disconnect autopilot during cruise

[mgenete]

Hello,

I've see in the 747-400 virgin atlantic itvv dvd that crew disconnect then reconnect the autopilot during cruise to make sure the aircraft is correctly in trim.
Is it just a compagny specific action or is it a boeing 747-400 operation manual specification ?
Why the autopilot could not trim correctly the aircraft ?

Thank's

Mathieu

[Jeroen Hoppenbrouwers]

The A/P can trim the aircraft (i.e. the horizontal stabilizer -- not the ailerons or the rudder). However, if for some reason this does not really happen, the A/P will dutifully fly with elevators out of null. This isn't desirable for several reasons.

By disconnecting the A/P and seeing what the plane does when left uncontrolled, you get a pretty good idea whether the A/P needs to work too hard.

Rudder trim typically isn't null for these planes. A bit left or right is required to neutralise bank tendency and center the sailing boat. With the A/P constantly adjusting, this cannot be done really well.


Jeroen

[Mariano]

Hi Mathieu,

I believe they must be checking for aileron trim. The autopilot trims the horizontal stabilizer, but it has no control of aileron or rudder trim. If in flight, a slight fuel imbalance develops (quite normal) the autopilot will compensate by using ailerons. This, in turn, creates extra drag that you may not want if you will be needing extra range (or more fuel at destination and, who doesn't?) Very small aileron deflections over the course of many hours may cause noticeable extra fuel burn. Sometimes, rudder trim is also checked (the previous crew may have had a need to slightly displace aileron / rudder trims to fit their needs.) The less control surfaces out of neutral during flight, the better for range / endurance.

Mariano

[mgenete]

thank you for your answers.

In fact, it's a specific effect on long range flights ?

Mathieu

[Jamie]

A bit off topic, but the CRJ2 for example has a tendency to rock the wings constantly during cruise IF not perfectly trimmed along the yaw and/or roll axis. Looks and feels a bit like being in a Dutch-Roll.

On the Fokker you can trim the rudder with AP engaged. I've seen a former 'long haul' captain switching of the autopilot @ FL350 then trimming the ailerons and rudder for a few minutes then engaging the AP again.

What I do is fly it manually after take-off, if load permits, and trim the aileron so it flies straight. During cruise or straight climb I adjust the rudder (with AP on). However this is when flying the Fokker 70/100.

Sometimes there is a system failure causing the AP to disconnect spontaniously. Always be aware this can happen.

[Mariano]

Mathieu,

It doesn't have to do specifically with long range flights. You can always save fuel (no matter what distance you are flying) by 'streamlining' the aircraft.

One of the reasons for having a trimmable horizontal stabilizer (as opposed to trimming the elevators) is to reduce drag. When you trim an elevator, it sticks into the airflow and that causes undesirable drag (called 'elevator drag.')

If you have aileron drag (caused by not-neutral ailerons) the effects on fuel burn due to the ailerons not being flush with the trailing edge is going to have to be multiplied by the number of hours this condition continues unchecked. On long range flights, this becomes more critical than on short segments. Keep in mind that when you trim the ailerons, you are re-setting their null point, and they will still stick into the airflow a little bit. When you re-engage the autopilot, it will not have to displaced the ailerons any longer since there will be no unbalanced force to counteract.

Anytime an aircraft is flying with the ball not centered, the relative airflow is not 'hitting' the aircraft from' straight ahead' and unnecessary drag is created. When reaching cruise altitude, the ball is checked and if out of center, the autopilot may be disconnected to correct this condition, therefore getting rid of the associated drag for the rest of the cruise portion of the flight. Also, if during cruise the autopilot is compensating for an unbalanced condition (by using ailerons) when the pilot disconnects the autopilot during the approach, the wings may suddenly rock undesirably.

In a similar manner, the main reason aircraft are loaded with an aft CG is so that the payload is now helping 'save' fuel by creating a tail-down force that otherwise would have to be created by displacing the elevators or moving the horizontal stabilizer (depending on the airplane.) As said earlier, the less any flight control is displaced from neutral, the less drag the engines will have to counteract by burning fuel.

Mariano

[Will]

Jamie, you said: "What I do is fly it manually after take-off, if load permits, and trim the aileron so it flies straight."

How would load factor into whether you can fly manually or not?  Thanks!

Will

[Zinger]

Trimming the aircraft during flight in the roll and yaw channels can be performed in autoflight utilizing EICAS Status. Boeing designed it so but some airline manuals specify aileron trimming in autoflight as prohibited.

[Jamie]

@Will: with load I mean brain-load  :roll: Some airports require special attention e.g. heavy traffic in combination with a lot of ATC instructions (LHR > never flew the SID, always heading this heading that  :mrgreen: ), special departures, we even have an airport which requires switching 3-4-5 times ATC before we even are at 3000ft, in these cases we engage the AP ASAP.

[Lasse]

There are two trim procedures well discriped in the flight crew training manual FCTM.
This one is the one we most often use:
Check symetrical thrust
Balance fuel...
Select HDG SEL...
Wait 30 sec...
Trim rudder until trim indication on control wheel is 0. (Be aware that the sticker might not be 0 where the control wheel has its edge... It might be slightly off to left or right so it might look like the control wheel is slightly turning)

You can perform this procedure with or without autopilot...Normally we keep it on due RVSM requirements.

