The Atlantic Glider revisited - official report released (Merged)
FUEL FU/FOB DISCREPANCY recommendation
The part I like is the recommendation for mandating the pin change to the Flight Warning Computer programming to present an amber warning message when Fuel Used + Fuel on Board compared to Initial Fuel on Board at Departure exceeds 3500 Kg.
The part I don't like is that this capability was already in the FWC, but disabled
The part I don't like is that this capability was already in the FWC, but disabled
Join Date: Sep 2002
Location: Europe
Posts: 182
Likes: 0
Received 0 Likes
on
0 Posts
RatherBeFlying,
Not entirely true: The FU/FOB calculation is done in the FCMCs, and there only at standard 9 and above. FCMC 9 came up sometime late 2001, I seem to remember.
You only need to modify the FWC to make it "responsive" to the information from the FCMCs.
BTW: As of 2006, you need to have this warning function activated to perform ETOPS...
Cheers,
J.V.
Not entirely true: The FU/FOB calculation is done in the FCMCs, and there only at standard 9 and above. FCMC 9 came up sometime late 2001, I seem to remember.
You only need to modify the FWC to make it "responsive" to the information from the FCMCs.
BTW: As of 2006, you need to have this warning function activated to perform ETOPS...
Cheers,
J.V.
Thread Starter
Join Date: Dec 1999
Location: Ottawa, Ontario, Canada
Posts: 330
Likes: 0
Received 0 Likes
on
0 Posts
In asserting that the Portuguese authorities have laid "all the blame on [the crew] doing a check list from memory", Mac_Scott is, I believe, misreading their report, a report which appears eminently comprehensive and fair to me.
Mac is right when he points out that, as usual, a whole chain of events led to the emergency. However, the Portuguese investigators, in their report and particularly in their recommendations, draw attention to deficient training of aircrew in fuel leak detection and to inadequate provision of warning systems to alert crew to abnormal fuel situations. Certainly, had the Air Transat crew referred to their Quick Reference Handbook when dealing with the fuel imbalance they might have noticed the caution not to follow the procedure in the case of a suspected fuel leak - but this was only one of the events in the chain.
What shook me (SLF) most about this incident at the time - and still does - was the fact that the crew of a modern airliner on a long-haul flight had only 53 minutes notice of total fuel exhaustion.
Rockhound
Mac is right when he points out that, as usual, a whole chain of events led to the emergency. However, the Portuguese investigators, in their report and particularly in their recommendations, draw attention to deficient training of aircrew in fuel leak detection and to inadequate provision of warning systems to alert crew to abnormal fuel situations. Certainly, had the Air Transat crew referred to their Quick Reference Handbook when dealing with the fuel imbalance they might have noticed the caution not to follow the procedure in the case of a suspected fuel leak - but this was only one of the events in the chain.
What shook me (SLF) most about this incident at the time - and still does - was the fact that the crew of a modern airliner on a long-haul flight had only 53 minutes notice of total fuel exhaustion.
Rockhound
Join Date: May 2004
Location: Canada
Posts: 52
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Jul 2004
Location: Canada
Posts: 46
Likes: 0
Received 0 Likes
on
0 Posts
For Mac:
You are being overly kind to the crew. After reading all 103 pp. of the report, my conclusion is that the Portuguese are also being overly kind.
To begin, one of my old instructors said "A superior pilot is one who uses his superior judgment to avoid having to use his superior skill." This crew failed that directive.
I don't understand why the crew felt the need to use a crossfeed valve when they didn't understand what was going on. Unless there was significant aerodynamic evidence, i.e. one wing very heavy, there was no urgency to cross feed. This is one of the reasons why regulators now ensure that cross feed is only done by very purposeful actions of the crew.
There should be no remotely normal situation, with all engines operating, when a lateral imbalance gets out of hand. Once one of the engines is out, any half competent multi engine crew should be able to work out that the dead engine side will be heavy. Until that is clear, why break the isolation of the two sides?
"Computer malfunction" suspected is a pure cop out. If it was a computer malfunction, there would be no reason to cross feed, unless the crew suspected that the computer had performed some side to side transfer. In that case, cross feed is an uncontrolled reaction to an undiagnosed situation. The crew made a bad situation dire by lack of basic airmanship. (What ever happened to "dead foot - dead engine" and the corresponding effect on heavy wing?)
