Minimum Drag Trim Technique

[NGjockey]

This is what Boeing calls the "Minimum Drag Trim Technique":
　
Autopilot engaged (preferred method):

Stabilize the airplane on a constant heading with HDG SEL
Trim the rudder in the direction of the down wing. Apply rudder trim incrementally, allowing the bank to stabilize after each trim input. Large trim inputs do not give repeatable results.
The airplane is properly trimmed when the bank angle is zero as displayed on the ADI sky pointer. If the airplane is properly rigged, this should result in an approximately neutral wheel.
Disengage the autopilot. With the autopilot disengaged, hold the wings level with the control wheel using the ADI for reference. Trim out any control wheel forces using aileron trim.
Note: Use of aileron trim with the autopilot engagaed is prohibited.
If desired, re-engage the autopilot.
　
Autopilot disengaged (alternate method):

Hold the wings level with the control wheel using the ADI sky pointer for reference.
Use rudder trim to correct (stop) heading drift.
For airplanes with IRS, display heading on the IRS overhead panel. Use the tenths of a degree window to zero heading drift.
Use aileron trim to remove any control wheel force.
If the control wheel is displaced more than approximately 1.6 units (10 degrees) spoiler deflection will be initiated and a significant increase in aerodynamic drag will result.

What bugs me is the part "If the airplane is properly rigged, this should result in an approximately neutral wheel", because most of the time when I trim the airplane (737NG) according to this procedure I get an approximately 0,5 to sometimes even 1 unit right wing down indication on the control wheel.

Where does that come from?

I have a theory, but I couldn't find anything backing up my woolgathering so far. Maybe somebody here will be able to shed some light on this. Here it comes:

The CFM 56-7B has a redline N1 of 5380 rpm, so I would think that during normal cruise those fans spin at around 4600 - 4800 rpm. Fan diameter is 1,54m, so if I recall this correctly from my physics classes decades ago there should be an enormous amount of torque on the wing. Both engines rotate in the same direction (clockwise if seen from behind), so the resulting turning tendency is opposite to that, making the airplane roll to the left. That left turning tendency has to be trimmed out with the aileron trim, resulting in that tiny but nevertheless appreciable right wing down indication on the control wheel.

Every pilot with a multi engine rating remembers the four factors which make one engine more critical than the other in case of a failure. Three of those factors do not apply to jet aircraft, but torque certainly does, or doesn't it? And since jet engines usually are not counterrotating, there is definitely an asymetric torque situation even with all engines operating normally.

I know for a fact that Boeing employs thousands of engineers who are at least a trillion times smarter than me, so why wouldn't they mention that?

Is there maybe some design feature that already counteracts this asymmetry? Are the low and high pressure shafts of the engine counterrotating? That might reduce some of the resulting torque, but not all of it. Hm...



Anybody?

[David Horn]

When it says "rigged" I assume it refers to the setup of the various control cables - as in boats etc.

[PantLoad]

No...

Boeing's expanded procedure:

Make sure the wings' fuel levels are in balance. Make sure the thrust, between the two engines is the same (N1, EPR....whatever), neutral aileron trim. Hold wings level...perfectly level. Watch the heading. Counter any heading drift with rudder trim. Once the airplane holds a constant heading with the wings perfectly level, trim out any wheel force (aileron) with aileron trim.

Airplane out of rig means that the thing may be 'bent'. Or, spoilers may be not completely down, or airlierons may not be faired...or whatever.

[NGjockey]

Absolutely. But even with all the mentioned prerequisites fulfilled I'm almost always getting this right-wing-down indication on the yoke. In fact, I can remember only one certain aircraft in our fleet (and we have a bunch of them) that showed a LEFT-wing-down indication, but that ship must have been bent or rigged the wrong way.

So it's either me constantly mistrimming the aircraft, or Boeing's statement that a correctly trimmed airplane should result in an approximately neutral wheel is not completely thought-out.

I've had the pleasure of flying high performance turboprops for a long time, and that made me very sensitive as to trimming an aircraft. Every now and then the autopilot would quit, and we had to fly manually for days on end. Nowadays I like to klick George off for a few seconds after reaching our cruise level, just to get an idea what the flight controls are doing, and that's when I noticed this strange need for a little right wing down aileron trim most of the time.

There is one thing that would back my theory that it is a torque related problem: When we start our descent and the thrust goes to idle, the airplane does not keep the wings level any longer on a straigth path. Since the speed doesn't change that much during the transition from cruise to descent this can only be caused by two factors: A change of the AOA and/or decreasing torque due to the thrust reduction. But then again, the AOA wouldn't change that much if speed doesn't change, right?

Which leaves only torque as the culprit...

[TyroPicard]

If the airplane is properly rigged, this should result in an approximately neutral wheel.

according to this procedure I get an approximately 0,5 to sometimes even 1 unit right wing down indication on the control wheel.

Therefore the a/c is not properly rigged, unless your "approximately" is the same as Mr. Boeing's.

