Hand flying a 737 flight simulator

[Tee Emm]

When I first flew a real 737-200 in 1977 the Boeing instructor pilot told me that it could be flown with just two fingers holding the control wheel. In retrospect, he probably wasn't far wrong. But I don't recall ever having later problems using one hand to fly the 737 Classic series. But that was over 35 years ago and memories are not always reliable. This recent extract below from a You Tube movie about flying a Super Constellation seems to also back up the comment (or is it a myth) that even hand flying a 747 or MD-11 can be done with a light grip with two fingers in normal flight.

“We have a little bit of hydraulic helping us, but it needs a little force to handle it,” says Frei, who is operations director of the “Super Connie’s” Basle-based operator, the Super Constellation Flyers Association. “It’s not flying like a Boeing (http://www.flightglobal.com/landingpage/Boeing.html) MD-11 (http://www.flightglobal.com/landingpage/McDonnell%20Douglas%20MD-11.html) or 747 (http://www.flightglobal.com/landingpage/Boeing%20747.html) with just two fingers"

I haven't flown a real 737 for over 20 years but I recently did a one hour session in a Level D (full flight) B737-300 simulator with most of it hand flying raw data. Flying with the left hand on the control wheel and right hand on the thrust levers, I found it quite heavy on the ailerons and at times during practice steep turns of 45 degrees angle of bank I needed both hands on the control wheel due to the stick forces involved. Even on instrument approaches the forces to actuate the control wheel in roll and pitch were heavier than I remembered from the -200 or 737 Classic old days. This was especially with trim changes occurring with flap configuration and thrust lever adjustment during manual flying of an ILS or VOR approach.

There was no way in the world that this particular simulator could be flown with just a two fingered technique. On the other hand maybe no 737 simulator can be flown with just two fingers. This leads me to believe if the "just two fingers" spiel was nothing more than an exaggeration used by instructors to accent the need not to over-control. Mind you, it could be that one's hand strength gets a bit weaker with advancing years :E


Like all Level D full flight simulators, this particular machine undergoes regular scheduled fidelity checks with a current 737 Classic pilot "flying" it. There have been no comments with regards to its fidelity and presumably that includes stick forces in normal manual flight. Is the 737 Classic any heavier in the rolling plane than (say) a 737 NG? Comments invited.

[cosmo kramer]

If the aircraft (737 300-900) is in trim and only minor changes are needed (like say in cruise), you can fly the aircraft with one finger. :p

If you are making major changes to flightpath/thrust settings/configuration you need a firm grip. Basically anything which causes the aircraft no longer to be trimmed for the current conditions...

...like you experienced with the steep turns - obviously the aircraft was not in trim to continuously do a steep turn without control column input. If you had trimmed it out for the condition (which we normally wouldn't do during a turn), you could have continued the turn with a light 2 finger grip as well.

By the way a Classic is a 737 300-500. A 200 would be referred to as an "Original" or colloquially as "Jurassic", now a days. ;)

Edit: Was a typo Denti, hereby corrected.

[Denti]

Actually, the -600 is the NG version of the -500, so the classic is -300 to -500. The -600 never really caught on, same as the A318. It was around five to six tons heavier for around 115 passengers than the -500, which is a lot of extra weight per passenger. I know SAS is still flying a few ones, dunno who else.

[Cough]

Jurasic had lighter controls than the classic...

[oceancrosser]

I have been type-rated on 5 different transport category types where we trained on full-flight simulators (of various levels). In every case, the airplane has been easier to fly than the simulator. Now the DC-8 was not a 2 finger airplane...

[111boy]

Have to agree, the aircraft is much easier than the sim, in the sim things keep going wrong !!

[AirRabbit]

For whatever it is worth, the following is quoted directly out of the regulatory requirements for flight control tests for airplane flight simulators.

It should be apparent to anyone that these tests, when competently completed satisfactorily, should verify that the control “feel” in the simulator should be VERY much like that in the airplane.


