actually, the argument, at least as far as I am concerned, is that a TL that is moving with AT gives you two things:

First, you have tactile feedback of the actual thrust setting when you keep your hand on the throttle during the flare.

That is a very immediate, direct means of the aircraft to communicate to you. It does not put any additional load on your visual communication load (which should be monitoring the primary flight instruments and / or the approach minima). It's also a lot less confusing than the aural channel that might be able to differentiate the engine sounds from the rest of the cockpit noises including RA calls, radio communications (not your own, hopefully) and what not.
So without this extra channel, you're quite simple one system parameter short on approach.

Second, upon AT disconnect, the manual thrust setting matches the last AT thrust setting. This, I think, is key. If indeed the #2 TL was left in the CLB detent or shortly below, the AT disconnecting after selecting reverse on #1 would have the #2 engine spool up to the thrust setting defined by the #2 TL.

I've said earlier that good system design should try to prevent errors. That's what this second factor up there does for you. Whether you're in manual or AT mode, moving levers always match the actual thrust setting.

and never fail to do so and comply with the rules that the system requires.

I've said it before, I'll say it again. I'm horrified by the number of people who seem to take a "won't ever happen to me" position. I sincerely hope that all those people are indeed not in a position to ever have that happen to them, that is, the furthest forward seat they'll ever take in an aircraft is 1A or so.

It has happened, at least twice, and the system has neither prevented it from happening nor handled the error nor failed gracefully. Indeed, in this event here, it has failed horribly.

pj

I think somewhere before in this thread somebody asked where it says that a full stop is mandatory after the reversers are used and a go around is no option any more.

It's simply the manual:
A320 FCOM No. 3, Chapter 3.03.22, SOP "Landing"

They did that part right by not attempting a go around. It didn't help the outcome however. I still don't think they would have waited 11 seconds to apply manual brakes. Where did that information come from? It took 20 seconds from touch down to going off the runway. If they didn't brake they might have considered using all that time with that speed to go around, no matter what the manual says.

I may have to double-check, but I *think* this is what the AB system does also until the lever is moved for whatever reason.

Remember that the direct cable connection is considered a thing of the past, not just by AB, new aircraft designs are predominantly FADEC throttle-controlled. Even with force-feedback supplied by software, such a system could transmit the *wrong* tactile information to the pilot.

Maybe because they were not expecting a high deceleration rate (runway advised as "wet and slippery"), so it took a while to register that they were not slowing down at all, by which time an accident was inevitable?

Thanks TripleBravo, that answers my question. The corollary must surely be then that TR(s) must not be selected until spoilers are confirmed deployed. (Maybe brakes too.) If selecting reverse removes the option of a go-around, then you need to be sure you will be able to stop first.

That should be an easy one, if you have the DFDR data (and the coordinates of the runway). LON / LAT is recorded, the air / ground state as well as radio alt, so that you can determine the position where the aircraft settled within GPS precision (limited by the time step interval of the recording and the precision of the aircraft's transducers).

See http://www.risingup.com/fars/info/part121-344-FAR.shtml for required parameters, LAT / LON is #39.

Hmmm... my impression from what has been said so far was that AT disconnect would set the thrust as commanded by the TL setting.

Yes, point taken, there's a failure mode to moving levers as well, and I agree, wrong information is even worse than no information. However, that's not relevant in the current situation. We're not discussing a system that reports wrong information, we're discussing a system that reports no information at all, at least not through this sensory channel.

pj

no2 Engine Thrust
 

PJ The AT disconnected when the No1 engine went into reverse. The No2 engine would now maintain it current power setting. This was mentioned earlier with a power setting of 1.02.

"That involuntary ATHR disconnection allowed the thrust to be frozen on engine2 whose lever was at CLB notch." -Airbus.

http://www.asc.gov.tw/acd_files/188-c1contupload.pdf

The table on page 142 is interesting.

The engine would not spool up it would just remain as it was.

Jim

I think we must be careful not to overcomplicate things :ooh:

We all know a "go around" is possible until at / prior a normal touchdown / approach. The performance has been checked, the procedure briefed for every approach etc.

A "touch and go" is very different. If initiated during / after a long float, the performance calcs are already compromised. Once you have touched down, and say Gnd Spoilers deployed, if you now decide to GA, you are "gambling" this is better than "sticking with it". Either solution might be correct, or incorrect, but you are on your own performance wise, and therefore a "pre-planned touch and go" because, say the spoilers seem not to have deployed is an illogical step, until and when all Landing Performance calcs are redefined for this procedure.

IMHO, the time to call a late GA is prior touchdown (e.g. a long float). So long as you push the TLs forward, then even if you touchdown so what? Leaving to "go / no go" decision until after touchdown is rather too late to discuss things in a 2+ man crew. Technically, until Rev is selected, yes, there is a GA option... but with 2 people able to call "GA", and 1 selecting Rev ASAP after touchdown, it is a minefield.

We cannot cover every eventuality, and a performance limiting landing is effectively relying on "everything working". If the brakes / spoliers / reversers (wet) fail then tough - life will get difficult... and the crew will have to take the best decisions according to the circumstances...

NoD

You're welcome, SysEng (didn't find your post in this thread anymore).

The Airbus SOP (which may differ across different airlines) seem to be not really consistent here. On the one hand it lets the PNF call out first "ground spoilers", then "reverse green" (if appropriate). On the other hand the FCOM says "Select MAX REV immediately after the main landing gear touches down." (SOP landing)

But I agree, it seems to be important to first be sure that spoilers are deployed before you pull the TL back. The short wait before pulling will eat up a few meters of runway length though, which were of importance here ...:(

With the TLs at "CLB", and the ATHR suddenly being deselected / deactivated / switched off, 2 possible results:
1. Thrust on that Engine => Climb Power. Ask any Airbus pilot who just manually disconnects the ATHR and forgets to retard the TLs first :{
2. Thrust remains current setting. More inclined to where ATHR "fails"... and you get the "Thrust Lock" amber warning etc.