Nevertheless, once they had placed themselves in such a glider situation, they did not need the altitude they kept approaching Lajes. In their defense, I have been to Lajes at night, and you really can't see anything until you are almost lined up on the runway. (The lights have some directionality, and there is high ground - not a cliff, but a rise - north of the runway). Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better. They had time to do the old descent vs. distance calculation in their minds when they could see their stabilized descent rate.
As an ex-military pilot, perhaps the oft practiced calculation of altitude needed to get to one of:
1. Overhead on runway heading at 2 mins. worth of altitude, or
2. Downwind at 1 min. worth of altitude plus a buffer to drop gear and flap, or finally
3. Join on base at something less than 1 minute of altitude
is too simplistic for an airline crew long separated from such thoughts. Still, it does not show a good appreciation of time/altitude/distance to get so close so high.
They did, however, scrub their altitude and speed successfully. 200 K over the threshold is not a terrible result. However, in light winds there does not appear to be a good reason to hammer the gear onto the ground as hard as they appear to have. Full braking was perhaps their best decision.
I'm not at all impressed by this crew. Obvious comparisons are to the crew of an Air Canada 767 glider which successfully landed on a 6500' runway from over 100 NM out, or the celebrated DC-10 pilot who saved most of the passengers after physically losing an engine and all flight controls.
The accident report focuses heavily on training re the computer pages and fuel imbalance situations. It appears that the investigators have accepted that airline crew are essentially automatons. This line of thinking will eventually lead to robot airliners. Only recently a fellow engineer told me that the high rate of human error means the machine must avoid human input at all times. Crews who show the lack of airmanship and understanding of mechanical systems evident here reflect badly on our community.
You are being overly kind to the crew. After reading all 103 pp. of the report, my conclusion is that the Portuguese are also being overly kind.
To begin, one of my old instructors said "A superior pilot is one who uses his superior judgment to avoid having to use his superior skill." This crew failed that directive.
I don't understand why the crew felt the need to use a crossfeed valve when they didn't understand what was going on. Unless there was significant aerodynamic evidence, i.e. one wing very heavy, there was no urgency to cross feed. This is one of the reasons why regulators now ensure that cross feed is only done by very purposeful actions of the crew.
There should be no remotely normal situation, with all engines operating, when a lateral imbalance gets out of hand. Once one of the engines is out, any half competent multi engine crew should be able to work out that the dead engine side will be heavy. Until that is clear, why break the isolation of the two sides?
"Computer malfunction" suspected is a pure cop out. If it was a computer malfunction, there would be no reason to cross feed, unless the crew suspected that the computer had performed some side to side transfer. In that case, cross feed is an uncontrolled reaction to an undiagnosed situation. The crew made a bad situation dire by lack of basic airmanship. (What ever happened to "dead foot - dead engine" and the corresponding effect on heavy wing?)
Nevertheless, once they had placed themselves in such a glider situation, they did not need the altitude they kept approaching Lajes. In their defense, I have been to Lajes at night, and you really can't see anything until you are almost lined up on the runway. (The lights have some directionality, and there is high ground - not a cliff, but a rise - north of the runway). Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better. They had time to do the old descent vs. distance calculation in their minds when they could see their stabilized descent rate.
As an ex-military pilot, perhaps the oft practiced calculation of altitude needed to get to one of:
1. Overhead on runway heading at 2 mins. worth of altitude, or
2. Downwind at 1 min. worth of altitude plus a buffer to drop gear and flap, or finally
3. Join on base at something less than 1 minute of altitude
is too simplistic for an airline crew long separated from such thoughts. Still, it does not show a good appreciation of time/altitude/distance to get so close so high.
They did, however, scrub their altitude and speed successfully. 200 K over the threshold is not a terrible result. However, in light winds there does not appear to be a good reason to hammer the gear onto the ground as hard as they appear to have. Full braking was perhaps their best decision.
I'm not at all impressed by this crew. Obvious comparisons are to the crew of an Air Canada 767 glider which successfully landed on a 6500' runway from over 100 NM out, or the celebrated DC-10 pilot who saved most of the passengers after physically losing an engine and all flight controls.