Static Control Tests:
Conducted in Ground Conditions
There are objectively measured tests for controls in all simulators. For example, the Roll Controller Position vs. Force and Surface Position Calibration test repeats what was actually conducted on an operational airplane. The test is conducted by recording the results of the aileron position and the spoiler angle while the cockpit roll controller (the wheel or side stick) is moved throughout an uninterrupted control sweep, meaning that the control (wheel or stick) is moved from neutral position to either the full left or full right position to the stop, and then reversed to the other stop, and finally back to the center position. Throughout the control displacement from minimum to maximum aileron/spoiler displacement in both directions, the position of the control wheel/side stick and both left and right aileron and associated spoiler positions are recorded, as is the pressure necessary to move the control wheel/side stick. The comparisons are to be exact to within the following limitations: ±2 lb (pounds) or 0.9 daN (deca-Neutons) for the breakout force (the force required to initiate movement); and then throughout the recording, ±10% or ±3 lb (1.3 daN) force, ±2° aileron, ±3° spoiler angle. These test are then spot-checked against in-flight data from tests such as, as engine out trims, or steady state sideslips.

These values are applicable to all 4 levels of airplane flight simulator, Levels, A, B, C, and D.

Dynamic Control Tests
Conducted in Takeoff, Cruise, and Landing conditions.
For underdamped systems: ±10% of time from 90% of initial displacement (0.9 Ad) to first zero crossing, and ±10 (n+1)% (where “n” is the sequential period of a full cycle of oscillation) of period thereafter.
±10% amplitude of first overshoot, applied to all overshoots greater than 5% of initial displacement (.05 Ad), ±1 overshoot (first significant overshoot must be matched).
For overdamped systems: ±10% of time from 90% of initial displacement (0.9 Ad) to 10% of initial displacement (0.1Ad). For the alternate method see paragraph 4 of this attachment. The slow sweep is the equivalent to the static test 2.a.2. For the moderate and rapid sweeps: ±2 lb (0.9 daN) or ±10% dynamic increment above the static force.

Data must show normal control displacement in both directions. Tolerance applies against the absolute values of each period (considered independently). Normal control displacement for this test is 25% to 50% of the maximum allowable roll controller deflection for flight conditions limited by the maneuvering load envelope

These values are applicable to Level C and Level D simulators.

[rudderrudderrat]

Hi AirRabbit,
It should be apparent to anyone that these tests, when competently completed satisfactorily, should verify that the control “feel” in the simulator should be VERY much like that in the airplane.
As human pilots we "feel" the delta g during turns and during any maneuver which changes the vertical speed (take off rotation, landing flare, TCAS maneuvers etc.) The delta g feed back loop is how we judge the aircraft is responding to our pitch inputs and prevents us from exceeding +2.5g (in unprotected aircraft) or uncomfortable reduced / -ve g during a rapid level off from a climb. I don't know of any simulator capable of reproducing those motion cues.

So mathematically, the sim and aircraft performance may be identical, but a human pilot will still tell you it doesn't feel the same, and he is correct.

[AirRabbit]

As human pilots we "feel" the delta g during turns and during any maneuver which changes the vertical speed (take off rotation, landing flare, TCAS maneuvers etc.) The delta g feed back loop is how we judge the aircraft is responding to our pitch inputs and prevents us from exceeding +2.5g (in unprotected aircraft) or uncomfortable reduced / -ve g during a rapid level off from a climb. I don't know of any simulator capable of reproducing those motion cues.

So mathematically, the sim and aircraft performance may be identical, but a human pilot will still tell you it doesn't feel the same, and he is correct.

I was attempting to point out the lengths to which at least some regulators go in order to provide the best possible match between aircraft and simulated aircraft “feel.” Of course, when you describe a 60-degree bank level turn at a constant airspeed and constant altitude in a typical airliner, you’re describing a “2-g” turn – and I certainly agree that such a turn in even the best flight simulator meeting today’s qualification criteria will not give the occupants a 2-g feel (and I’ll qualify that by saying that while I don’t fly the Airbus inventory, since its a matter of physics, not manufacturers, we wouldn't see any substantial deviation with that manufacturers products). Additionally, as I was referring to the post by Tee Emm in which he said:

