Quote:
"@HN39 - Understood, and it wasn't a criticism, I just wanted to make sure that as the thread meanders to its late stages we got as many of our facts in a row as we can."

IMO, that has to be one of the most patronising comments I've seen on this thread - from someone who is inclined to bluster in areas of which he understands little.

Believe me, it wasn't meant to sound like that! Honest and heartfelt apologies if it came across that way.

I felt it was important to make the distinction because, as I've learned from you all, a simulator's behaviour around the edges of the flight envelope will usually be approximate, whereas doing it for real would give a more accurate picture.

As far as I'm aware, I knew it was Bechet who flew the reconstructions at Toulouse because he said he did on the ACI/Mayday programme on the subject.

Thank you, Doze. What I wanted to confirm. I feel we are on the same page, and I have actual experience flying the "limiters" or "protections" in a FBW system.

???!!!what!?

Today all the airline pilots are type rated on these simulators for all their training and formation!!!!

That's true today, but not on the A320 at Blagnac in February/March 1988. The Thomson-CSF flight simulator was very unreliable, and certainly not approved for zero-flight-time type-ratings! Admittedly, it would have been a bit better 6 months later.

Bonsoir Chris,

Thank you for your time reference.

Asseline said he trained his 100 FT height flight on the Thomson simulator (with sealed software) used to train AF first A320 pilots. Isn't ?

Which informations and simulator quality had Asseline, Bechet, the BEA, about AoA in Alpha protection and flight laws ?

So You were type rated by Test pilots who had to do instruction and qualification ?

roulis!

More of a history reference, after 26 years! Sorry, I should have explained better, so today I have been looking at the Aeroformation course information, the schedule (with many amendments), and my own notes annotated on them.

Capt Asseline was, I now understand, to be in charge of A320 flight training for Air France, and would have been on Aeroformation (later renamed Airbus Flight Training) A320 course #1 ("FC1"). I was one of two ordinary line captains (each of us paired with a line first-officer) on course FC2 - the first BCAL/BA course - a week behind FC1. (It is sometimes forgotten in French aviation circles that there were three launch-customer airlines for the A320, and Air Inter was the lowest in the pecking order.) The rest of my course members were training and management pilots, and two pilots from the CAA. We did not mix socially with the AF pilots, which I now regret; but we would occasionally meet them briefly at the simulator platform on handover, or at the drinks machine.

After a couple of weeks of self-tuition on the VACBI, and passing the technical exam, we started the FBS (fixed-base simulator) - originally planned to be 14 three-hour sessions - on the third Monday in January. The two FBSs were hors-service (u/s), so the first half of the series was cancelled and we started at FBS session #8, using the FFS (full-flight simulator). I cannot remember if we used one or both FFSs, but I think only one was available most of the time. It or they were unreliable - throughout our 7 FBS and 7 FFS sessions. Neither of my two instructors for the 14 FBS and FFS sessions was A320 type-rated, as far as I know. We lost 1h40 on the first two FFS sessions, and regained 0h45 on the next three; lost 0h45 on the sixth, and regained 0h10 on the seventh.

We then went home to England to wait for the A320 to be type-certificated, returning to Blagnac in March for our type-rating training. The latter consisted of one FBS session for re-familiarisation; 3 FFS sessions with instructors who were A320 TRE/IRE-rated, during which we did our instrument ratings; and one base-training sortie on the a/c for our type ratings.

Quote:
"So You were type rated by Test pilots who had to do instruction and qualification?"

Most of the TRE/IREs were Aeroformation trainers, not test pilots. There were, however, two or three test pilots who were involved in our refresher sim sessions in March. They may not have been TRE/IRE-current, but I'm not sure. They included GC, UE, and maybe NW. (Later, GC and NW were closely involved in our line training from Gatwick.) My single sortie on the a/c for base-training and type-rating was with an Aeroformation training captain, Dick Steele (who later conducted my "final" line check out of Gatwick). We reached alpha-max at an altitude of 4000 ft...

Quote:
"Asseline said he trained his 100 FT height flight on the Thomson simulator (with sealed software) used to train AF first A320 pilots. Isn't?"

That's quite possible, as he would have been able to negotiate extra sim time in his capacity as chief trainer for AF. But I cannot comment on "sealed software". What is that?

Quote:
"Which informations and simulator quality had Asseline, Bechet, the BEA, about AoA in Alpha protection and flight laws?"

I think the simulator(s) had been increased in number and reliability by the end of June. I know little about the working of sims, but all the necessary information would surely have been available from the Aerospatiale EFCS design team to create the necessary algorithms? Perhaps someone else can comment.

You have clearly no idea what would be to reproduce such a flight for real with the necessary required level of detail, accuracy, and fidelity, to have any value. The flight over the Toulouse runway was not meant to be a "reproduction" but was merely a wish by members of the commission to personally assess the behavior of the aircraft during maneuvers near the ground, to see if the switching of control laws could cause difficulties for a pilot.
HN39 is referring to the 'graph' on page 63 of the PDF which refers to a simulated flight not a real one.

