Too Slow, Too Low and Obstacles Ahead: Air France flight 296

6 Jul 18 19 Comments

On the 26th of June in 1988, one of the first in-service Airbus A320s flew a demonstration flight over Habsheim airport and ended in tragedy. I’ve written an introductory piece which explains the background along with the technical issues which are required to understand what happened that day.

Rudy Jakma has mentioned he has some information which may differ from the BEA sequence of events, so be sure to check out the comments to this piece.

At the time, no one saw a reason to ban passengers from a demonstration flight. A raffle was held at the Habsheim Airclub and the winners, along with a number of journalists, were invited for the flight, making for a total of 130 passengers.

The flight plan was for the Airbus A320 to board the passengers at Basel and then fly to Habsheim runway 02/20. There they would overfly the airfield twice at 100 feet above the ground: once low and slow at the maximum angle of attack and a second at high speed. Then the aircraft would fly to Mont Blanc to circle the famous peak before bringing the passengers back to Basel.

The flight crew were both experienced and the captain, a test pilot for Air France, knew the aircraft as well as anyone. He had pushed it beyond its operational limits and found the Airbus A320 could do things he’d never been able to do. He was confident about the demonstration.

The co-pilot, acting as Pilot Monitoring, had 25 years experience at Air France, six of which as a training captain. He had over 10,000 hours flight time and had qualified as captain on the A320 three months earlier.

Air France’s Operations Director and the Flight Safety Department had approved the demonstration.

No one ever saw any need for a test run.

The two pilots received their flight plan the morning of the flight. No details of the aerodrome or the flyover were included.

They flew from Paris to Basel where Air France held a press conference. Then the passengers boarded and the captain briefed his co-pilot. They would take off from Basel, retracting the landing gear but leaving the flaps at 1 (10° when going below 100 knots). Then a nice and easy turn to the right and look for Habsheim.

Because Habsheim was a small airfield with only a 1,000 metre runway (3,280 feet). It was not listed on the flight computer; the flight crew would need to locate it visually and then prepare for the low and slow pass as they descended.

The captain told his co-pilot that, as soon as they had the runway in sight, he should take the flaps to 3 (20°) and extend the landing gear. Then he would descend to 100 feet and disengage the alpha floor protection, so that the aircraft would not increase the thrust as they approached stall speed. They would then pass 100 feet over the runway at just above stall speed.

The co-pilot’s monitoring duties included calling out when they’ve reached 100 feet and watching the instruments.

The captain told him, “if I tell you that it is hard, you help me, and you hold the power to keep zero vertical rate.”

Flying at right at stall speed, the aircraft will become sluggish. At 100 feet above the ground, it’s important that they don’t lose any height at all during the low pass. The co-pilot is thus briefed to watch the power settings to make sure that the aircraft doesn’t sink and that the captain isn’t struggling. Because the captain is disabling the alpha floor protection, he has to rely on the co-pilot to adjust the engine thrust so that they can maintain straight and level flight. They didn’t discuss airspeed but a typical approach speed on the A320 is 130-140 knots; I assume that they had planned to stay at the low end of this.

Then, at the captain’s signal, the co-pilot is to increase the thrust to TO/GA power. As they climb away, the captain would turn, ready for the high-speed pass at 340 knots (391 mph/ 629 km/h).

The briefing was clear and covered the details of the flight. The captain split up the tasks and made sure that the co-pilot understood the plan.

But things didn’t go according to plan.

After the briefing, the A320 departed Basel and climbed to 1,000 feet and turned right, just to find that they couldn’t see the airfield. It was 8.6 nautical miles (16 km) to Habsheim, plenty of time to set up descend; however, by the time they’d made visual contact, they were only 5.5 nautical miles (10 km) from the airfield. They throttled the engines back to idle in order to allow for the faster and steeper descent.

They extended the landing gear and selected flaps 2 and then, a few seconds later, flaps 3 as they descended through 500 feet. Everything was happening very quickly.

At 200 feet above the ground, the flight crew suddenly saw that the crowds were not at the concrete runway (02/20) for which the demonstration was planned. Instead, they were lined up along the sides of a grass strip at a 40° angle from the main runway. Descending at 600 feet per minute, the captain quickly turned right to align the aircraft to the grass strip.

