A/T autodisconnection during manual flare

[Britjet]

Sorry, I didn't get to check the override function. I will try again next week.
I did try a manual landing with the autothrottle engaged (very difficult due to pitch-power couple, of course).
I am pretty sure that the autothrottle did disengage in the flare - I will try it again a little more scientifically next week..

Peter

[emerydc8]

Thanks, Peter. I think it will disengage, but I have never confirmed it. I'm curious to see what the A/T FMA looks like when it disengages just prior to touchdown. Cheers. Jon.

[emerydc8]

Regarding my question about whether the A/T would stay engaged on a manual landing with no reverse, this answer from a HKG-based 744/748 captain confirms what you are saying and adds some detail as well. I thought they would disengage at least by touchdown, but I would have never bet that they would go to IDLE at any point on the landing. Also, I guess we can add CX aircraft to the list of aircraft that display ROLLOUT only after an autoland touchdown.
Jon D.

Short answer: No they won't.

Manual landing, with AT engaged.
At 25'RA, the AT goes from SPD to IDLE.
<5'RA TOGA is inhibited.
At touchdown IDLE goes blank on the FMA and AT is disengaged.

So....AT cannot and will not try to maintain your SPD; its is now "off".

To abort the landing you must:
Push the throttles up as you have neither TOGA or any AT at all !!

Above 5'RA, the AT and TOGA are re-armed (the FMA is still blank)
When you hit TOGA (above 5'RA) you get THR on the FMA again.

Why is it designed like this?
My "guess" is to stop you hitting TOGA accidentally as you touchdown or reach for the reversers.

****
Note: on an Autoland the FMA says SPD/LOC/GS
In the flare it says IDLE/LOC/FLARE
After touchdown is changes to
-----/ROLLOUT/----

ie: there is no THR, SPD or IDLE.
The AT is "off".
The AT behaves exactly the same way for a Manual Landing with AT engaged as it does in an Autoloand.

http://www.pprune.org/tech-log/577059-b744-autothrottle-manual-landing-2.html#post9337015

[tango4]

Looks more and more it is behaving exactly like a B777...well, apart for the wake-up part (well, after all the 747 is an old lady, you would not dare wake her up if she is asleep...).
Anyway, as usual, thanks guys for this very interesting discussion !

[emerydc8]

I'm hoping to get more on this in writing, but he mentioned that the A/T behaves exactly the same whether on autoland or manual. It's hard to tell from the video I have of the autoland, but I'm pretty sure it is still indicating IDLE after the airplane touches down and the reverse makes the FMA blank the A/T mode. In any case, it sounds like it will at least be in IDLE at touchdown for either autoland or manual and will not attempt to maintain the MCP speed if no reverse is selected. I guess that would be the same whether in autoland or manual.

[John H Watson]

My 744 engineering notes repeat what has been said:

The flare retard mode, when engaged, causes the A/T servo to drive
the throttles to the AFT mechanical stops at 5 degrees per second.
The AFDS sends a flare retard request during a manual or automatic
landing condition. This occurs at <25 feet radio altitude and the A/T
SPD mode engaged.

I find it odd that the thrust levers go to idle during a manual approach as it's the AFDS which commands the idle. Obviously, the idle would not occur if the FCC CBs were pulled also.

[Hardy Heinlin]

Two details in that quote from PPRUNE are not 100% accurate.
So, as usual, take the whole quote with a pinch of salt.

I say:

1. TOGA is inhibited below 5 ft RA plus 2 seconds.

2. ROLLOUT engages at 5 ft RA already, not at touchdown.


|-|ardy


One questions remains: When the idle job is done, will the A/T mode stay in IDLE mode and, when I have advanced them again during rollout, will the A/T force the levers to return to idle? Or will the IDLE mode disengage once the idle job is done (like it does in flight)?

[emerydc8]

Yes, I saw that too. Regardless, I still think it makes sense that the A/T would act the same whether in auto or manual landing. I may get some snapshots of that section he referenced when he gets back to base in a few days.

This is coming together like a puzzle. First the study guide I have that said the A/T goes off at 5', then Peter verifying that the A/T goes off by itself on landing, now this guy saying that the A/T works the same whether in auto or manual flight. 

[EDIT]:

I say:

1. TOGA is inhibited below 5 ft RA plus 2 seconds.

2. ROLLOUT engages at 5 ft RA already, not at touchdown.

|-|ardy

I agree with you, Hardy.