Use of ailron trim with autopilot enganged is prohibited. Thats a certification limit stated in the FCOM/AFM.

Small out of trim conditions affect fuel flow by less that 1%, if no spoilers are deflected. Since an aircraft twists and bends during flight due temperature, speed or fuel use then trimming is done now and then during a flight...

There are well defined limits how much it can be out of trim for delivery or after major checks, but not when flying actively on the line...
What some operators do is haveing technical pilots test it before the check to inform the maintenance how many units its off for the specially selected configurations so they can adjust flaps according to the AMM...Since the aircraft is not perfectly straight then we use flaps to trim it into limits by adjusting the flap drive...Its very clear to see when flaps are up then aft are not flush with the rest of the trailing edge...

Then you can say the flight crew will do the "fine" tuneing for the present day...

/Lasse

[Zinger]

1. B747-458 current FCOM:
Aileron Trim
Pushing both Aileron Trim switches in the desired direction causes the feel and
trim mechanism to reposition the aileron neutral point. Both control columns have
an aileron trim indicator.
If the Aileron Trim switches are activated with an autopilot engaged, the aileron
neutral point is repositioned. When the autopilot is disengaged, the wheel and
ailerons move to the repositioned aileron neutral point. The airplane responds with
roll proportional to the amount of aileron trim input.
The same FCOM prohibits aileron trim in autoflight as you state and I did above.

2. According to flight mechanics, the roll and yaw channels interact,  therefore it is futile to trim one without the other. I'd therefore add to the aileron trim procedure stated, to trim in yaw first using the trim knob, verifying that the aircraft bank angle is level and no yaw acceleration exists.

[Jeroen Hoppenbrouwers]

My guess is that repositioning the aileron neutral point (i.e., in autoflight) does not at all change anything in flight dynamics. No visible roll, no indication on EICAS control surface readout, no change in drag, no nothing -- because the A/P could not care less where the neutral point is. It stupidly drives the ailerons wherever it wants them.

But at A/P disconnect time, the ailerons will be "released" and snap into neutral position. If you have trimmed the ailerons far enough out of the position they should be in, you induce a severe roll and nobody expects this at A/P disconnect time. Let alone when A/P disconnects unexpectedly.


Jeroen

[Lasse]

Thats the exact reason for not being allowed to use aileron trim with AFDS on. AFDS will just compensate untill you disconnect the AFDS and then you will have a uncommanded roll which is considered unsafe.

True both roll and yaw goes hand in hand, the FCTM Alternate Rudder Trim Technique trims aileron and rudder...

/Lasse

[OKD]

Jeroen

I think your reasoning is almost exactly of what happened to China Airlines (747- Combi - correct me if I am wrong?) sudden spiral dive over the pacific on its way to SFO back in the mid 80's (Air Crash Investigation - National Geographic Channel).

I could not recalled the weather condition during that said flight, but clearly remember that it was at FL410!!  For some reason, the Captain disconnected that the A/P, and that's when all hell broke loose.  Obviously that the A/P was "correcting or flying into its almost limit in aileron condition).  By disconnecting the A/P was just like sudden snap (or let go) of a rubberband with a piece of object on one side.

End finding was that had the Captain gently did some gentle rudder input at its initial dive, he wouldn't had gone into the serious trouble.

But the main thing in end of this story was that the Captain did what he was trained to do, SAVED HUMAN LIVES & AIRCRAFT.  Landed in one piece, but with parts missing "left-right-centre) from the 747!!

My guess is that repositioning the aileron neutral point (i.e., in autoflight) does not at all change anything in flight dynamics. No visible roll, no indication on EICAS control surface readout, no change in drag, no nothing -- because the A/P could not care less where the neutral point is. It stupidly drives the ailerons wherever it wants them.

But at A/P disconnect time, the ailerons will be "released" and snap into neutral position. If you have trimmed the ailerons far enough out of the position they should be in, you induce a severe roll and nobody expects this at A/P disconnect time. Let alone when A/P disconnects unexpectedly.


Jeroen

[Hardy Heinlin]

Pilots,

when half of the pax seats are unoccupied and you fly in regions where an attractive landscape lies on the left side of the aircraft, a significant number of pax will probably move over to the left side.

Has anyone of you experienced such a case and noticed the consequences on the aerodynamics?


Cheers,

|-|ardy

[Jamie]

non-747 driver but yes we notice. E.g. fuel imbalance on Fokker is accepted if its below 350kg (!) which is quite a lot. Needs a few cranks on the aileron trim to keep it straight The Fokker APU drains only the left tank, furthermore we taxi out single engine often (engine #1 also left tank). Quite possible. In F70/100 no indications of controls and roll/ruddertrim is manual.

Flying manual above say FL100 requires significant input force. Not quite 'comfortable' (artificial feel system for control load feedback). Dont know how this is done on the 747.

[Will]

Hardy, I can tell you from personal experience that it makes no difference in a DHC-6.

http://www.grandcanyonairlines.com/gca/show_article.aspx?article_id=71&lang=en-US

Will