This must have been insanely disorientating.

The aircraft reached 100 feet above the ground. The pilots had no time to think about their fast descent at idle power or the changes to the plan. The co-pilot called out 100 feet (monitoring their height above the ground) but the captain never responded. He never gave the command for the co-pilot to increase thrust to keep them at 100 feet.

The radio altimeter also called out 100 feet. And then 50 feet, 40 feet, 30 feet. The captain finally levelled out somewhere between 35 and 30 feet above the grass runway. The speed decayed to 122 knots.

The aircraft is now too low and too slow and the alpha floor protection is turned off. There’s nothing to save them from a stall but themselves but both pilots must have been struggling to process what was happening.

To make matters worse, the grass airstrip had trees at the end: trees which were not marked on the chart and that the flight crew didn’t know about. The trees weren’t a problem for aircraft taking off from the runway; under normal circumstances an aircraft would climb away to 50 feet above the ground as it cleared the runway and pass over them.

But this was a low straight-and-level flight along the length of the grass strip, planned for a runway which was clear of terrain on both ends.

The aircraft was set up for the low pass: wheels down, flaps fully extended. Their airspeed was only 122 knots. They were flying lower than the surrounding obstacles. And their engines were at flight idle.

It seems pretty clear that the captain applied TO/GA power at this point. The captain later said that he applied the power and nothing happened, so he had to pull back and apply it again, losing precious seconds. But this doesn’t match the logs on the Flight Data Recorder or the sounds on the Cockpit Voice Recorder where the clicking of the power being increased is clearly audible.

The problem is that the total energy available to the aircraft at that moment was not enough to clear the trees. The angle of attack was already near maximum, intentionally as close as possible.

The captain increased the pitch but without the engine thrust, that aircraft could not climb. The engines needed more time to accelerate to a useful level of thrust.

For certification, an engine must be able to spool up in eight seconds. For the first four to five seconds, there’s very little increase in thrust, it all comes at the end. On a normal approach, the thrust is stabilised, usually above idle. The pilots are targetting a specific approach speed and only minor changes are needed. In this case, the speed had been quietly decaying since they started their descent.

The aircraft flew into the tops of the trees. The engines, almost to the point of spooling up, began to suck in the leaves and small branches. At 4.5 seconds after the application of TO/GA power, both engines failed and the aircraft impacted the trees. The right wing was torn off and the fuel tank breached, spilling jet fuel which immediately ignited.

Inside the aircraft, the passengers were dazed and confused, having hit their heads on the backs of the seats in front of them. The senior cabin crew membmer attempted to give instructions for an emergency evacuation, but the PA system had broken on the impact. He opened the left-side door which only partially opened, blocked by the trees. The emergency escape slide immediately began inflating, still stuck inside the fuselage. The cabin crew and passengers forced the door fully open but fell from the aircraft. Another cabin crew member attempted to evacuate the remaining passengers but they were unable to get clear of the trees and branches at the bottom of the slide. The cabin crew member continued to evacuate the passengers while those below cleared a route for passengers to move away from the aircraft. Cabin crew members at the rear also evacuated passengers as quickly as possible but the overwing emergency exits weren’t usuable for those in the middle of the aircraft.

By now a fierce fire raged on the entire right side of the fuselage and the fuel in the left wing had ignited. The fire breached the cabin and passengers trying to push forward from the middle of the aircraft were burnt and suffering from smoke inhalation. The fire trucks could not make their way to the scene, blocked by the trees. The aircraft was destroyed by the fire.

Three passengers were unable to escape the burning aircraft: two children and a woman who turned back to help them. Of the 133 who were successfully evacuated, 34 were treated for burns and injuries. Honestly, the fact that anyone actually survived this is amazing.

The investigation found that the accident was unavoidable from the moment that the aircraft passed the tower: by then, it was too late to get up and over the trees.

Captain Asseline, First Officer Mazière, two Air France officials and the president of the flying club sponsoring the air show were all charged with involuntary manslaughter. All five were found guilty. Asseline was initially sentenced to six months in prison along with 12 months of probation. The others were sentenced to probation.

The captain argued that the barometric altimeter was set wrong; he said that it displayed 100 feet when he went to climb away and so he never knew that he’d descended below the target height. Even if this was true, flying at such a low height, he should have been using the radio altimeter for his precise height above terrain, but he said it was hard to see.

