PDA

View Full Version : 737 Single-channel autoland


BOAC
11th Feb 2005, 13:13
I would like to pick up something that arose in this thread (http://www.pprune.org/forums/showthread.php?s=&threadid=158753&perpage=15&pagenumber=1) a while back and repost two posts:-

From ' Hudson'

I don't know about the real aircraft but the 737 Classic simulator will execute a perfectly respectable smooth coupled landing just on one autopilot. Under this one autopilot operation there is definately no autotrim back at 400RA though, nor at any other stage. It is comforting to have that fall back availibility on one autopilot if you do not have a second autopilot available for some reason and you just have to get in in bad weather on the ILS.


From 'Seat1APlease'

Hudson I am glad you mentioned that because the same thought struck me. We were told that Mr Boeing had to allow for the case where a crew for whatever reason did an approach with only one a/p engaged. (200/300/400 series)

It obviously has to stay in until cat one decision height but what then? either drop out or continue and land or fly stright into the ground, so the default was it would continue for A/land but it wasn't approved and therefore not published nor documented.

As you say there is no guarantee that the sim and the real beast will behave the same, and we couldn't of course try it on the real thing, so it became one of those urban myths which no-one is quite sure about.

Perhaps someone can clear this up?

I now have it on 'good authority' (a 737 sim trainer!!) that the 737 (sim) will indeed perform a 'reduced flare autoland' single channel.

Can anyone throw any light on this ie:-

1) Is it a 'deliberate' function?

2) What 'cues'/inputs does it use for the landing manouevre?

3) Will it work on the real a/c or is it a sim software 'wildy'?

Joyce Tick
11th Feb 2005, 13:54
If you think about it for a minute - of course Boeing designed it to land on just one A/P - it's called 'fail-safe' or 'fail positive'. It allows for the case where one A/P fails just before the flare - what do you think would happen if the flare mode failed in toto, just because one A/P fails?

You will get no Flare Armed indications with just one channel selected, or if one fails, but it will still flare, just fine!

LEM
11th Feb 2005, 13:58
Hi BOAC, what do you mean with reduced flare autoland?

Mr Levitator
11th Feb 2005, 15:24
My understanding was that each AP must be able to conduct a full auto land individually. How else could each AP monitor and know that the other was doing the correct job. I think that the Master (FCC) actually does the job and the other slave just watches. If there is a disagree, a message is sent (somehwere) for the AP's to disengauge.

What I would be interested in is what you have to do to arm this function with just a single AP engauged(if at all)? Does it happen automatically?? I have never seen the (dreaded) uptrim come in when on a Single AP CAT I landing. Or as I supspect, the Auto land fuction is only armed when both AP's Engauge.

Of course I stand to be corrected!!

Mr L.

BOAC
11th Feb 2005, 15:34
LemHi BOAC, what do you mean with reduced flare autoland? - no idea! Just the words used to me.

Hobo
11th Feb 2005, 18:05
Not only will it land with one channel it will A/Land on one engine with one channel.

In extremis taking off in low vis with an uncontrolled engine or wheel well fire what would you rather do, fly on fire to your cat I return alt - maybe 1 hour away. Or attempt a non standard A/land?

BOAC
11th Feb 2005, 20:15
Not only will it land with one channel it will A/Land on one engine with one channel. - hobo - if I KNEW it would do it, probably the latter, but you offer NOTHING to back up your 'statement! It is not often that I feel like a test pilot.:eek:

ZSLHF
12th Feb 2005, 04:39
Just a couple of minor corrections on the above postings. In two channel mode, the system is fail-passive, which means the FCC giving the lower control input/change is used (neither is master/slave - that's why both M/A lights are illuminated on the MCP). In single channel mode, she will fly you accurately on the ILS to the ground with no flare (not reduced flare), but because the 737 classic sim does not simulate hard landing properly, it feels like a reduced flare (ask me - did it yesterday). But then again this single channel capability of the 737 is nothing special, because I've flown down to 30 feet an ILS on autopilot (as a test in fine weather) in a Mooney 231, BE58, C421, PC12 and King Air 200. They can all fly you down to 30ft AGL without much difficulty.

BOAC
12th Feb 2005, 08:07
Thankyou, ZS - that is what I suspected. As you say, any autopilot will fly down to ZERO in fact (unless there is a built-in Rad-Alt disconnect) - it is how it does it that can hurt!

Hudson
12th Feb 2005, 11:58
ZSLHF. My initial info about the 737 Classic simulator landing on one AP came from a very senior Boeing instructor from the Boeing company. I assumed that it would apply to the real aircraft too, otherwise no point in bringing the subject up. On one autopilot (sim, again) the flare and throttle reduction is exactly the same as a two AP operation. In the sim I work with you cannot pick the difference between two or one AP autoland.

My initial post stated that there was no backtrim applied at any time. I was wrong. It does back trim just at the flare -whereas a two AP approach trims back at 400 ft.

If you think your particular Classic sim does not replicate a hard landing, you want to try the sim I work on. It fair rattles your teeth with the impact and as I have done more than my fair share of let us say firm landings in the real thing, I can tell you that our sim at least is very much the same as the real thing. When I see a late flare in the offing, I brace myself in the sim for a hard touchdown.

On the other hand I have flown 737 sims where the techs have deliberately adjusted the machine to do only soft greaser landings regardless of high sink rate at impact. Very satisfying of course but useless. I believe that is to minimise jarring of the machine and thus maintenance costs. Whoever does the fidelity checks on those sims must look the other way because no way is that a accurate replication of a hard landing.