The accident report focuses heavily on training re the computer pages and fuel imbalance situations. It appears that the investigators have accepted that airline crew are essentially automatons. This line of thinking will eventually lead to robot airliners. Only recently a fellow engineer told me that the high rate of human error means the machine must avoid human input at all times. Crews who show the lack of airmanship and understanding of mechanical systems evident here reflect badly on our community.
Join Date: Jul 2001
Location: north of CYMX
Posts: 82
Likes: 0
Received 0 Likes
on
0 Posts
plt_aeroeng
"they did not need the altitude they kept approaching Lajes."
"Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better."
"They had time to do the old descent vs. distance calculation in their minds when they could see their stabilized descent rate. "
Are you out of your mind?????
"As an ex-military pilot"
OK, I see...
Hasn't your old instructor also told that your better off losing extra altitude than... not making the field?
"they did not need the altitude they kept approaching Lajes."
"Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better."
"They had time to do the old descent vs. distance calculation in their minds when they could see their stabilized descent rate. "
Are you out of your mind?????
"As an ex-military pilot"
OK, I see...
Hasn't your old instructor also told that your better off losing extra altitude than... not making the field?
Join Date: Sep 2004
Location: canada
Posts: 4
Likes: 0
Received 0 Likes
on
0 Posts
Transat pilot flying high in popular opinion
plt_aeroeng
It's very easy for you talk the talk of a superior pilot whilst sat at your computer. In the cold light of day and especially with the benefit of hindsight everyone will question why they chose to cross feed as it's obvious that they should not have done. At the time that the imbalance was noticed they were already trying to fault find a fault which was not covered in the QRH and like the fuel leek fault one that they had not previously experienced in the sim. This was obviously enough to have hampered their thought process and made them doubt the ECAM. You must also remember that the Air France pilots did exactly the same thing.
"Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better"
I would have thought it was much better to go for best glide range arrive high and then lose the height whilst preparing for the landing rather than increasing your workload trying to accurately plan your glide angle only to find that you've c**cked up and end up in the drink.
"200 K over the threshold is not a terrible result"
I would have though that it was very close to the stalling speed for an A330 with only slats deployed. A very good result I would imagine.
"However, in light winds there does not appear to be a good reason to hammer the gear onto the ground as hard as they appear to have"
When you consider that they had no flaps, no spoilers, no reverse thrust, no autobrake, no antiskid limited pitch control and no second chance I would of thought that getting the thing down was all that mattered. Perhaps a superior pilot such as yourself would have managed to do a perfect landing and controlled the braking sufficiently to allow the aircraft to coast to the gate, however I think that they did a stunning job of getting themselves out of the mess that they got themselves into.
I'm only a wanabbie so maybe my thoughts are a little off the mark, and if so I apologise. I'd be very interested to hear if any line pilots agree with me.
It's very easy for you talk the talk of a superior pilot whilst sat at your computer. In the cold light of day and especially with the benefit of hindsight everyone will question why they chose to cross feed as it's obvious that they should not have done. At the time that the imbalance was noticed they were already trying to fault find a fault which was not covered in the QRH and like the fuel leek fault one that they had not previously experienced in the sim. This was obviously enough to have hampered their thought process and made them doubt the ECAM. You must also remember that the Air France pilots did exactly the same thing.
"Still, 13,000 feet up at 8 miles distance suggests that they could have planned their final approach a little better"
I would have thought it was much better to go for best glide range arrive high and then lose the height whilst preparing for the landing rather than increasing your workload trying to accurately plan your glide angle only to find that you've c**cked up and end up in the drink.
"200 K over the threshold is not a terrible result"
I would have though that it was very close to the stalling speed for an A330 with only slats deployed. A very good result I would imagine.
"However, in light winds there does not appear to be a good reason to hammer the gear onto the ground as hard as they appear to have"
When you consider that they had no flaps, no spoilers, no reverse thrust, no autobrake, no antiskid limited pitch control and no second chance I would of thought that getting the thing down was all that mattered. Perhaps a superior pilot such as yourself would have managed to do a perfect landing and controlled the braking sufficiently to allow the aircraft to coast to the gate, however I think that they did a stunning job of getting themselves out of the mess that they got themselves into.