This also may have been true but when it comes down to it, his job was to be looking outside the aircraft and relying on the co-pilot to monitor the instructions. The co-pilot (and the radio altimeter) called out 100 feet while the aircraft was still descending but the captain, likely fixated, never heard it.

The captain also argued that he would have cleared the trees but the aircraft forced him down. He said that the alpha protection wouldn’t let him climb.

Airbus attempted to recreate the situation as closely as possible, but on a much longer runway which was clear of terrain at the departure end. They found that they were able to recreate the aircraft elevators moving down (as the pilot said) as a part of the alpha protection. Although Airbus confirmed that the elevators moved down as he attempted to clear the trees, that’s not damning proof that the Airbus system is flawed. It simply underscores what we could already see from the FDR data: that aircraft was not capable of flying over those trees in the configuration it was in.

The captain’s argument shows a flawed understanding of the alpha protection. The alpha protection wasn’t stopping him from climbing, it was stopping him from stalling. If the aircraft had been capable of climbing, it would have climbed, even an Airbus. Especially an Airbus.

If the trees hadn’t been there and the aircraft had been flying over the concrete runway, ending up in the same position, then because it was a fly-by-wire aircraft with alpha protection, they would have got away with it. The end of the runway was clear of terrain, so they would have just needed to keep straight and level until the engines spooled up and the aircraft gained airspeed, at which point they could climb away. The alpha protection, in this case, would allow the pilots to do this with maximum efficiency and ensuring that they recovered from the stall.

If the aircraft had levelled off at 100 feet, as planned, there would not have been an issue, even if the power had not been increased until the very last minute, because they would have remained clear of the trees.

If they’d been flying a Boeing, it would have sounded alarms at the flight crew. It would not have made it over the trees. It couldn’t. The aircraft would have stalled as it crossed the tree line.

The Airbus overrode the pilot’s setting in order to prevent the stall. It couldn’t prevent the crash.

The accident was caused by a combination of factors:

  • no overflight or practice runs for the demonstration, so the pilots did not know the details of the airfield
  • the A320 was configured for slow flight, not for climbing away
  • the organisers changed the target runway at the last minute (or at least, herded the crowds to the wrong runway)
  • the flight crew was distracted with their late descent and change in plan at the last moment
  • the airspeed had dropped to below what they had planned
  • the aircraft was only 30 to 35 feet above the ground
  • the engines were at idle
  • by the time go-around power was applied, it was too late

As a result of this crash, passengers are banned from demonstration flights and flight crews are expected to receive the proper reconnaissance of airfields, although the Sknyliv air show disaster in 2001 suffered this same fault, where the pilots did not know the airfield and, at the last minute, realised that the display line was not where they expected.

It’s not often that an accident report looks like poetry but I could not resist sharing this excerpt from the Airbus report with you.

Airbus concluded that by the time the captain pitched the nose up, there was nothing that could have been done to avoid crashing. The only question was whether the aircraft would crash nose down or nose up.

Airbus theorised that the captain had became overconfident after the success of his test flights in the A320. The aircraft had so precisely done everything that he wanted it to do that he hadn’t realised that he had pushed the aircraft beyond the limits of its ability to fly.

There’s another discussion to be had here about Just Culture, where front-line operators are not punished for their actions or decisions within the scope of their role and training. Specifically pilots and ATC are at great risk of being ‘punished’ so that the public can see that something has been done, restoring their faith in safety. It doesn’t mean that no one ever can be prosecuted but it means that there needs to be a case of gross negligence. If the captain had decided that he was going to fly at 30 feet for the sake of it, when the organisers told him to stay 170 feet above the ground, then there’s a case for negligence. If he had refused the opportunity to overfly the runway and do practice runs and then changed runways on the day and then purposefully flown at 30 feet – yes, I would be more open to the idea of criminal charges.

But the captain had not simply decided to try a risky manouevre of his own volition. Air France planned and approved the fly-over without offering the appropriate support. There were no practice runs. The flight crew were given the flight plan at the last minute. And then the unbelieveable last minute changing of the runway, a decision most certainly made by someone who simply did not understand or respect the aircraft and the flight crew.