When I first flew a real 737-200 in 1977 the Boeing instructor pilot told me that it could be flown with just two fingers holding the control wheel. In retrospect, he probably wasn't far wrong. But I don't recall ever having later problems using one hand to fly the 737 Classic series.
… if an airplane is capable of being “flown” with “2-fingers,” I would presume the reference would not include a 2-g, 60-degree bank level turn while maintaining the entry airspeed, even if other, more typical control was capable of being maintained with only “2 fingers.” My suspicion is that the “2 finger” reference was with respect to maintaining normal cruise flight, with normal maneuvering - and, perhaps, even in a nominal visual traffic pattern. Believe me, I certainly understand the limitations that airplane flight simulators confront every day – and, for the most part, it is my opinion that simulators quite satisfactorily represent what is likely to be experienced in the airplane – at least the “on-set” motion cue, which is then supplemented with visual and instrument (and sound when/where appropriate) indications throughout the completion of the maneuver … and, at least in the airplane operations with which I am familiar, I don’t see, and haven’t seen, even sporadically, a pilot achieve a 2.5-g circumstance, even at lift-off or go around, and certainly not a negative version of that during a level off.

[cosmo kramer]

… if an airplane is capable of being “flown” with “2-fingers,” I would presume the reference would not include a 2-g, 60-degree bank level turn while maintaining the entry airspeed, even if other, more typical control was capable of being maintained with only “2 fingers.”
Sure you can fly a 60 deg turn with 2 fingers in a 737. You just have to trim the back pressure away. Which we would of course not normally do, but indeed it is possible.

[Tee Emm]

Many thanks for all your replies. They are much appreciated. As a matter of interest, the steep turn normally required as part of an instrument rating test is based upon 45 degrees angle of bank and 250 knots. 60 degrees is real Gung Ho stuff and not really relevant other than unusual attitude practice.

Opinions vary, but some pilots prefer not to use the stabiliser trim in steep turns. One reason being steep turns are usually short term manoeuvres and if back trim is used during the turn to off-load the control force, rapid forward trim as well as slight power reduction is needed when rolling back to wings level. It can get complicated especially if conducted on instruments.

[ClimbSequence]

T.E.

I second what cough wrote.
The original are lighter than the classics and for me much more pleasant to hand fly as well.
One thing I have noticed so far in my experience is the larger the fuselage, the heavier the feeling of the controls.

Just my two cents

[AirRabbit]

As a matter of interest, the steep turn normally required as part of an instrument rating test is based upon 45 degrees angle of bank and 250 knots. 60 degrees is real Gung Ho stuff and not really relevant other than unusual attitude practice.
Of course, steep turns are typically flown, as you say, at some defined airspeed (often it is 250 knots) and at a bank angle of 45 degrees. The only reason I referenced a 60 degree bank, 2-g turn was in response to rudderrudderrat’s post …
As human pilots we "feel" the delta g during turns and during any maneuver which changes the vertical speed (take off rotation, landing flare, TCAS maneuvers etc.) The delta g feed back loop is how we judge the aircraft is responding to our pitch inputs and prevents us from exceeding +2.5g (in unprotected aircraft) or uncomfortable reduced / -ve g during a rapid level off from a climb. I don't know of any simulator capable of reproducing those motion cues. So mathematically, the sim and aircraft performance may be identical, but a human pilot will still tell you it doesn't feel the same, and he is correct.
…that seemed to be centered around a 2.5-g maneuver that would likely not be able to be realistically simulated in a flight simulator – regardless of it’s pedigree. I was attempting to defend the viability and excellent accuracy with which most flight simulators represent realistic, if only “on-set” motion cues – including the initiation of appropriate “g-force” cueing. In addition, I wanted to address the fact that since typically we don’t see maneuvers in the airplane that result in anywhere near 2.5-g accelerations, and since the references noted were to flying with “2 fingers,” I was merely attempting to point out that whatever task was being flown with “2 finger control” was not likely to result in something that would take at least a 60 degree bank, level turn – which requires right at 2-g’s of acceleration to accomplish.

[Tee Emm]

I was merely attempting to point out that whatever task was being flown with “2 finger control” was not likely to result in something that would take at least a 60 degree bank, level turn


Fully understood. Please keep your posts coming. They are always thoroughly informing.

[cosmo kramer]

Anyway, the point is, YES, a 737 can be flow with 2-fingers, so as long as it is in trim - regardless of bank angle.

[3 Point]

2 fingers works just fine throughout the flight envelop in an Airbus!!