I'm only a wanabbie so maybe my thoughts are a little off the mark, and if so I apologise. I'd be very interested to hear if any line pilots agree with me.
stilljustanothernumber
Join Date: Apr 2002
Location: the night sky
Posts: 624
Likes: 0
Received 0 Likes
on
0 Posts
Well I agree with you Sky Wave.
That crew were let down by Airbus, Rolls, their own engineers their sim trainers and themselves.
I wouldn't like to try that glide. Aside from the practical challenges of it, imagine the fear, stress, guilt, panic and embarresment they must have been feeling.
That crew were let down by Airbus, Rolls, their own engineers their sim trainers and themselves.
I wouldn't like to try that glide. Aside from the practical challenges of it, imagine the fear, stress, guilt, panic and embarresment they must have been feeling.
Join Date: Jan 2001
Location: UK
Posts: 2,044
Likes: 0
Received 0 Likes
on
0 Posts
plt_aeroeng
Out of order... Once the Engines had flamed out, I think the crew did exceptionally well.
Interesting report, shame about no FDR for the glide in!
If you have not flown a modern Airbus, they are fine when everything works. However, once you get to Emer Elec Config, life gets quite hard, but more significantly very different to normal ops. We practice it in the Sim - but that's with the engines working. I have also done, in the Sim, (successfully) the glide approach and landing - but with an APU, and having planned it. This crew at night, with most systems not working (remember we are talking a FBW aircraft here, with the FBW system in it's final degraded mode short of packing up), got the aircraft down and all OK. It doesn't get much worse than this in a FBW Airbus, short of total disaster... and cannot be compared to a Boeing in this situation.
I also flew the A340, and an early turnback, and dumping required caused a number of "anomalies" to occur e.g. it failed to actually dump while this Trim Xfr was going on, despite the pretty screens showing it dumping (just the quantity did not reduce!). Hopefully the system has been improved, but when you have computers shifting fuel around largely at their whim, this will mask some of your basic observations / analysis.
IMHO that went some way to recover the siutation that they did, to be fair, contribute to in the first place. These things happen - let's learn, not 20:20 hindsight sniping at the crew. Could be you next time?
NoD
Out of order... Once the Engines had flamed out, I think the crew did exceptionally well.
Interesting report, shame about no FDR for the glide in!
If you have not flown a modern Airbus, they are fine when everything works. However, once you get to Emer Elec Config, life gets quite hard, but more significantly very different to normal ops. We practice it in the Sim - but that's with the engines working. I have also done, in the Sim, (successfully) the glide approach and landing - but with an APU, and having planned it. This crew at night, with most systems not working (remember we are talking a FBW aircraft here, with the FBW system in it's final degraded mode short of packing up), got the aircraft down and all OK. It doesn't get much worse than this in a FBW Airbus, short of total disaster... and cannot be compared to a Boeing in this situation.
I also flew the A340, and an early turnback, and dumping required caused a number of "anomalies" to occur e.g. it failed to actually dump while this Trim Xfr was going on, despite the pretty screens showing it dumping (just the quantity did not reduce!). Hopefully the system has been improved, but when you have computers shifting fuel around largely at their whim, this will mask some of your basic observations / analysis.
IMHO that went some way to recover the siutation that they did, to be fair, contribute to in the first place. These things happen - let's learn, not 20:20 hindsight sniping at the crew. Could be you next time?
NoD
Interested to see the rates of descent. 1000-1100 fpm gliding to field. 1700-2000 during orbit, then 1800 fpm when established on final approach.
Would these be typical for most modern jet aircraft ? (I imagine the A330 makes a particularly good glider ?)
Would these be typical for most modern jet aircraft ? (I imagine the A330 makes a particularly good glider ?)
Join Date: Jan 2001
Location: UK
Posts: 2,044
Likes: 0
Received 0 Likes
on
0 Posts
Smurfjet
There are various degrades on an Airbus. You start with "Normal Law", then some flavours of "Alternate Law" (fairly similar to Normal), then "Direct Law". Direct is where the control positions are directly related to the sidestick - rather like any other aircraft, and you need to use the trim wheel. A330 might be a bit different... It might sound easy, but when you're not used to trimming, and have all the other degrades, it gets rather complex.