Personally, the only actual negligence I can see is on the part of the organisers, who clearly did not understand the impact of what they were doing.

Another aspect of Just Culture is that the investigation, as handled by the transport authority responsible, is kept separate from criminal investigations. There was a recent case in the UK, where the AAIB refused to give up the details of their investigation to the police, because they argued that they would lose the trust of the community. They would be unable to conduct interviews and investigations in future if people thought that the details were going straight into a criminal investigation.

In this case, I feel that a lot of the controversy about this flight and what caused the crash comes down to the French decision to prosecute and sentence the flight crew for manslaughter.

The official report gave the following probable cause of the accident as being a combination of:

  • Very low flyover height, lower than surrounding obstacles.
  • Very low speed, slowing down to reach maximum possible angle of attack.
  • Engines at minimum idle flight power.
  • Late application of go-around power.

The BEA concluded that although the descent below 100 feet was not deliberate, it showed a failure of the crew to take proper account of the visual and aural information available to them (that is, they should have been able to hear the radio altimeter and see how close they were to the ground.

The following are the references I used for my points and analysis.

I expect there to be some robust conversation so I’d like to remind people that I expect civility and calm in the comments. I will not publish comments which engage in name calling or attack the intelligence of the other posters.

Category: Accident Reports,


  • OK folks,
    What I will write now is HEARSAY, and I have not been A320 rated, so you must accept that it may be totally and utterly untrue. At the time, many stories were told, including some that suggested that Air France and Airbus Industries colluded to hide the facts. However, the concept on which the Airbus design and it’s fly-by-wire philosophy was based is sound and, provided fully understood by properly trained pilots, has the potential of adding substantially to air safety.
    So why does an at the time ultra-modern aircraft, flown by a highly experienced crew, crash during a fly-past at an airshow?
    I have, now many years ago, done sightseeing flights over the Dutch tulip fields. The aircraft were Fokker 50 and, of course, we flew VFR.
    The F50 is not nearly as sophisticated as the A320, but if we descended below 1000 feet GEAR UP a cockpit warning would sound. The warning would be generated by the radio altimeter.
    The version of the AF A320 crash, as told to me by a pilot who was A320 rated was as follows:
    The intention had been to make a low pass at Habsheim with the GEAR UP. Low being one THOUSAND feet, not 100.
    Flaps 1 or 2 with gear retracted, engines spooled down and stick fully back would cause the aircraft to maintain 1000 feet rad.alt.
    Approaching the stall the aircraft alpha protection would prevent the aircraft to descend any further, automatically increasing power and the aircraft would continue to fly at a pitch angle looking impossibly high.
    The A320 was a brand-new type, just introduced. The crew were highly respected, highly experienced but obviously had recently transitioned from the B737.
    The crew therefore were relatively new to the A320 and may have overestimated its possibilities and underestimated the limitations.
    Not very comfortable with VFR possibly, they did not line up properly for the low pass and (again, this is conjecture !) may have come in a bit too fast. In order to solve this, they extended the landing gear.
    Doing so, they removed the protection that would keep the aircraft flying straight and level at 1000 feet. With engines spooled back, in landing configuration, the aircraft was now programmed for landing. The protection does not mean that the aircraft knows that the terrain ahead is not suitable for landing, it just does what it has been programmed to do.
    The aircraft now was sinking below the one thousand feet protection, configured for landing with engines still at idle but also at a speed that was very low. The protection would still apply to stall avoidance.
    When the pilots started to realise that the sink rate would bring them to an unintended touch-down they must have added some power, probably still with a mind-set of making a fly-past. When it became evident that they were heading for the trees the captain belatedly activated TOGA.
    Not only did the engines need time to spool up – time they no longer had – even if they had, an aircraft at very high angle of attack is flying at the wrong end of the so-called “power-drag curve”. After maximum power has been established, the aircraft first must accelerate in order to have sufficient lift to climb.
    My own experience, coupled to what I have been told makes little sense to me of the version Sylvia published and which no doubt has in turn been based on official reports.
    To attempt a low pass at 100 feet, gear down, engines idle, will probably generate rad-alt warnings like “one-hundred”, or maybe “too low, flaps”. As the aircraft was configured for landing there would NOT have been a “whoop-whoop TERRAIN” or “PULL UP”.
    There would have been routine call-outs from the system to inform the crew that they had programmed the aircraft for landing. But landing was something they had not intended to do.
    No, I still do not buy the version as published here.
    Sylvia did an excellent job describing the operation of the A320 and it’s protective protocols (or “laws”).
    I may well be ignorant of a few facts.
    But to attempt a low pass as described, deliberately, would have been suicidal.