[Love_joy]

I'm not that up on the CFM56, but most engine designs with multiple spools have different parts of the core counter rotating to mitigate torque effects.

Of course, triple spool will have two in one direction but I would guess these are balanced so there is an equal amount of mass going in each.

73's are still cable rigged, and such large rigging will always result in some tolerance issues - thus your little bit of trim in flight.

[kenparry]

so if I recall this correctly from my physics classes decades ago there should be an enormous amount of torque on the wing. Both engines rotate in the same direction (clockwise if seen from behind), so the resulting turning tendency is opposite to that, making the airplane roll to the left. That left turning tendency has to be trimmed out with the aileron trim, resulting in that tiny but nevertheless appreciable right wing down indication on the control wheel.

Every pilot with a multi engine rating remembers the four factors which make one engine more critical than the other in case of a failure. Three of those factors do not apply to jet aircraft, but torque certainly does, or doesn't it? And since jet engines usually are not counterrotating, there is definitely an asymetric torque situation even with all engines operating normally.

No - you are forgetting Newton's Third(?) Law, that action and reaction are equal and opposite. There is only a torque on the airframe (via the engines) if those engines exert a torque on the airflow/gasflow through them, which they do not. The exhaust stream is not rotating. Look up the back end of the engine, and you will see that the support struts for the tailcone behind the LP turbine have a slight twist, which is there to straighten the gas flow.

If you don't like that explanation, try this one. Your LP spool is rotating at, say, a steady 5000 rpm: is there a net torque on it? No - if there was, it would change speed. There is torque in the LP shaft, of course, transferring power from the turbine to the compressor & fan stages that it drives. Same with the HP spool. And, for RB211 & Trent aficionados, the IP spool, too.

The point is that all the internal torques balance out to a net zero in the design case for which those support struts are set. Off-design, there will be a very small torque. Enough to see through the described trim procedure? I don't know.

Getting the airframe rigging right involves a huge number of settings: flap segments, ailerons, spoiler panels, even leading edge slats & flaps, and there is almost endless opportunity for small tolerances to add up to a measurable out-of-trim condition. On the 737s I flew, each was different - and I suspect that some of the older airframes were a little bent, something else to consider. Remember aircraft off the same production line are not identical - every part is made to a set of defined tolerances.

[misd-agin]

737 NG, using HUD, results in zero bank deflection when trimmed.

[aa777phil]

Rigging

First of all the construction of the aircraft must be square and true. It is not uncommon that aircraft from the same fixtures will all exhibit the same right or left roll tendency. Now don't forget that after they take a few punishing landings they can become a little "twisted". I recall a few 727-100's that would have quite a surprising roll on liftoff.

[glhcarl]

What bugs me is the part "If the airplane is properly rigged, this should result in an approximately neutral wheel", because most of the time when I trim the airplane (737NG) according to this procedure I get an approximately 0,5 to sometimes even 1 unit right wing down indication on the control wheel.

Prior to this statement only the "rudder" has been trimmed. The "rudder' is not connected to the control wheel. So control wheel should be in neutral position.

With the autopilot disengaged, hold the wings level with the control wheel using the ADI for reference. Trim out any control wheel forces using aileron trim.
Note: Use of aileron trim with the autopilot engagaed is prohibited.
If desired, re-engage the autopilot.

This step trims the "ailerons". Any movement of the "roll trim wheel" will result in wheel displacement, as the wheel is connected directly to the ailerons.

[Terraplaneblues]

Trim the rudder in the direction of the down wing

The "down" wing, this direction goes against any theory of flight ideas you would normally apply regarding lifting a wing with the rudder on a swept wing machine (it's because the autopilot is engaged) The technique appears to force the autopilot's hand by worsening the wing drop to take the broad roll null beyond the null margin. i.e. to an amount of roll that registers with the autopilot, in order that it can correct it with a roll input in the opposite direction and thereby lift the down wing.

Although you don't mention which of the 2 methods you are using, if you use the preferred method, do you follow up by disengaging the ap as in the latter part of the preferred method?

Or have I totally misunderstood as usual?

[muduckace]

I recall a few 727-100's that would have quite a surprising roll on liftoff

That is a result of improper shimming of the flaps. I have personal experience of only one operator checking this and it took 3-4 test flights to correct. The O/B flap shim would be checked at lower speeds clean and the I/B flaps at higher speeds, we had charts with calculations for this it was a scientific process. Now on liftoff with flaps extended a bit this would have effect as one O/B flap could be producing more lift than the other.

[AerocatS2A]

The "down" wing, this direction goes against any theory of flight ideas you would normally apply regarding lifting a wing with the rudder on a swept wing machine (it's because the autopilot is engaged) The technique appears to force the autopilot's hand by worsening the wing drop to take the broad roll null beyond the null margin. i.e. to an amount of roll that registers with the autopilot, in order that it can correct it with a roll input in the opposite direction and thereby lift the down wing.

This happens because the autopilot is using aileron to correct a constantly changing heading due to the out of trim rudder.