Whatever, the next degrade for FBW is "nil"! You do now have Rudder and Pitch Trim... It's flyable, and some people practice that that to landing - but with engines <G>
NoD
...in it's final degraded mode short of packing up
Whatever, the next degrade for FBW is "nil"! You do now have Rudder and Pitch Trim... It's flyable, and some people practice that that to landing - but with engines <G>
NoD
Join Date: Dec 2002
Location: Beverly Hills 90210
Posts: 88
Likes: 0
Received 0 Likes
on
0 Posts
On an extremely popular, funny, and smart TV show the Captain Piché will be one of the main guests and will be interviewed and humored.
Sunday 8pm - 10pm Radio-Canada
le commandant Robert Piché,
http://radio-canada.ca/television/toutlemondeenparle/
Sunday 8pm - 10pm Radio-Canada
le commandant Robert Piché,
http://radio-canada.ca/television/toutlemondeenparle/
Join Date: Jul 2004
Location: Canada
Posts: 46
Likes: 0
Received 0 Likes
on
0 Posts
Gumbi:
After rereading the report, I agree with you: the crew did do a 360 before landing (P 8), and thus did handle their descent profile well.
In any case, I had no real issues with their approach to Lajes, my primary point was their actions before the double flameout.
Sky Wave:
The quote was from early student days being told what a fool I was, and it stuck. I stand by my position that there is a need for reasoning using basic airmanship. This should have suggested to the crew that crossfeed was a bad idea until either they better understood what was going on, or the imbalance got so bad that control became difficult.
My note re 200K over the threshold should have been interpreted as being a good result after the previous flight path. However, if an A330 is near stall at 200K, why did it float 1,800 feet during the first bounce? I do accept that with the very degraded flight control mode, it may be unreasonable to expect a smoother landing.
Just as a comment, I never suggested that I personally should be regarded as a superior pilot. Given that I have never had to practise a glider approach in a large aircraft, who knows whether I would have displayed superior skill. The real point is that judgment is generally more important than stick and rudder skill. Still, other large aircraft crews in glider situations seem to have done at least as well, perhaps better, and this is really the standard. I do think that the press reports describing the crew as "heroes" is somewhat excessive.
Nigel:
Perhaps you are right, in that modern aircraft do get very ropy with systems out. (Although - what about the DHL crew who suffered missile damage ex Baghdad, and successfully brought the aircraft back?) The real point, as I suggested, was the period before both engines flamed out.
Del Prado:
Fair analysis, although as Gumbi has pointed out, one would not want to guess wrong on final. Up to final, it should be relatively straightforward to predict altitude at destination. However, modern jet aircraft do have very large fans. With no fuel to turn them, these tend to worsen glide performance.
After rereading the report, I agree with you: the crew did do a 360 before landing (P 8), and thus did handle their descent profile well.
In any case, I had no real issues with their approach to Lajes, my primary point was their actions before the double flameout.
Sky Wave:
The quote was from early student days being told what a fool I was, and it stuck. I stand by my position that there is a need for reasoning using basic airmanship. This should have suggested to the crew that crossfeed was a bad idea until either they better understood what was going on, or the imbalance got so bad that control became difficult.
My note re 200K over the threshold should have been interpreted as being a good result after the previous flight path. However, if an A330 is near stall at 200K, why did it float 1,800 feet during the first bounce? I do accept that with the very degraded flight control mode, it may be unreasonable to expect a smoother landing.
Just as a comment, I never suggested that I personally should be regarded as a superior pilot. Given that I have never had to practise a glider approach in a large aircraft, who knows whether I would have displayed superior skill. The real point is that judgment is generally more important than stick and rudder skill. Still, other large aircraft crews in glider situations seem to have done at least as well, perhaps better, and this is really the standard. I do think that the press reports describing the crew as "heroes" is somewhat excessive.
Nigel:
Perhaps you are right, in that modern aircraft do get very ropy with systems out. (Although - what about the DHL crew who suffered missile damage ex Baghdad, and successfully brought the aircraft back?) The real point, as I suggested, was the period before both engines flamed out.
Del Prado:
Fair analysis, although as Gumbi has pointed out, one would not want to guess wrong on final. Up to final, it should be relatively straightforward to predict altitude at destination. However, modern jet aircraft do have very large fans. With no fuel to turn them, these tend to worsen glide performance.