    • I think that Airbus certainly had reason to want to bury this. However, the accusations that the CVR was tampered with all center around when the captain put the power back on and whether the engine failed to spool up — the captain never denied the briefing, which clearly states one hundred feet and to extend the wheels.

      In addition, the captain’s counter attack specifically says that the barometric altimeter was fauly, showing them at 100 feet and not at 30 feet, so he couldn’t have known that the aircraft was so low.

      So although I agree there are things that make no sense, I can’t see where 1,000 feet comes in at all. It seems clear to me that the captain agreed to attempt a low pass at 100 feet and briefed it with the gear down.

      The radio alt warnings were recorded. I don’t know why the others weren’t, but it was only a few years after the Cali crash, which brought the focus on EGPWS, so maybe the first A320 didn’t have it sophisticated enough or that it was specifically inhibited (just as the alpha floor was) by using landing configuration.

  • Or in summary:
    A deliberate very low pass at very low speed with a very low power setting, in a high drag configuration over an area with limited space and obstacles at the end, in an airliner flown by a crew with limited experience on type is a set-up for a near inevitable accident.

    • And yet almost every multi-engine rated pilot in the world will insist that given sufficient time to prepare, with foreknowledge of how hazardous it is, and by learning from the Habsheim tragedy, this maneuver can safely be accomplished.

  • In line with Sylvia’s Just Culture comments: What would have happened to the captain if they had said “nope, the crowd’s not where they’re supposed to be, and you do not get to throw a runway change on a demo that’s this tricky already”, and called off the pass entirely?

    Because as far as I can see, that’s really what he should have done!

  • Another catalogue of errors!
    The thing I find so sad though is the phenominal number of fatal “accidents” where the pilot would have “got away with it” if gas turbines didn’t take anything up to 8 seconds to spool up! Where seconds count, a gas turbine can so easily kill you! eg any last minute decision to go around!

  • Sylvia et all,
    My assessment was not based on the reports Sylvia published but on the assessment of an A320 rated pilot and my own experience as a pilot.
    – Even the lowly Fokker 50 had a protection that would warn the crew if the aircraft would descend below one thousand feet, gear UP.
    – Extending the gear removes that protection. The crew are configuring the aircraft for landing, it will accept that the crew intend to make an approach and will allow it to descend to either minima or landing.
    – Especially so if flaps are being selected.
    – To rely only on the QNH (or QFE) for a low pass in landing configuration makes no sense, in such a case the rad alt would and should have been used.
    – Has anyone ever tried to make a low pass in a medium-sized aircraft at exactly 100 feet? No? I thought as much! It is very very difficult to do this without either losing a bit of altitude – altitude that is now in dangerously short supply, or oscillating around the 100 feet..
    – The aircraft is set-up for LANDING, even more so with a reduced power setting.
    – But worst: the protection floor has been reduced. The A320 fly-by-wire still would prevent a stall, but if the aircraft has a bit of excess speed and the crew were pulling power back to demonstrate how it would be able to fly, high pitch-up, without stalling, the aircraft would have continued to sink with rapidly decaying speed.
    – It would increasingly rapidly build up drag, drag that would be needed to be reduced first before the aircraft would be able to climb away again. With the engines at flight idle the flight would have been set up for “the perfect storm” or, rather, a now virtually unavoidable crash.
    The approach speed Vref is based on 1.3 Vs or stalling speed for that very good reason. Steve, a turbine-powered aircraft set up in a stable, well-coordinated approach, would have
    a. A bit of excess speed to cope with wind gusts and even sloppy flying and
    b. A reference power setting which is HIGHER than flight idle. In other words: already spooled up.
    A jet aircraft in a stable approach would have a rather high power setting. In the Citation 500 it was about 62-65%.
    In order to prevent the speed increasing, a typical jet will fly the approach with a lot of flaps hanging out.
    So, if a sudden go-around is necessary, the spool-up time will be a lot less than 8 seconds, more like 2. Selecting TOGA will nearly instantly give the required power setting, followed by a part-retraction of the flaps to give maximum lift and minimum drag. So normally the aircraft should be able to climb away at once.
    And how in the name of all that is holy did the crew not have a briefing that incorporated the obstacles (trees) at the end?
    This all still does not make any sense to me.