It seems in this case the latter took place, leaving #2 in some fwd thrust, but not high power. My reading of the Airbus info was 1.20 (not 1.02) which is higher than approach power, but nowhere near a takeoff / climb setting.

NoD

TB - disagree - sorry :eek:

You need to select Max Rev straight away for best performance. From selection of Rev to getting "Rev Green" (via Rev Amber) takes some 3s? 5s? Whereas the Gnd Spoilers indicate pretty much straight away (and of course do not require human reaction time). So the call in the SOP order fits in with the order they shoukd be achieved...

I also 100% disagree with your statement That is directly contrary to SOPs, probably on all aircraft types... and if you did it on a Sim / Route Check, I trust you would be failed :{

NoD

NoD/Triplebravo,
Please forgive/correct me if I am wrong but I thought that the certification basis for all types takes no account of the contribution of reverse thrust in the calculation of landing roll distance required. "Best performance" will certainly be achieved with full reverse, especially in the wet, but its main function is to reduce brake and tyre maintenance costs. If this is correct then the selection of reverse before spoilers confirmed out puts DOCs above safety.

NoD Sorry, I should have denoted my posting as "thinking loud", did not intend to state any SOP with that.

Dry - yes. Wet - no - IMHO ;) i.e. in the Wet, Reversers are assumed to some extent, and why "1 Rev U/S" MEL requires no Perf corrections in the Dry, but does in the Wet...

But you miss my main point ;) To await selecting Reverse until you see the spoilers out, and presumably "touch and go" if you don't see them (?), is a complete new SOP / Performance issue. All training I have done to date is basically once you touchdown do all you can do to stop ASAP ;) That means selecting all stop aids ASAP - not a flow diagram of A, then B, and if you don't get A then try C ;)

You are correct. We've mentioned that around pages 48,49 (many of us) that by CFR 14 Part 25, part 121 NO TR IS ACCOUNTED FOR CERTIFIED LANDING.

NoD,

NO, for wet there is no account of TRs in any certification documentation. OEM recommendations, QRHs etc withstanding.

Only Part 25/121 are official.

Cheers

anyone see the recent AIT from Airbus? -- no acft malfunction.
also, anyone know when/which MMEL revision changed operational procedure on thrust levers with inop reverser? I know rev 29 and 30 say both levers on rev max, including inop side.

Safety Management
 

A wider view of this accident (based on information in this thread) might conclude that the operation had insufficient safety margin – ‘too close to the edge’.
The ‘short’ runway without an overrun area lacks the safety margin available at other airports. An operator could consider a landing weight limit, less than the certificated performance. There is no regulatory need to for this - so it is not done.

An operation with reverse MEL’ed also reduces the safety margin, and it adds to the crew’s workload with the need to control asymmetric thrust. Thus a rev inop configuration could be prohibited, or an additional wet / crosswind limit applied, again with commercial implications.

Do the crew / operators consider the variable (reduced) safety margins in the range of wet runway conditions, or how close a wet operation might be to a contaminated condition with a heavy shower or ‘damming’ wind’. What effect would / should this have on the policy to tanker fuel?
Do the crew consider (plan / train for) a spoiler failure? In performance terms the increased landing distance may well be within the certificated safety margin (do we know our own type?). The continued landing requires immediate recognition of the problem and a change to the planned operation – maximum braking. It also assumes that the landing was made at a reasonable speed and touchdown position, as both of these reduce the existing safety margin that will be relied on for the spoiler failure. Also consider the importance of the ‘spoiler’ call, how often is this made by habit when the spoilers have not deployed. IIRC at least two accidents involved aspects of this (146 at Aberdeen and MD80 Little Rock); safety recommendations were that only the failure of a system should be called, removing the superfluous ‘by rote’ call.

With a temporary ungrooved runway other restrictions could be applied. Would either the airport authority or an operator consider the reduced braking effectiveness as a need to reconsider performance, and additionally any effect due to new tarmac ‘sweating’? Perhaps more commercial issues.

Thus there are many seemingly inconsequential issues to consider, but we rarely think about these or the probability of a sufficient number of them happening together to lead to an accident. Unfortunately history shows that it only takes one of these issues in conjunction with a human error (perhaps originating from other avoidable issues – documentation, SOPs, training) to cause an accident.
The premise of a Safety Management System is that operators and individuals think about these issues, their combinations and the hazards. A judgement is required which increasingly relies on specialist knowledge, and correlation with previous events, incidents, and accidents elsewhere or involving other types – a problem of communication. The managers in this decision process must provide a balanced judgement of safety against commercial interests (safety is a commercial issue – the first issue), they must avoid bias, and heed the lessons from the apparently increasing number of accidents where this balance appears to be incorrect.

This might sound like CRM for management within a SMS – an issue of oversight by the regulatory authority?
An intriguing thought is, that this is exactly what thread is undertaking - a SMS process, with hindsight, and with all of the associated CRM strengths and weaknesses, to seek ‘the’ safety judgement for this operation, but perhaps without commercial bias?

If you confirm spoilers before selecting reverse, you preserve the option of a go-around if the spoilers do not deploy.

On the other hand, if you select reverse immediately on touch down on a limiting runway (especially without EMAS), you have just made a big bet that the spoilers will work as advertised and that none of the interlocks will get in the way.