OK, OK, I'm leaving now!! :p

[john_tullamarine]

One certification side issue. Stick load/g requirements are necessary to provide pilot safeguards in respect of overstressing during manoeuvring.

Ergo, it probably is quite poor form to trim in the turn. I suggest that the trim should be appropriate to the steady state situation (climb, cruise, descent) and that pilot plus his/her morning Willie Weeties is the answer to steep turn stick load fun and games.

[parabellum]

One thing I have noticed so far in my experience is the larger the fuselage, the heavier the feeling of the controls.


Have flown B737-200 & -300, B747-400, B757-200 and B767-300ER, honestly can't say there was a whole lot of difference! The B747 is a big, refined lady but control forces never seemed proportional to her size!


Never trimmed in steep turns, releasing the back pressure as the bank came off was, during my training sessions, the 'norm'.

[cosmo kramer]

[rant=on]
Arrrg... Read my WHOLE posts, instead of replying out of context. The original poster asked if it is possible to fly a 737 with "two finger". The answer is yes, if the aircraft is in trim. Even if in a steep turn you can fly it with two fingers, if you trim it out... That doesn't mean it's a good idea or practice. Infact I wrote in every post "which we would not normally do".

Now... Who the hell wants for fly a steep turn with two fingers anyway? [rant=off]

[john_tullamarine]

cosmo - we understand your view.

However, due to the educational aspect of this forum for the new chums, occasionally we need to put in some clarification to provide for their balanced reading.

[BARKINGMAD]

Re feeling like the real aircraft, simulators are #*$t¥&$ :mad: !!!

Just climb in, wobble the control thingy in the direction and at the rate which puts the needles/digits where the test requires.

Perform the correct procedures, carry out Q R H items religiously, demo good CRM, don't argue with the demigod assigned to assess you, then leave with your licence signed and get back on the line.

Next flight rejoice in the handling qualities of your real plane, and spend the next 6 months practising for the next sim session.

Isn't that the way to do it..........?

[piratepete]

Sadly you are 100% correct.

[Mimpe]

I suspect most people would agree the real thing is a bit easier!

[Jwscud]

There was an excellent thread recently on simulator fidelity where a lot of these issues were discussed in depth. I recommend a read:

http://www.pprune.org/tech-log/540987-simulator-training-strong-crosswind-landings.html?highlight=Simulator+crosswind+landing

[AirRabbit]

There was an excellent thread recently on simulator fidelity where a lot of these issues were discussed in depth. I recommend a read:

Simulator Training for strong crosswind landings


“JW,” my friend, you are 100% correct … and that you took the time to remind those here of that, speaks volumes about your professionalism as well as your commitment to facts and honesty. :D:D:D:D:D:D:D

[BARKINGMAD]

So why is it, despite the 9625s, FSTDs and part 60s etc bandied around over the years, that after 26 years of full-motion ZFT all singing and dancing simulators, I failed to find a single one which accurately reproduced the handling qualities of the REAL thing???

Maybe I've got totally insensitive muscular feedback and vision perception, though I managed to con Class 1 medicals up to 64.5 years old, but I still maintain they are a good procedures and emergencies practising tools.

Please trainers, management and company test pilots, don't delude yourselves, and by implication others, that the average airline sim is a faithful representation of a line aircraft, 'cos it ain't!!! :(

[captjns]

After 38 years on the Boeing, I've come to the conclusion that the simulator is more realistic than the actual airplane:}

[latetonite]

It sure is!

[AirRabbit]

So why is it, despite the 9625s, FSTDs and part 60s etc bandied around over the years, that after 26 years of full-motion ZFT all singing and dancing simulators, I failed to find a single one which accurately reproduced the handling qualities of the REAL thing???

Maybe I've got totally insensitive muscular feedback and vision perception, though I managed to con Class 1 medicals up to 64.5 years old, but I still maintain they are a good procedures and emergencies practising tools.

Please trainers, management and company test pilots, don't delude yourselves, and by implication others, that the average airline sim is a faithful representation of a line aircraft, 'cos it ain't!!!

