If G-THOF Were An A320...

[Winnerhofer]

Would it have passed without incident thanks to AlphaFloor, A/THR and AutoTrim?
Air Accidents Investigation: 3/2009 G-THOF

[A4]

On the Bus the sequence would have been something like this:

1. At platform altitude, GS LOC armed with A/THR in speed mode, T/LVR's in the CLB Gate. Target speed would probably be "S" speed (Flap 1). On the Bus once the APPROACH Phase is activated and the speed is "managed" (as opposed to "selected" by the crew) the target speed automatically reduces with additional flap selection back towards VAPP (final approach speed).

2. Established on GS and LOC selection of Flap 2 would result in a speed target of "F" speed and A/THR would reduce thrust to achieve this and increase when approaching target speed.

3. Failure of the A/THR at this stage would result in a single chime Master Caution and an ECAM. Crucially though, because the A/THR has disengaged in a non-standard manor this would result in THR LOCK i.e. The thrust remains at it's last commanded value with a flashing amber annunciation in column 1 of the FMA. Of course if it failed at idle thrust would be locked at idle. To recover from THR LCK the T/LVR's only have to be moved fractionally out of the CLB detent to instantly recover manual thrust. So if in the CLB Gate, cracking them back just 5mm will result in near CLB thrust.

4. If the disengagement/failure is not noticed the Bus will behave like the Boeing in that the AP will try to maintain the Commanded path (GS) - BUT as soon as it enters the low speed regime the protections will kick in. So, with A/THR off/failed and continued configuration to gear down and flap full, the speed would decay and the AP AUTOTRIM would increase pitch to maintain GS. As speed reduces below VLS and towards V Alpha Prot (Alpha Protection Speed), AUTOTRIM will STOP and the AP will disconnect with associated master warning. With no pilot input, the aircraft would now descend at IDLE at V Alpha Prot - sort of "mushing down at high alpha but not stalled. If the crew attempt to arrest the descent with use if sidestick only (pull back), speed will decay further towards V Alpha MAX - this will result in ALPHA FLOOR engagement AS LONG AS A/THR IS AVAILABLE and irrespective of the T/LVR position.

ALPHA FLOOR provides TOGA thrust and even if the crew hold full aft stick the aircraft will not stall and will climb away at ALPHA MAX. Recovery from ALPHA FLOOR is achieved by lowering the nose, accelerating above VLS which then results in TOGA LOCK - DISENGAGE AND RE-ENGAGE A/THR to recover normal A/THR operation.

The above assumes the aircraft is in NORMAL LAW. Another benefit of NORMAL LAW is there is no pitch thrust couple so the application of TOGA does NOT require a PUSH on the control column/sidestick

Additionally, the pitch limitations would not allow the aircraft to go beyond 25 in CONF Full (20 low speed) or 30 in CONF 0-3 (25 low speed).

So, whilst it's not impossible that a Bus could find itself in a similar scenario to the 737, the protections and pitch limitations should significantly reduce the chances of it ending badly. This incident emphasises the critical monitoring of the aircraft systems (FMA) and also the speed / thrust interaction as deceleration and configuration towards final approach speed occurs. These crucial few seconds should be closely watched by both crew - DONT get distracted by trivia - ON slope, ON speed, THRUST UP. The Turkish 737 at AMS is a terrible example where 3 crew failed to spot the speed decay and lack of A/THROTTLE response before it was too late

Hope that helps.

[Piltdown Man]

This incident demonstrates the criminal design inherent with most aircraft A/P - Trim systems. Can someone post a good reason as to why the trim system should carry on trimming (nose up) when the aircraft is below "bugged" speed. That is about as much use as a chocolate teapot. The elevator trim logic should be such that if the aircraft is below bugged speed by a certain value, additional nose up trim is prevented. Then, if the A/P finds the elevator loads too great, it should fail - leaving the aircraft in a flyable state. The Turkish 737 at AMS had the same problem. It would also be nice to know if the 777 has the same trim logic (I'm thinking of Asiana at SFO here).

This nasty trap lives inside most aircraft. It should be removed.

[Fursty Ferret]

The other thing you'd get on the approach from an Airbus is an auto-callout of "SPEED, SPEED, SPEED" once the airspeed dropped below the point where power is required to maintain a stable flight path.

[Winnerhofer]

AutoTrim

By T.S.
The autotrim complaint is nuts though.
If you don't keep trimming you run out of elevator authority in one direction. Autotrim preserves the crew's pitch authority in both directions for as long as possible, on the assumption that the crew knows better than the airplane on what's actually required in a given situation.

[Denti]

After the Turkish accident an additional "Airspeed Low" aural alert has been added on the 737NG. It activates at the same moment the airspeed low alert in the MCP speed window becomes active if the airspeed is in the lower amber caution range.

[TowerDog]

Back in the old days we used something called scan to monitor airspeed:
It worked quite well actually, try it instead of blaming accidents on some auxillary system that may or may not work...