ZSLHF
12th Feb 2005, 20:31
Hudson, I'll double check with a senior instructor later this week, maybe he can give me more info. Watch this space...

junior_man
12th Feb 2005, 21:08
737 will not autoland on one auto pilot. It will fly the GS into the ground. No flare at all. May look okay in the sim. not something to try in the aircraft.
Need to be dual channel to get "Flare armed" will not occur on one channel. No flare armed, no flare.
Aircraft is fail passive for auto approach. Ie it will not kill you if something goes wrong. It isn't fail active, where a failure will still leave the capability to autoland.
If one autopilot fails Both will disconnect. The reason for the nose up bias trim at approx 400 feet RA is so that when the disconnect, the aircraft will pitch up away from the ground. Will get your attention. Add power for a go around (or from auto throttle to maintain speed) and the aircraft really pitches up.
737 is also not approved for any single engine autoland. The auto thrust is not approved for single engine ops either. 737 autopilot doesn't control the rudder like 757 767 or A 320. So with an engine out it would yaw all over the place.
This pertains to -200 -300 and -400. Applies to SP 177 auto pilot. Early -200s with SP 77 auto pilot have no autoland capability.
I have never flown the next Gen 737. So can't speak about the capabilities of it.

alexban
13th Feb 2005, 06:29
It is written very clear in the FCOM: on single channel app the flare feature IS NOT AVAILABLE. It's real,no joke.
It will smash into the rwy with less than 600'/min,but no flare .Maybe some ground efect,but not a nice landing.
I think that if this would've been possible,Boeing would've said something like....autoland on SE is not permitted,recommended,authorized...not unavailable
The sim it's ALMOST like the plane,it's not programmed to do all the things like the real plane,if they are not to be expected.
I.e. try a loop ,or a barrel,it will be easy to do.Just like a FS on pc. I wouldn't try it on a real 737 though,suspect it to be harder,not like the sim...:)
Brgds...
Alex

LEM
13th Feb 2005, 07:49
I've never tried autolanding on one AP, so just a guess: my bet is Boeing, in it's typical legal telegraphic wording, could not make any reference to such a sensitive issue like autoland on only one AP.
But actually the real thing is just capable of doing it, although not certified thus not documented.
This is true on other airplanes also, like the Dornier 328, which will "autoland" if you leave the AP engaged till touchdown.
Of course, this was observed in the sim, so we can't be sure 100%.
Only way to be sure is trying on the airplane, with no pax, of course; if you see no flare by 30 ft, still time to disengage and pull.
Any volunteer?
;)

74Freight
13th Feb 2005, 08:01
never had the pleasure to fly one like you chaps but had a little play on the sim, question- if the flare is not armed on a single channel approach, and does not annunciate on the FMA, does it still do it or just fly firmly into the ground ?

LEM
13th Feb 2005, 09:07
74Freight, you haven't done your homework!
Just read carefully this thread, and the one related to it...

;)

safetypee
13th Feb 2005, 09:40
“737 will not autoland on one auto pilot. I It will fly the GS into the ground. No flare at all.”

I cannot vouch for the capability for the 737 to autoland on one auto pilot, but I would be very surprised if were certificated to do so.

However, flying the GS into the ground may be a simulator problem and / or a source of confusion. In the real world the ILS GS beam is formed by a reflection off the ground in front of the transmitter, the ground position being the GS origin. The geometry of this at very low altitude bends the GS beam into a ‘flare’ shape as the azimuth angle between aircraft and transmitter increases. The point of reflection is between the aircraft and transmitter; in some cases the GS beam may not touch the ground / runway. This is one of the reasons that autopilots fade out the GS signal at low altitude.

One aspect of this thread and the previous ones, is the lack of conclusive knowledge; thus it was interesting to read Flight Crew Reliance on Automation (http://www.caa.co.uk/application.aspx?categoryid=33&pagetype=65&applicationid=11&mode=detail&id=1598)
The primary conclusions were that there is a lack the right type of knowledge (training) for automation, and that there are deficiencies in CRM training re thinking skills. This of course involves how knowledge is used and in particularly about making assumptions

LEM
13th Feb 2005, 15:59
I cannot vouch for the capability for the 737 to autoland on one auto pilot, but I would be very surprised if were certificated to do so.
You don't need to be surprised. The fact it's not certificated to do so, everybody knows.

Deficiencies in how knowledge is used and in particularly about making assumptions?
Making assumptions is the only thing we can do over these grounds, until somebody actually tries in the airplane.

Yes, here we are pushing our mind beyond the strict limits of functioning like robots, the anal behaviour you must certainly love.

:yuk: :yuk: :yuk: :ugh:

sets
14th Feb 2005, 22:29
Can only speak re the 737NG. Will autoland on single channel on the real airplane. Main differences from dual channel are :

Single channel autopilot authority only
Much reduced failure protections (e.g. no hardover protection)
No flare spring bias stab trim at 400 feet
Will not annuciate FLARE
No guarantees made as to performance (sink rates, position etc.).


Why is it available ? Its there because its lesser of two evils (i.e. if the pilot has let it go that far it might as well try and save his ass rather than disconnect).

DO NOT try this on a real airplane !!!! Its a safety net - not a certified function of the system. The AFM limits use of the single channel autopilot to 50 feet for good reason.

LEM
15th Feb 2005, 07:25
Hi sets.
You say "Will autoland on single channel on the real airplane."

One of the doubts raised in this thread is that, although some of us have observed this capability in the sim, we wonder if simulators always reflect the behaviour of the real thing (soft landings might be programmed in the sim for hardware protection...).

It's good to hear from somebody who says firmly that this will happen in the real airplane.

Just tell us what makes you so sure, please.

Have anybody of your acquaintance tried in the real machine?

LEM
23rd Feb 2005, 09:52
Fantastic, ASFKAP!

At last something not questionable!