Following this thread with all its diversions from the erudite to the banal and especially to the hopelessly biased has been quite an education. However, if there is one positive that can be extracted from any accident it is precisely that: that the analysis and discussion will make us all better educated and more wary of the conditions that brought our colleagues to grief.
Unfortunately, one of the predominant themes of this thread on the part of some has been the attempt to cast immediate judgment and blame on the non-moving aspect of the Airbus thrust levers without even the slightest evidence that they played a causal role in the accident. If the latest information here quoting the Airbus AIT is accurate, then though the apparent failure of the PF to retard the TL2 to idle is the initiating factor for the accident, the fact of whether the TL's move a la Boeing or remain fixed as per Airbus is irrelevant. Retard only one of them in either case and on the other TL you will end up with a fixed thrust equivalent to the power setting of the engine at the time the first TL was retarded. It seems logical and it is logical, and it is most definitely not a case of the *computers* taking over and over-riding the pilot's intentions. In fact it's totally the opposite, with the *computers* doing all that's possible to match the thrust to the pilot's inputs (TL1 to idle, idle you get on #1, TL2 not moved, ok no change on #2 until you command it).
This logic also makes sense in terms of the ground spoiler actuation and autobrake application. The thrust levers are the pilot's primary tool to communicate his intentions to these systems both on a Boeing and on an Airbus. When a pilot leaves one TL at approach thrust and retards the other to idle the intention becomes entirely ambiguous ... does he want to land or is he preparing to go around? The *computers* cannot read minds and any arbitration on their part absent a clear signal of intention has a significant potential to be catastrophic. So, the resulting logic has to be: where the intention of the pilot cannot be determined because he has provided directly contradictory inputs, the system will remain in its current condition (spoilers and autobrakes armed) until other signals from the pilot (via TL's, Rev, manual braking or manual speedbrake) clearly indicate his intentions.
And really, this is the way that we pilots want it to be. Almost all of the high vitriol here is about how it must be the pilot who controls the aircraft and not the computer. In this event it seems the aircraft gave the pilot everything he commanded through his actions (REV on #1, approach power on #2, later manual brakes when applied), the problem was that those initial actions were not sufficiently unambiguous for the system to deduce that a landing was intended and that automatic ground spoiler and braking should occur, as opposed to requiring further manual inputs from the pilot to make that happen.
On the apparently confusing subject of whether the ground spoilers would have deployed had the speedbrake lever been extend, I believe the explanation is as follows: Per the FCOMs for the aircraft I have flown
Ground Spoilers are only ever automatically actuated through the arming conditions and thrust lever positions as documented in earlier posts.
Speedbrakes (or Flight Spoilers if you prefer) are another matter. These, which are physically the same control surfaces (3 of 5 anyway), can be actuated manually through the extension of the Speedbrake lever. On the A320's I flew this function was inhibited in CONF FULL (but that was 10 years ago). However, more recently flying A330s and A340s for several operators I have discovered that different aircraft of the same model have different inhibitions in this regard. It depends on the particular modification status of the aircraft which in this case is dictated by the flight control computer standard (for the A320, the SEC) installed. Quite likely all A320's are not the same in this respect either and hence the differing opinions from pilots flying models with different SEC standards installed.
Regardless of this though, the potential or not to manually extend the ground spoilers or speedbrakes is only relevant if an attempt to manually extend them actually occurred and failed to do so. So far there is no indication that this is the case. And, while on the subject of the ground spoilers, their effect on stopping distance is significantly less (though certainly still valuable) than it has been made out to be by some posters. In the TransAsia accident report regarding the over-run at Taipei Sungshan airport on page 91 it is noted that:
The Landing Performance in Sec. 1-27 “Flight Controls” of TransAsia’s A320 Minimum Equipment is states that if the ground spoilers are not extended, the landing distance of his type of aircraft (A320-232 B22310) shall increase by 15%.
This can be confirmed by reference to MEL 27-92-02 which describes means for calculating takeoff and landing distances with the Ground Spoiler Control System inoperative. If we accept PJ2's calculations from post #120 of braking distance with Max manual braking and with Autoland and medium autobrake (ALD between 4100'-5400') as being reasonably accurate then the lack of ground spoilers would only have added 600'-800' to the actual landing distance. A noticable amount to be sure, but not at all the critical factor in this instance. As an aside, I'd hope that this would put paid to the idea of holding off on reverse or executing a go-around if the ground spoilers fail to deploy. For them to become
the critical factor in stopping safely other failures or contaminant performance degradation would also have to occur at the same time.
In terms of the facts so far revealed (at least if we take the AIT information to be correct) there has been remarkably little discussion about the factor that stands out to me as being most directly responsible for the failure to bring the aircraft to a stop on the runway:
The failure to apply full manual braking as soon as a lack of normal deceleration was detected.
If the time of 11 seconds from touchdown to initial manual braking is accurate, this represents 2200'-2500' of the runway from the touchdown point that was traversed without either crew member taking an affirmative action to stop the aircraft. This delay seems to have occurred despite the fact that both crew members were aware in advance of the very limited stopping distance and the runway conditions and became aware within 3-4 seconds of touchdown that the spoilers had not extended. Why an extra 7 seconds passed before an attempt to do something about the lack of normal deceleration was made has to be the real crux of the investigation.
Reviewing the TransAsia accident, the same factor, a 15 second delay from touchdown to the first application of manual braking is also found to be a critical factor in the cause of that over-run. In fact, it is very instructive to note the Airbus performance calculations on page 143 of the report. There are too many variables to make a direct comparison to TAM 3054 (TNA had a lower weight but a higher ground speed), but it is worth noting that Airbus calculated the ground stopping distance on a wet runway with 1 engine in max reverse, 1 maintaining 1.08 EPR and no ground spoiler extension as being 1907m (about 6200') if braking equivalent to autobrake medium was used. With maximum manual braking that figure might be as much as 25% less, which leaves open the possibility that even without the crew ever recognizing that TL2 had not been retard and that the ground spoilers had not extended, the TAM aircraft
might have still stopped on the runway or within the confines of the airfield if maximum manual braking had been applied at the first recognition of a failure in deceleration (18:48:29.5 HOT-2 "Spoilers Nothing").
All of this leads back around to whether there's a procedure in place for this that could have been identified and used in this circumstance. There is, and it is one of the very few required memory items for an A320:
Loss of Braking
If Autobrake is Selected:
Brake .....................Pedals Press
If No Braking Available:
Rev........................Max
Brake Pedals............Release
A/Skid & NW/STRG ...Off
Max Brk Press .........1000 PSI
If Still No Braking:
Parking Brake ..........Use
One might argue about whether the "Rev......Max" is best placed before "If No Braking Available" or after, but the essential point is that this procedure is the manufacturer's intended guidance on what to do at any time when the rate of braking experienced doesn't match the rate expected or the rate necessary to stop on the runway. And the first step is obvious, instinctive and requires no particularly arcane knowledge of flight control systems or spoiler and autobrake activation logic.
So, at the end of the day, though the question "Why did the crew member retard only 1 TL?" is a very important one, the really critical question to be answered, in this as in many other runway over-runs, is "Why did the crew allow so much time to elapse and so much runway to go by before attempting to brake manually?" More particularly, why was this so when they were completely aware of the runway length, runway surface condition and thrust reverser inoperative limitations?
To my eyes there are critical underlying training issues here that are far more relevant than any discussion about preventing this accident in future with things such as moving, vibrating thrust levers or planned go-arounds in the event that ground spoilers fail to extend.
ELAC