With wings level and an out of trim rudder, the aircraft will gradually turn. The autopilot senses this and does the only thing it can do, turn in the opposite direction by banking. By trimming the rudder you remove the heading drift and the autopilot removes the bank in response.

(This all assumes the autopilot doesn't have direct rudder control beyond a simple yaw damper.)

[Idle Thrust]

Despite kenparry's detailed rebuttal on the subject of torque, I think NGjockey is onto something.

Years ago my mob re-engined six of our DC-8-63's with CFM-56 engines making them -73's, at the same time retaining two aircraft in the -63 configuration with the P&W engines installed.

From then on all of the -73's required one or two units of rudder trim in cruise when the procedure described above was carried out. The -63's did not. I assumed that there was some sort of "torque" effect involved due to the big fan but kenparry pretty much torpedoes that theory. And I agree with his logic and explanation.

So what was going on?

[Terraplaneblues]

Thanks AerocatS2A

Now I see the logic of it, thanks.

Regarding the original question, there doesn't seem to be any provision for the gyro effects of the fan(s), which must be huge?

On the Classic & NG the troubleshooting for excessive trim test flight gives Flaps up Aileron 3/4 unit & Rudder 1/2 unit in Cruise and is usually less when tested. The troubleshooting section for excessive trim is very comprehensive due to the huge 737 database.

[kenparry]

Regarding the original question, there doesn't seem to be any provision for the gyro effects of the fan(s), which must be huge?

There are gyroscopic couples, but only when the aircraft is pitching or yawing. Huge? I don't think so in transport aircraft, where the rotating mass is small in relation to total mass. But the design of the engine mounts has to allow for it.

In an earlier post, somone supposed that 2-spool engines have counter-rotating spools to reduce that effect. AFAIK, the only counter-rotator is the Pegasus in the Harrier & AV-8. The rotating masses are probably a higher percentage of the total than in any other type. The prototype engines had same rotation, but it was changed very early on to the counter-rotating spec purely to reduce the gyroscopic effects.

[Idle Thrust]

GEnx is counter-rotator:

The GEnx Theatre

[singleseater]

Not sure about the 73 but for the 777 here goes:
The auto pilot has no input to the rudders or ailerons during normal climb/ cruise/ descent, only during the app with flaps not up (This is why the fly by wire can not compensate). So the flaperons are the only control it has available.
Kenparry is right, rotoating mass torque is very small, but the airflow has the "twist" taken out of it by the airflow stab fins, but in doing so they are subject to side forces, if viewed from the back, they are twisted clockwise.
This results in a net twisting of the whole engine assembly.
We effectively have a captured gyro, moving forward, spining right, so preccession is left, causing the right wing to rise. The auto pilot puts in flaperon to compensate. By trimming right wing down, you are taking this load to the rudder, which is lower drag for the same control result (due to moment arm ) and a more efficient flight.
In the 777, amount varies with power, speed (TAS) and engine type (Fan Diameter and mass).
Some older turbo-props had offset rudders to compensate (2 Degs for the 748/Andover) Not sure about newer ones.

[NGjockey]

I would never have thought of precession as an approach to this problem, but this seems to be a very interesting train of thoughts.

I'm still wondering, though, why Boeing would simply ignore that effect and stick to this "If the airplane is properly rigged, this should result in an approximately neutral wheel" - statement? My observation has been that because of this statement most pilots trim the wheel to the neutral indication, resulting in some very small, but still noticable bank. In the strict sense, this could be called "uncoordinated flight", which is definitely not what we are aiming for in terms of fuel economy and airframe stress. I know that this small out-of-trim condition will not bend the airframe or increase fuel consumption drastically, but I still find it peculiar that it is simply being ignored by the manufacturer.

[kenparry]

Idle Thrust : thank you, I stand corrected.


A few misconceptions here though:

but the airflow has the "twist" taken out of it by the airflow stab fins, but in doing so they are subject to side forces, if viewed from the back, they are twisted clockwise.
This results in a net twisting of the whole engine assembly.

Well, no. The twist was put into the gas flow by the engine, so taking it out simply retores the status quo ante. What you call "stab fins" generate a torque in roll, not a side force. And this results in zero overall net torque on the gas flow, and thus on the engine: straight flow in, and straight flow out.

We effectively have a captured gyro, moving forward, spining right, so preccession is left, causing the right wing to rise.

No, not so. Linear motion has no precessive effect, whatever its direction. Precession only occurs when there is a torque applied to the gyro in a plane in which the gyro's axis lies. No torque, no precession. In the case of an engine rotating about a longitudinal axis, a yawing torque will produce precession in roll, and vice versa. In this case, there is no roll or yaw torque applied to the gyroscopic mass.

Some older turbo-props had offset rudders to compensate (2 Degs for the 748/Andover) Not sure about newer ones.

Props - different case altogether. The engines apply a net torque in roll to the airframe (except in the case of aircraft with coaxial contraprops, such as the long-gone Gannets and Shackletons, and those with handed engine rotation - and then only in the balanced all-engines case).