  • Dave: I agree calling off the passes would have been safely cautious. (cf “Plan the dive; dive the plan.” for scuba.) However, do you know the case of William Glen Stewart? He was a BA captain who safely landed a squirrely 747 at Heathrow in bad conditions after a bad missed approach; he should have diverted to Manchester but knew he’d be penalized by management for costing them money, because the idea of the captain being the absolute final decider is long gone. He had no audience; the crew here did, so they probably figured they’d be in serious trouble if they didn’t find a way to make it work. (Stewart was also convicted (of endangering passengers), possibly because he had the temerity to challenge being downgraded.) Sylvia’s description does suggest to me that this crew should have taken a longer approach (to reduce descent rate) after changing plans, but I get the impression that western-European airspace is intensely crowded; they may not have had airspace to do this.

    Rudy: it’s not clear there were trees at the end of the runway that was supposed to define the pass; given the complexity of the flight plan and the near-certainty that nobody would try to land a 6-across jet on grass, I wouldn’t be surprised if the information were left out of the planning. It should have been on a chart (physical or electronic), but looking for hopefully-irrelevant information was probably the last thing on the crew’s minds.

    Michael: whose arrogance? Having toddlers on a demo flight does suggest to me that Marketing (possibly in multiple companies) should have been sacked, but they had plenty of front-line people (pilots and ground organizers) to blame; can you see the circular firing squad? The Wikipedia article says that barring passengers from demos was one of the recommendations in the final report, but I don’t find whether such was actually outlawed.

  • Rudy,
    I accept what you say but I still say that there are far too many disasters, many fatal to passengers, where the instant application of extra thrust would have saved the day.
    A pair of inexpensive rocket boosters activated by extra pressure on the throttles would solve the problem. Sack the pilot if used maybe but at least save the passengers!

  • From the reactions I think that many do not quite understand where I am coming from.
    From a pure aviation perspective, this flight was a disaster in the making.
    Absence of knowledge about the obstacles (trees) is NO excuse. Air France is a major airline and could – and should – have made sure that all relevant information was available and promulgated to the pilots.
    The crew were new to the aircraft. After all, it was a brand-new type just recently introduced to the airlines.
    The A320 has a lot of safety features that, provided that the aircraft is properly operated, by a well-trained crew, will introduce a whole new safety platform.
    If the intention had been (as now seems to have been the case – to my utter surprise !!) to make a low pass at 100 feet, gear down, an important safety had been knowingly and deliberately removed.
    Transport aircraft are not designed for such a manoeuvre. In modern days, even during a non-precision approach, the idea of flying level at the minimum descent altitude (which is NEVER as low as 100 feet !) has been discontinued in favour of a continuous descent to a missed approach point or VDP. After which the crew must react in the same manner as during a precision approach when insufficient visual clues mandate an immediate missed approach.
    Jet aircraft typically fly an approach at a high power setting and with a flap setting beyond simply giving optimum lift: they are in addition supposed to add DRAG.
    The resulting high power setting required to overcome this dag allows for a quick spool-up in the event of a missed approach, the high drag flap setting can be instantly eliminated by retracting them to the optimum high lift / low drag configuration.
    In addition, the Vref is calculated to be at least 1.3 Vstall, often with a bit extra for gust correction.
    So in the event of a go-around the power will be there nearly instantaneously, the wings will provide lift, the .3 safety margin above the minimum Vref will ensure that the aircraft will have a little bit of excess speed to climb away.
    To allow a crew new to the type, over an airport they were not familiar with was at best lunacy.
    At worse? I do not want to get solicitors’ letters from the lawyers representing a certain airline.
    So Steve, your suggestion normally should be totally unnecessary.
    I will remind you once again of the warning that used to be placarded in the aircraft belonging to the Tiger Club, in the old days at Redhill and possibly still today:


    No rockets will safeguard against foolishness. Which in my considered opinion had been built into this display from the beginning.