[fmgc]

Back in the old days we used something called scan to monitor airspeed:

And there were a lot more accidents in the old days!

Technology has improved safety enormously but it's not fool proof.

[misd-agin]

There wasn't enough available indications?

Airspeed indicator.
Speed tape color change.

Where are we going in this industry if one of the most critical instruments is ignored and then considered 'not enough available indication'?

[RVF750]

Can't help but agree totally on the last. A good scan is essential to any pilot's basic skills. If you can't do that, WTF are you doing sitting up front in an airliner?

[Check Airman]

If you can't do that, WTF are you doing sitting up front in an airliner?

...or a C172 for that matter. How many more bells and whistles do we need? A modern aircraft will give a visual, aural, then (excluding Airbii) tactile warning that you're getting slow.

One wonders how those courageous Cessna 172 pilots manage to keep those things in the air with only an airspeed indicator and a stall horn.

[cosmo kramer]

While I completely agree with the posts, that a proper scan is essential - with thousands of aircrafts takeoff and landing each day, it is a matter of statistics that sometimes, something, for some reason will be missed, like in this case.

For that reason I agree with Piltdown... Why not design the autopilot to disengage when reaching Vref -5 knots (or another value as deemed practical). That would certainly get the attention of the crew.

Denti:
After the Turkish accident an additional "Airspeed Low" aural alert has been added on the 737NG. It activates at the same moment the airspeed low alert in the MCP speed window becomes active if the airspeed is in the lower amber caution range.

It is a customer option only.

[vilas]

Atleast on the Airbus there are enough warnings both audio and visual also the alfa floor and aircraft stability at alpha prot. The problem is not with the aircraft but the way it is flown by some pilots. It is in the interest of the pilot community that they fly properly so that the aircraft manufacturers are not forced to invent more safety features that will make the pilot redundant.

[Piltdown Man]

Autotrim preserves the crew's pitch authority in both directions for as long as possible...

If that was the case, I wouldn't have made the point I did. A causal factor in G-THOF's stall was, as stated buy the AIB: "The trimmed position of the stabiliser, combined with the selection of maximum thrust, overwhelmed the available elevator authority." Now tell me how you can justify designing and manufacturing an autotrim system that leaves an aircraft in an un-flyable condition? Furthermore, is it acceptable that Boeing knows this a problem and offers advice for stall recovery that includes:

Apply as much as full nose-down elevator
Apply appropriate nose down stabilizer trim
Reduce thrust

I believe it would be simpler to let the A/P fail before a stall, leaving the aircraft in a recoverable configuration (trim vs full thrust vs elevator authority)? To say "the crew should have..." is not acceptable and such a comment is not worthy of a professional pilot. My suggestion was also slightly misinterpreted. I think the maximum "out of trim" force limit for the A/P should be reduced, especially when the aircraft is flying below bugged speed. This will make the stall recovery a simpler affair and allow the use of maximum thrust as part of the procedure.

Call me "old fashioned" but we must not build traps in aircraft and if we find they have sneaked in, they must be removed.

[Winnerhofer]

Control Authority

As you slow down, you need more control deflection to achieve the same control authority.
The only way to preserve authority in positive AND negative pitch at low speed is to bring in the stabilizer, the elevators are not large enough by themselves.
If you let autotrim cut out early you will run out of nose up authority...this is great for stall recovery but will kill you in other situations.
You can't point design a control system for one specific crew screw-up scenario when doing so will endanger the aircraft in other, more probable, situations.

[Daysleeper]

I think what has been forgotten in this thread is that there was a technical failure of the aircraft (the autothrottle). The failure was subtle, was masked by the approach profile and the aircraft's alerting of that failure was inadequate to get the attention of the crew. I would argue that the human machine interface of the alerting system was antiquated and the design, while certified, was not made in consideration of how the system was actually being used in normal operations. That the crew were unable to capture and then mitigate these upstream failures was unfortunate but hardly worthy of some of the comments appended here.

Regarding the original question, I'd suggest that an A320 with a subtle and poorly alerted technical failure would be in just as bad a place. Perpignan in 2008 showed just that and the eventual speed and pitch figures there are similar to THOF's.

Remove the failure and the incident wouldn't have happened.

[A4]

I believe it would be simpler to let the A/P fail before a stall, leaving the aircraft in a recoverable configuration (trim vs full thrust vs elevator authority)

...which is what the 'Bus does in Normal Law. AUTOTRIM is Frozen, AP disconnects and aircraft stabilises at V AlphaProt. You're further "protected" by the FBW masking the pitch/thrust couple. Of course in Normal Law the aircraft should not stall - AlphaMax being the protected limit.

The Perpignan crash was a whole different scenario....

[alonso1986]

Airbus A320 Fly By Wire Demo - YouTube

Just to complement what has been explained here about the stall protection of the Airbus, here's a little demo of the Fly by wire.

Regards