Iconoclast
Join Date: Sep 2000
Location: The home of Dudley Dooright-Where the lead dog is the only one that gets a change of scenery.
Posts: 2,132
Likes: 0
Received 0 Likes
on
0 Posts
Who is really at fault?
I noticed in the various postings since the release of the report some individuals have taken sides as to who was at fault and why. Granted I did not read the report (loss of concentration) and this may have been covered. I have spent the last 35 years as an assurance engineer. I have seen how airframe manufacturers skim over potential problems due to time constraints or financial constraints and in some cases not wanting to follow the input of assurance engineers like myself.
This problem did not start at 35,000 feet over the Atlantic. It did not start in the Air Transat maintenance facilities at Mirabel, Quebec. It started at the Rolls Royce facility in (I assume) Derby, England.
Had a proper maintainability program been implemented by Rolls Royce they wold have prepared a design checklist. One aspect of this checklist would be directed at parts or assemblies that could be installed backwards. In the case of the Air Transat engine this would include the fuel line and the hydraulic line both of which could be installed backwards. It is obvious that such a checklist was never prepared. Since the parts could be installed backwards the technical manual for the engine should have contained a cautionary note telling the mechanic that caution should be taken when installing the respective lines. If the line(s) was/were installed incorrectly the design clearance between the two lines would close up allowing one line to abrade on the other resulting in a leak and in the case of the Air Transat engine total structural failure of the fuel line.
This problem had been experienced on both General Electric and Pratt & Whitney engines resulting in minor leakage that was discovered during inspection of the engine and pylon support.
Both companies issued ADs outlining the problem and providing detailed instruction relative to establishing the clearance between the two lines as well as periodic checks to see if the clearance had closed up.
This happened long before the Air Transat incident so Rolls Royce should have known about the potential for the same type of incident on their engines.
I don’t have access to Rolls Royce tech manuals and if any of the readers of this forum do please check to see how Rolls Royce is handling this situation.
According to my sources Air Transat took delivery of a brand new A-330 after the Atlantic Glider incident and during receiving inspection they found that the hydraulic and fuel lines were contacting each other.
This problem did not start at 35,000 feet over the Atlantic. It did not start in the Air Transat maintenance facilities at Mirabel, Quebec. It started at the Rolls Royce facility in (I assume) Derby, England.
Had a proper maintainability program been implemented by Rolls Royce they wold have prepared a design checklist. One aspect of this checklist would be directed at parts or assemblies that could be installed backwards. In the case of the Air Transat engine this would include the fuel line and the hydraulic line both of which could be installed backwards. It is obvious that such a checklist was never prepared. Since the parts could be installed backwards the technical manual for the engine should have contained a cautionary note telling the mechanic that caution should be taken when installing the respective lines. If the line(s) was/were installed incorrectly the design clearance between the two lines would close up allowing one line to abrade on the other resulting in a leak and in the case of the Air Transat engine total structural failure of the fuel line.
This problem had been experienced on both General Electric and Pratt & Whitney engines resulting in minor leakage that was discovered during inspection of the engine and pylon support.
Both companies issued ADs outlining the problem and providing detailed instruction relative to establishing the clearance between the two lines as well as periodic checks to see if the clearance had closed up.
This happened long before the Air Transat incident so Rolls Royce should have known about the potential for the same type of incident on their engines.
I don’t have access to Rolls Royce tech manuals and if any of the readers of this forum do please check to see how Rolls Royce is handling this situation.
According to my sources Air Transat took delivery of a brand new A-330 after the Atlantic Glider incident and during receiving inspection they found that the hydraulic and fuel lines were contacting each other.
Join Date: Dec 2002
Location: Beverly Hills 90210
Posts: 88
Likes: 0
Received 0 Likes
on
0 Posts
His book came out 2 years ago and I got a copy and got it autographed by him at a book signing.
He was quite talkative considering the legal stuff was going on. This detailed conversation wouldn't have happened in the states.
Robert Piché aux commandes du destin
http://www.amazon.ca/exec/obidos/ASI...405940-4644328
He was quite talkative considering the legal stuff was going on. This detailed conversation wouldn't have happened in the states.
Robert Piché aux commandes du destin
http://www.amazon.ca/exec/obidos/ASI...405940-4644328