Mid-air collision over Brasil

[jondc9]

Dear Aviators:

PLEASE HELP ME OUT.


I hope to further the cause of aviation safety, and any thoughts you might have on this collision over and above what has already been written might help me.

I am to be on Paula Zahn's CNN show wednesday evening to talk about this amazing event. In the typical way that modern television is done, I will be asked questions for half an hour and perhaps 2 seconds will be used on tv.

It is my hope to stir a proper debate to influence an improvement in air safety.


One question that might be asked is why the Legacy crew heard nothing according to the reporters article. The reporter might be on the show too.


I've considered this question, my initial response is that modern planes are insulted against sound from the outside. also, the air is much thinner and less able to transmit the sounds at the altitude involved. That the relative speeds of the planes were so high, there was little chance for the sound to register to a listener.

Any thoughts?


And are there any other thoughts that you as fellow aviation professionals might like expressed.


There are some on this forum who don't care for what I do. Fine. But this is your chance to have a fellow pilot speak about things that should interest you and you get some input at this forum.


regards

jon regas

perhaps 12 hours to comment, prior to my studio time, which is different than show time.

[ATC Watcher]

3miles, the answer lays in the Folha article ( asuming what is reported is correct of course): if my understanding of Portugese is correct , basically the flight was indeed planned at 360 but ATC lost contact with the legacy and it stays at 370.
If ATC did warn the other unit ( which had the 738 in contact) is still not known. The article mention some ATC errors ( to be taken with caution )

The article mention a loss of transponder ( which could explain the non-RAs) and some " delivery tests" done during that flight by the Legacy, " without informing ATC" but that comes from "rumours" from the military.

The direction of the FL according semi circular rule is not the key factor here ( We modify this all the time ad-hoc to ensure separation or expedite traffic ) The coordination between the 2 ATC centres is .
I would bet this is where ly one of the main cause of this collision.

[jondc9]

atc watcher

you make some fine points. if indeed the legacy was to be at a lower altitude and somehow lost contact with ATC, what happened to the "expected" portion of the clearance.

like:

maintain FL370 expect clearance to FL360 in blank minutes?


one of the hallmarks of modern flying is communications...radios are so reliable now perhaps the backup of EXPECTED was over looked.

and ATC watcher, sadly I think you are right on about what caused this crash


j

[Shuttleworth]

Such a sad day for the families and relatives,
What an absolute tragedy.

the Legacy had switched off the transponder in order to go up

If there is any truth in that; then someone needs to go to prison for a lifetime.

[SM4 Pirate]

Track offsets

Ventus,

The quality of an offset is equally and proportionately useless to the equipment of the other airframe involved in the conflict; Big sky theory has more knobs than ever before with increasing accuracy of GPS based nav etc.

Remember direct tracking, User preferred routes, flex tracks etc not flying 'on the airway'. An offset may out you directly nose to nose with someone going to a waypoint early off a SID etc.

Take a 90 degree crossing event, an offset may help, but equally might put you it, when 'on track' might, repeat might work when all the other holes in the cheese line up.

Would you want to be the one on a self initiated offset (without advising ATC) that actually 'casued, the last hole' and the impact?

Cheers

[Shore Guy]

Ventus45 said "Where to from here?”

ADS-B.

The issues of equipment redundancy still apply, but a collision avoidance system based on ADS-B would be infinitely superior. The accuracy is such one could do lateral escape maneuvers.

As we all know, TCAS is far from a clean sheet of paper design. It is a (for lack of a better word) bastardized system added on to a transponder design not envisioned to be an in-flight collision avoidance system.

With a CDTI (cockpit display of traffic information) and ADS-B, each aircraft would have "seen" each other over 100 miles away.

[Capt Pit Bull]

Ventus,

We'll have to wait and see, but...

In addition, where most aircraft have multiple redundancy for most systems, this is not so for TCAS, most aircraft only have ONE transponder fitted. If that transponder fails in flight, that aircraft, and all others around it, are in a much higher (statistically) state of potential - or actual danger of collision.

Not really. Most aircraft have 1 TCAS unit and 2 mode S SSR transponders.