I suspect that the answer to your question – why is it that, with the years of experience you’ve achieved, you’ve never found a single simulator that accurately reproduces the handling qualities of the “real” thing?” – probably has a LOT to do with how you define “accurate.” If “accuracy” to you means that you cannot tell the difference between the simulator and the airplane it represents, then you are certainly correct – simulators don’t have that degree of “accuracy.” But then, as I suspect you are already aware, there are some airplanes, on any given ramp, that would provide you a similar degree of difference between any one, specific airplane, and any other airplane you might want to designate … and even widening that consideration, as I’m sure someone who has your level of experience can certainly attest, rarely will one find 2 airplanes that “feel” exactly alike. Also, I think almost anyone with the level of experience you have will certainly agree that both the overall quality of simulation, and the quality of any specific characteristic of any airplane simulator has considerably improved over the past 3 or 4 decades – in both 1) the measureable categories (any measureable category you may specify) and 2) the finely tuned “feel” of very highly experienced flight test pilots.

There is a major piece of the premise for the allowed use of simulation for training and testing of pilots, that makes up a significant portion of the basic tenent for the dependable use of simulation, which is not, in any way, limited to the “accuracy” of the simulation (again, determined by both specific measurements and by personal opinion of those pilots with the most current and broadly based experience on the subject airplane) but is significantly based on the ability and dedication of the instructor/evaluator personnel assigned to provide such services. In fact, the specific reason that relatively old simulators (maintained against their original qualification standards) can, and regularly do, produce highly competent and very proficient pilots, is due to the fact that these instructor/evaluator personnel use simulation as a tool to train and test pilot applicants – the more modern and, therefore, the more accurate, is that simulation, the better the quality of the student in a shorter time, and often requiring less specific, repetitive involvement of the instructor/evaluator. Unfortunately, this fact has often been the substance of the unfortunately increasing problems seen in the quality of some recently trained pilots. The bottom line HAS to be that, regardless of how good – or not so good – the simulation is or is not ... simulation, in and of iteself, cannot be the SOLE participant in the training and/or testing of airplane pilots.

As I’ve said here, certainly airplane simulators have become vastly more accurate (regardless of how you define that term) with respect to the airplane being represented, but it has been, it is, and it will always be … simply a tool used by those who teach and test pilots. When a tool is used incorrectly or incompletely, one simply cannot, and MUST not, blindly presume that the desired result has been achieved. One cannot throw a tool box load of wrenches, pliers, hammers, and timing lights under the hood of an automobile, and then drive that automobile around the block a few times, and conclude that because all the correct “tools” were available and “used,” that the automobile has been “tuned up” and is “ready to go!” The same with pilots and the simulation they use. Without the proper knowledge and experience of the persons charged with the responsibility of designing, refining, examining, and conducting the training in which simulation is to be used, the simulator will not provide any higher degree of success than those hastily thrown-in “wrenches, pliers, hammers, and timing lights.” No one … student or instructor … experienced line pilot or novice … can either accept or reject a specific airplane flight simulator without including in that acceptance or rejection the efforts of all involved … from the simulation designers, the engineers, the data collectors, the software engineers (who turn squiggles and numbers into believable control column forces, sounds, and surprisingly realistic visual scenes in front of your eyes), the regulators, the instructors, and the evaluators … all of whom must combine their talents and training to be able use a very special and very sophisticated training tool … where the emphasis MUST be on the last word in that description – “TOOL” - the flight simulator.

Therefore, Mr. BARKINGMAD, as long as airplane flight simulators are (as you said) “…good procedures and emergencies (sic) practising tools…” and if it is true that what is learned, practiced, and ultimately performed when operating the “real” airplane is proper and correct … I’d submit that the “tool users” and the “tool” itself have both performed the jobs they each were intended to do. Sure, we can have the "does it handle and perform JUST LIKE the airplane" conversation to satisfy whomever is interested. But I think our time would be better spent looking at the result of the use of simulation and the amount of explanation, the kinds of explanation, and the accuracy of any such explanation, that is required to produce a competent, safe, and knowledgeable pilot. I'm of the opinion that the quality of the pilots today - after using modern simulation - built and evaluated in accordance with "the 9625s, FSTDs and part 60s etc." that you referenced, are better prepared and more knowledgeable than ever before. My concern is the coming degradation in the experience levels of new airline pilot applicants - and what kind of training will be required to provide the same knowledge and understanding - and where/how that knowledge and understanding is going to be achieved ... and I am certain that simulation - of some kind or level - will most certainly be involved.