approach environment. In this case, it did not seem to exist.

[Vsplat]

Let's be clear:  We don't know, with any precision,  the level of situational awareness in the flight deck, how it evolved or why.  It's easy to say the crew did not know they were in trouble until it was too late, but such a statement can too easily be construed as suggesting that they missed something. 

Sometimes it's not a question of missing something vital, but including something vital that should be accurate and trustworthy but for whatever reason is not.

Put another way:  Any crew operating at this level, in these conditions, having already faced the decisions this one had, would be updating constantly.  High data flow is a double edged sword - Correct data leads to higher precision.  Incorrect data the opposite.   The faster you upload new data, the faster you correct obsolete information, - or propagate the effects of a bad data source.

There are methods to test and verify navigational and positional accuracy, but most of these are intended for use prior to final approach.  if a data source goes from accurate to inaccurate inside the FAF, depending on what it is, how it degrades and when,  the deterioration may be next to impossible to detect.

Remember this was not an ILS, so no glidepath flag, no vertical dots, just the aircraft seeming to fly the flight path it was told to fly.  There is no comparator within the instruments that will tell the crew if the aircraft is vertically displaced from a selected flight path angle by turbulence/ shear.  There is nothing to warn if the aircraft enters a region of unusual pressure affecting altimeter or vertical speed.  The environment, as we all know, was difficult, but did something in there result in the crew getting wrong or misleading information?  Only the TSB and whoever they consult with knows at this point. 

All we, as public onlookers have, are opinion and theory.

All that to say, I have looked at as much of this file as anyone I know outside of the investigators.  I have not yet seen anything to put a finger on the crew.

FWIW

Vs