But since you have 2 aircraft involved in the collision, and either TCAS can provide adequate safety as long as the other aircraft still has at least 1 SSR transponder working.

The "system" is the kit in both aircraft, not the kit in one aircraft. So in fact, there quite a lot of redundancy there. e.g. you could lose a TCAS unit in one aircraft, along with a transponder, and lose a transponder in the other aircraft, and still get a safe outcome.

pb


p.s. Before we start throwing money at more redudancy we ought to be making sure we are getting the best out of the kit we have. I remain convinced many TCAS training programs pay only lip service to the overall issues. I can think of many really nasty TCAS incidents caused by lack of crew knowledge / training (see note below) but far fewer (and certainly far fewer recently) caused by technical malfunction.

note: lack of crew knowledge / training is not equivalent to lack of professionalism, so don't go all defensive on me. But making sure there is robust training in place is a regulatory flight ops function, and I am not convinced it is happening globally. Just look at the misconceptions that crop up every time a TCAS related thread happens here for example.

night all

pb

[broadreach]

More news in Portuguese this afternoon (3 Oct) from the Correio Brasiliense, a Brasilia newspaper not noted for sensationalism: http://www2.correioweb.com.br/cbonli...ri_bra_193.htm

The article seems more accurate in terminology. It begins by focussing on the lack of timely communication between Brasilia ATC (Cindacta 1) and Manaus (Cindacta 4) and between these centres and the aircraft themselves, based on radar and comms recordings in the respective centres.

It then moves to actions on the part of the Legacy crew whose CVR and FDR began being analysed on Friday. The Legacy crew were interviewed in Rio on Monday by the aeronautical authorities who, by that time, had heard the CVR and conducted at least a superficial analysis of the FDR. Details not revealed but the article states that at some point after passing Brasilia the Legacy’s transponder was switched off, thus rendering the secondary radar returns that rely on the transponder and which would have more precisely established altitude, virtually useless (primary radar being able to track course accurately but altitude only to ±1,500’). Then, that the Legacy began a series of brusque manoevers up and down, left and right, descending at one point to FL 320 before resuming FL 370, those manoeuvers further complicating interpretation of radar signals.

Around that time Brasilia, unable to make radio contact with the Legacy, transferred control to Manaus, who then failed to take into account their own 738 and did not avail themselves of the Cachimbo mountain range radar assembly to more accurately track the Legacy.

At one point the article states that the accident committee believe the Legacy crew did see the 738 and attempted evasive action, too late. “Proof” offered of this being the transponder’s being switched back on shortly after the collision and, simultaneously, an emergency landing at Cachimbo AF base being requested.

The article contains a list of errors on the part of ATC and the Legacy, free translation:

ATC
- The Legacy was not alerted when it maintained FL 370 after passing Brasilia.
- ATC should have tried to alert the Legacy via any other aircraft.
- Manaus ATC alerted neithter of the aircraft of the possibility of collision

Legacy
- Did not follow the flight plan filed at Sao Jose dos Campos which included descent to FL 360 on passing Brasilia.
- Switched off transponder and performed dangerous maneouvers blinding the radar and making Cindacta 1 (Brasilia) and 4 (Manaus) work more difficult.
- Ignored repeated alerts from Brasilia ATC whilst varying altitude by up to 3,000 feet.

There is also a bit about standard commercial flight procedures ex Manaus, from which the Gol 738 was cleared to FL 410 normally to be accomplished in steps, on autopilot.

There are contradictions just about everywhere. Reports of “100 bodies in the tail section” yesterday have been adjusted to 50. The Gol 738’s recorders were recovered from the tail yesterday or today and the investigation team have been given 40 days to deliver a report. “Official” news is being delivered through daily press conferences by ANAC but reporters are getting more information and, of course, contradictory leads, from contacts on the accident site and in the investigation group.

[PJ2]

broadreach;

A thorough, balanced if not disturbing post, thank you.

[20driver]

Track offsets

Is there anything to stop an airline from incorporating a track offset - say .5 nm, right between 0 and 180 heading, left 180 to 360, in their SOPs?
A simple solution that requires no rule making. Still keeps you on the airway width.
20driver

[OVERTALK]

Previous Coverage of the Lateral Offset Tracking Trial

http://www.pprune.org/forums/showthr...elhi+Collision
.

further query

Anybody know whether these articles are still accessible anywhere (links appear dead)?