    • I agree that the stunt, as described by Sylvia, sounds extremely dangerous. Even military pilots in aircraft designed for low altitude operations rarely fly that low, and when they do they use considerable thrust and special equipment to look for obstacles.
      I had a thought/ question on the discussion of barometric altimeter vs. radar altimeter: With a significant angle of attack and at low altitude, the height difference from the ground of various parts of the plane will start to matter – ALOT! I don’t know how or if the rad alt compensates for angle of attack or where the lowest part of the plane is. I wonder if the crew was using the barometric altimeter because they knew they couldn’t trust the radar altimeter in this situation. Rudy, do you know how or if a radar altimeter compensates for angle of attack, or if it assumes the aircraft is level?

      • As far as the captain’s defense went, the reason he gave for not using the radar altimeter was that it was hard to read. I didn’t see any justification and I’m pretty sure I’ve never seen someone argue *against* using the radar altimeter for low-level flyovers other than varying terrain.

  • Once again Rudy, I totally agree with what you say. My point is that far too many pilots, and this is yet another example, have got themselves into an irretrievable ABNORMAL situation through stupidity, lack of experience or bad planning where the slow spool up time of the engines has contributed to a disaster that might have been averted with the availability of instant emergency thrust.
    So I am saying that rockets absolutely COULD safeguard against foolishness, there’s a lot of it about! Look at all the cases, some of them in Sylvia’s examples, of experienced pilots who made a last minute decision to go around too late!
    RAF fighter pilots during the war had an emergency option to push the throttle past the normal gate limit for extra thrust, this is in effect the same thing.
    Even the Tenerife disaster just might have been avoided had there been enough visibility to allow time to ignite emergency boosters!

  • This crash has interested me since day one. It has been seen as an indictment against fly-by-wire systems but it’s not that simple. As I see it, the entire flight was poorly conceived. Who, in their right mind, would think that a 100’ flyover by a medium sized transport airplane filled with passengers was a good idea? If someone wanted to accomplish such a feat, why would they plan it without, at the very least, a familiarization flight over the airfield?

    One of the things I was taught early on is that you can’t make a good landing from a bad approach. When the pilot saw that he was aligned with the wrong airfield, he elected to “save his approach” via last minute maneuvering and, while this was a fly-by and not a landing, as I see things, the principle was the same. By circling and making another approach, he would have reduced the workload in the cabin. That alone could have made for a better outcome.

    Whether the fly-by-wire system was at fault, which seems to be a matter of dispute, one thing is for certain, over reliance upon automation can lead you into trouble. The whole point of this exercise was to demonstrate the ability of the fly-by-wire system to keep the plane stable in a manner that a human would be hard pressed to match. It did that; it truly kept the plane stable, but . . . it did not prevent the pilot from flying into a very dangerous situation.

    As I see it, this tragedy was preventable on many levels and three lives were cut short unnecessarily. First off, don’t take passengers on flights which deliberately involve risky maneuvers. Secondly, leave the risky maneuvers to the test pilots. Thirdly, don’t ever put yourself in a situation where a systems failure can leave you in trouble with no time to react. A system to prevent stalls makes a lot of sense, but why tempt fate by voluntarily entering a situation where it is needed. Once again, leave this sort of thing to Flight Test.

  • I greatly enjoy Sylvia’s postings on her web page FearOfLanding. They are well-written and interesting. I have read all the comments re. this page re. the Airbus A320 crash at Habsheim France in 1988 and they are very interesting and well-written.

    Just as a simple, non-technical question from me, a lady who loves airplanes and only has my private pilot’s license, it says the aircraft flew by at around 30 feet, and I wonder how many feet were below the landing gear, which of course were hanging down from the belly?

  • I forgot to add that I’m astonished that only three people lost their lives in this crash (and of course any loss of life at all is a tragedy).

  • Planning a 100’ AGL flyover at alpha max with 100 pax in the cabin speaks volumes about the mentality at Air France in those days (I’m French). That sounds ridiculous nowadays, CRM having done a lot to make us pilots realize that we are not almighty nor invulnerable. The captain’s ego never recovered from this rough realization in Habsheim that day and he never accepted any responsibility for the accident – which also speaks volume about his sense of responsibility and arrogance. Blame the manufacturer, it will always influence the general public…

Post a comment:

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.