As you seem ready to give any value to those graph, how Bechet and Asseline would still keep the same speed after obtaining so different alpha ?
According to the poor quality ... graph we are presented, not if he had started the maneuver at the same altitude that Asseline did, and not 10 ft lower ...

Some remaining outstanding issues
 

Bonsoir Chris Scott,

Thank you for your long description of the A320 launch-customers pilots type rating. I could compare it to my own MD-80's TR at the same moment. Despite the MD-80's were already "old" aircrafts our TR had many similarities, but of course no flight with test pilots, not alphafloor flight. But like you, only one base-training sortie on the a/c, after VACBI, FBS and FFS. That only one flight was around 1 hour or less, verifying that the real plane felt still easier than the FFS, and to say we did a real flight !

I wanted to point that flying alpha prot or alpha floor on A320 FFS is reliable, as the FFS has the exact reproduction of the EFCS rules. That is very different from trying to test i.e. stalls who have never been done in flight, and whose behaviour is still unknown and could not be "reproduced" on the FFS.

My purpose was to say to DozyWannabe that the FFS simulation from BECHET/BEA was totally reliable, and consequently too that the Hazelnuts39's demonstration perfectly applies (about "sealed software", see (1)).

Continuing the demonstration, it suggests that the fact that pulling the stick three seconds later and getting in alpha prot which had not been previewed by Asseline -and missing in the short "briefing" (gums You are right, that was a very poor "briefing" !) - modified seriously the pitch control reponses.

I don't forget that Hazelnuts39's conclusion reminds that it was not enough to replace the late thrust of Asseline.

Asseline is writing many times in his book that he surely was wrong to trust his altitude. we know from the crash itself and the videos that the plane's real altitude was nearly the same that the RA. The question seems still to be open : did Asseline read 100 FT on his baro-altimeter, when the DFDR records an altitude complying with the Radar altitude, and Mazières telling him clearly about RA ?

The BEA report is saying too that the crew did not contest the FCS (at the time of the enquiry, which is some years earlier than Asseline's book).

(1) re:"sealed software" : Asseline said that working with Thomson they first asked to AI -or AS ?- they needed the copy of the EFCS software. Asseline said they recieved a negative answer, but AI-or AS?- accepted to give them an original EFCS software in a box they could not open or read or copy

I see no reason to doubt the accuracy of the simulator used by Bechet. He gained less than 10ft height in the five seconds after he went to full throttle and full back stick. That is equivalent to less than 1 kt of airspeed. In the accident flight the wind was light and variable (5 kt +/- 2kt headwind component), and in those last few seconds the airplane entered the lee of the forest. Those wind varations were evidently not represented in the simulator.

P.S.
Just before hitting the trees, the DFDR recorded RA was 24 ft, possibly even less because the recorded values are rounded to the nearest 2 ft. The average tree height is reported as 12 m (40 ft).

Thanks roulis,

My main purpose was to explain to anyone interested in these matters that - at the time Capt Asseline was doing his conversion course - the A320 sim was unreliable, and that zero-flight-time conversion was impossible.

Quote:
I wanted to point that flying alpha prot or alpha floor on A320 FFS is reliable, as the FFS has the exact reproduction of the EFCS rules. That is very different from trying to test i.e. stalls who have never been done in flight, and whose behaviour is still unknown and could not be "reproduced" on the FFS.

I think the a/c would have been stalled during the test programme. But it would not be necessary to incorporate any stall data into the FFS, because the aerodynamic stall is not part of the training syllabus. And, of course, it did not occur at Habsheim.

Quote:
re:"sealed software" : Asseline said that working with Thomson they first asked to AI -or AS ?- they needed the copy of the EFCS software. Asseline said they recieved a negative answer, but AI-or AS?- accepted to give them an original EFCS software in a box they could not open or read or copy

Thanks - that is interesting, and would make sense commercially.

Regarding the test flights at Toulouse, I'm going from Captain Bechet's interviewed statements on the ACI programme, and he at least seemed to be confident that it was close enough to the actual event (though obviously with some variation) to be useful. Other than that I make no claims whatsoever, other than HN39 (as usual) seems to be on the money regarding the effect of earlier vs. later application of full back-stick.

As for the software in the original simulators, it would have been standard practice at the time to provide complete and working hardware/software combinations as a unit - as long as the interface of the simulator and that of the aircraft was the same it should have been like-for-like. As such, the behaviour of the EFCS would have been one-and-the-same on the simulator and the real aircraft, but the simulator's flight model behaviour would still likely have been undergoing refinement. The "black box" (in engineering terms, i.e. meaning a 'sealed unit') was and still is for commercial considerations only - Airbus's competitors would have likely been very interested in the implementation of their systems, and - especially at that time - the fear of industrial espionage was a real one.