Long time Ppruners will be familiar with these threads from the very early days of Pprune.
.
http://www.pprune.org/pub/tech/MidAir2.html
.

http://www.pprune.org/pub/tech/MidAir.html

[bubbers44]

20, won't that put an east/west offset on the exact same line of flight? I am talking heading. How about .5 right for everybody and emergency descents always turn right. Kind of like a Texas road. Nobody crosses the center line. Or touches it. The Brits can just go to the head on rule of deviating right and forget the highway stuff.

[Bangkokeasy]

http://news.yahoo.com/s/ap/20061003/...DJlYmhvBHNlYwM-

Contains some relevant translations of Brazilian sources, for those non-Portuguese speakers among us.

[ORAC]

- Switched off transponder and performed dangerous maneouvers blinding the radar and making Cindacta 1 (Brasilia) and 4 (Manaus) work more difficult.
- Ignored repeated alerts from Brasilia ATC whilst varying altitude by up to 3,000 feet.

Primary 3D radar is only accurate to within +/- 5000ft except when used for comparative heights between 2 aircraft in close proximity. Digital radars exhibit track weave, which is why separation criteria is normally increased to 10nm laterally when more than 100nm away from the radar head.

The military "rumours" would therefore seem to be based on an ignorance of basic radar performance. I would suggest they be discounted and that the suggestion to exercise caution be implemented concerning all reports until more detailed FDR information is released.

[410]

20driver, think about what you've just suggested. It's the equivalent of two aero club aces deciding to do an opposing beatup of their local aeroclub. "I'll come in from the north, you come in from the south. I'll fly down the right side of the runway and fly down the left side."

1 mile right offset inbuilt into ALL systems when LNAV is selected when above F150 wouldn't be a panacea - ie, it wouldn't solve crossing conflicts, but it would have opposite direction traffic no longer on EXACTLY the same track for hours on end.

[jondc9]

controllers have a term for Wrong Altitude/Flight Level (for) Direction of Flight-WafDof (woff doff).

many years ago when given a WAFDAF, I would question it to get it on the tape...but it has become quite commonplace. Something like "CONFIRM WAFDAF".

I wonder if the TCAS installation on Legacy is like most in that turning your tansponder to TCAS or R/A etc is the only way to get your tcas on...perhaps when turning off transponder ( still unproven) they did not realize theyhad turned off TCAS.

j

[threemiles]

The direction of the FL according semi circular rule is not the key factor here ( We modify this all the time ad-hoc to ensure separation or expedite traffic )

There is always a danger in this thinking when not anticipating that loss of radar contact or comm failure may occur at some point. Exactly the same thinking was bringing the TU-154 and C-141 down.

Away from high dense airspaces and with the exception of the oceanic rush routes it is good practice to stick to semi circular rules over long thin traffic routes. No reason to not do so.

[Pom Pax]

Was there a collision?

I suggest that there does not have to have been any actual physical contact to have caused this accident.
On the road you meet a large lorry at a closing speed of 150 knots. As you pass your car experiences a severe buffet. Now consider 2 aircraft of similiar comparatives sizes closing at 6 times that speed. V squared has become much much greater. The winglet is subjected to this sudden force in the opposite sense (direction) to which it is designed to work, it fails strikes a glancing blow to the tail plane.
Back to the motoring comparison, Ranger One much earlier in the thread (post #178) made some calculations, that winglet now has as much energy as a very large car traveling at 40 mph. The winglet now strikes the the 738. Now refer to the Lockerbie (Pan Am) report to see how damage to 1 panel of the skin leads to the failure of others.

[Huck]

A corporate pilot never forgets who he's got in back.

This flight was loaded with big guns from both the owner corporation and Embraer - plus a travel reporter from an influential paper.

As a former corporate guy I can say this: I doubt seriously that they flew any maneuvers, other than straight and level and as smoothly as possible.

[410]

Yes Huck, I have to say a similar thought crossed my mind. I thought it sounded awfully like someone in officialdom preparing the ground for a "It wasn't our system that was at fault" defence.