[emerydc8]

One questions remains: When the idle job is done, will the A/T mode stay in IDLE mode and, when I have advanced them again during rollout, will the A/T force the levers to return to idle? Or will the IDLE mode disengage once the idle job is done (like it does in flight)?

What would it do on an autoland?

[Hardy Heinlin]

I'm referring to both autoland and manual landing. I now think there is no difference re IDLE logic.

[emerydc8]

That's a good question. Maybe we can have Peter do some further investigation on that one. Doing a real-world manual landing with A/T engaged to touchdown, without using reverse, is rare enough; but doing a rejected landing out of that would be really rare. I imagine it could happen with a check airman doing touch-and-gos (if anyone does that anymore in this airplane).

[John H Watson]

I'm having trouble getting my head around this. In manual mode, what is happening to the A/T at 5' (plus whatever). Is it disconnecting, remaining in idle, going into hold or is there just a TOGA inhibit? During Autoland, the A/T has to be manually disconnected by the reverse thrust levers.

[emerydc8]

During Autoland, the A/T has to be manually disconnected by the reverse thrust levers.

I think it will do the same on a manual landing. At least it's in IDLE -- not SPD -- so the throttles won't advance to maintain MCP speed as the brakes are applied. How it would behave if the throttles were manually advanced to do a touch-and-go as Hardy asked, I'm not sure. Would it go to HOLD? Would it fight it and return the throttles back to idle?

[Hardy Heinlin]

Is there any chance you can check this on your next sim check ride, Jon? If your check captain doesn't know the answer either, perhaps he's interested too ...

[emerydc8]

I have been out due to loss of my medical certificate, but when I go back I'll see if I can. I have found that the more questions you have for your check airman the less he's going to ask you.

[John H Watson]

I think it will do the same on a manual landing. At least it's in IDLE -- not SPD -- so the throttles won't advance to maintain MCP speed as the brakes are applied. How it would behave if the throttles were manually advanced to do a touch-and-go as Hardy asked, I'm not sure. Would it go to HOLD? Would it fight it and return the throttles back to idle?

If the thrust levers have been commanded to idle, then surely this will apply...

1) Manual override occurs when the throttle rate does not track the commanded servo
rate. This may be caused by the pilot manually opposing the A/T lever motion or the
levers reaching their physical position(limit). Whatever the cause, the control A/T servo will
detect it and cause the throttles to become dormant."

This leaves the pilot to move the thrust levers at will.

[John H Watson]

It was previously noted that the 737NG idles the thrust levers irrespective of A/P engagement , but there is some very specific logic for the 737NG:

The idle command occurs at 27' RA instead of the normal 24'. The idle command is known as "retard" on the NG.
Flaps must be more than 12.5 degrees
A/T must be in SPD mode
Aircraft must be deemed to be on the Glideslope.

The RA receiver/transmitters (R/Ts) send radio altitude to the A/T. The A/T uses this data to determine control law gains during approach and as a backup for flare retard. The 737NG appears to get its normal flare retard signal from the AFDS system (similar to the 744).

I wonder how much of this has been incorporated into the 744 logic?

Rgds
JHW

P.S. The 737NG Rad Alts feed RA data to their onside FCCs. The FCC's then send the signals to the MCP and then to the A/T computer. The A/T on the NG is not in the FMC. It's a separate computer.

[Will]

A friend who flew the 744, but who now flies the 777, tells me that there's a little more to the FLCH/HOLD story. As we all know, in the 777, when you counteract the throttle servos while in a FLCH climb, the autothrottle reverts to HOLD. But apparently if you move the throttles around, and then move them back to set the annunciated CLB thrust, the system goes out of HOLD and back into THR. The philosophy is that you can vary the rate of climb when needed, but then easily re-engage the servos if your need to vary the rate ends before you reach your target altitude.

Should we expect to see this on the 744? My friend can't remember hearing it talked about for the 744, but he has seen it in print for the 777.

[emerydc8]

So, when you move the throttles back to where they originally were, it goes back to THR. That's interesting. If you can get anything in writing on this from your firend, I would be interesting in seeing it. Thanks for the input, Will.

Jon D.

[Will]

I did receive it in writing from the operator, but I'm not at liberty to post it verbatim. I can paraphrase it though, which I did above. The meaning is quite clear: if you move the throttles back to the annunciated CLB setting, THR mode will re-engage, and the servos will once again control the auto-throttles.  This is for the 777, of course.