I know that the flight model was essentially fed with the data gleaned from the flight testing performed on the A320 prototypes, and I've been told that those prototypes were the most heavily instrumented and monitored machines of their type at the time. I don't know whether they were flight tested beyond the stall boundary, but wouldn't be surprised if they were at least taken as close as possible in order to gather data on airframe characteristics.

From my brief experience, I can say that modern A320 simulators do provide a plausible representation of behaviour going into stall and recovery, but can't vouch for how precise a representation that is.

In 2 sec the airplane gains 5 deg of AoA without altitude gain and without losing speed ... That A320 is a very special bird indeed ...

Yes, the atmosphere was calm.

Not what is in the report text and not what is on the graph you're commenting ...
How do you explain again such confusion in the report ... ?

The reports states that the wind was 5kt at the time of the accident. Comparison of the recorded airspeed and groundspeed shows variations of +/- 2kt of airspeed that can only be explained by variations of the wind speed.
The last value of 34 feet is obviously a typographical error and should read 24 ft as shown in the CEV print of DFDR data. In addition, the recorded accelerometer data show that the airplane cannot possibly have climbed 10 ft in one second between TGEN 333 and TGEN 334.
http://i.imgur.com/jOmWmjz.jpg

I thot Doze or another contributor tried to duplicate the stall/recovery for AF447 in the sim.

Unless early flight tests showed a poor pitch co-efficeint at a high AoA, as we had in the Viper, then an approach to stall at 20K would seem to provide great data for the simulator. No need to reach the AF447 complete stall conditions, but at least show the ability of the alpha protect and such to help the pilot. From the AF447 data, seems to me that the 'bus has poor nose down authority once past 15 or 20 degrees AoA. Our problem in the Viper was similar, in some ways. We had plenty of nose up, but our AoA limiter would not let us use it. Nose down was FUBAR due to the design of the jet and its pitch co-efficient at 40 - 50 degrees AoA. Hence, we had the manual pitch override switch that would only work if AoA was above 30 degrees.

I still feel the accident, if you can call it that, was the result of poor planning and practice, and improvised changes in the "plan". I won't comment any further on the airmanship of the pilot.

1. If it was one they would not have used it for the graph drawing
2. Tome 2 is more inclined to depict 34RA for TGEN334 than 24 ...

http://i55.servimg.com/u/f55/11/75/17/84/image12.jpg

Still nothing new here gums - We still all agree - But what we're looking for is the FULL story not only the part on the crew.
As HN39 is avoiding the question, you probably could bring your thought on that :

That is not what I said in post #622 based on the graph in the report. In the simulation the airplane gained 10 ft and that is equivalent to 1 kt of airspeed.The graph is a computer/plotter generated record of the simulated flight. How accurately were the plotter pens aligned with the millimeter raster on the paper ? The scales and the Habsheim points have been handwritten on that plot with lesser accuracy. The wind at Habsheim was variable. Thus a difference of one or two knots is easily masked by other factors. It is ridiculous to dismiss the simulation for no other reason than that you can't see a difference of one knot on that plot.

"It's a special bird indeed"

Well… Maybe.

During that time was it gaining thrust? If so, then increasing the angle of attack, and thus drag, keeping the potential and kinetic energy constant, but matching the increasing drag and thrust? That doesn't seem an unreasonable proposition.

Documentation discrepancy of Rad Alt datum at TGEN 334
 

Hello Confit,

I think many of us were already aware of the discrepancy between the figure of 34 ft in 1.11.4 of the BEA "Rapport Final", and the 24 ft shown in the DFDR "Tome 1" table of the same document.

Thanks for bringing to our attention that DFDR Tome 2 of the same document appears to show a figure of 34 ft, in line with the narrative. The version you show magnified in your post is, I presume, the PDF copy, the poor quality and assembly of which we discussed earlier on this thread.

Another curious aspect of this matter is that an HTML version of the BEA report - Habsheim F-GFKC - clearly shows a figure of 24 ft in both Tomes 1 and 2.

TGEN 334 seems to represent the last second of flight before the RA antennae (rear fuselage) reached the treeline. The ground between the northern extremity of the grass Rwy 34R and the treeline - a distance of about 70 metres - seems to have been flat, with minimal slope. One cannot, however, entirely rule out the possibility of a vehicle or other temporary obstruction being present within the wide, conical lobe of the aircraft's RA TRx. As the ground-speed of the a/c was slightly less than 60 m/s at that point, the RA values at both TGEN 333 & 334 could have been affected by vehicles or superstructure between the grass runway and the treeline.

Therefore, determining the precise RA datum at TGEN 334 would not in itself enable us to prove the trajectory of the a/c in the final second of free flight.