4
$\begingroup$

I am part of an open-source project developing a simulation of a Cessna 172P. We are tweaking our flight dynamic model at the moment and we are looking for feedback from people with real life experience on the 172. In particular, we would like to know how much rudder input does this aircraft require during a take off roll (with no winds) in order to keep it centred on the runway.

$\endgroup$
  • 6
    $\begingroup$ I don't think you can get a numeric answer, short of putting a force sensor on the rudder pedals. The only answer I could give is "enough". You do it by feel, not by numbers. $\endgroup$ – jamesqf May 26 '18 at 19:12
  • 2
    $\begingroup$ Shouldn't your team be flying the 172 pretty much everyday? Reminds me of a programmer friend who worked at an Amazon warehouse one holiday, there was lots of automation to help pickers do their job, but it was incompetent, like the developers had never seen an Amazon warehouse, and so pickers were badly underused and wasted time on an industrial scale. I'd have forced the Dev team to work in the warehouse 1 day a week. And the holiday rush. So get flying folks! $\endgroup$ – Harper - Reinstate Monica May 26 '18 at 23:32
  • 2
    $\begingroup$ What's harsh is that you're trying to write a simulator for something you have never experienced for yourself. Enthusiasm is grand, but know your limits and know when you need help. Get some actual 172 flyers NDA'd in and on your design team, and then listen to them. $\endgroup$ – Harper - Reinstate Monica May 26 '18 at 23:55
  • 3
    $\begingroup$ One thing that makes this difficult is that PC-level sim aircraft controls aren't flown like real aircraft controls. With the usual PC joystick, you fly by stick position. But with a light aircraft, you fly by control force rather than by control position. $\endgroup$ – Wayne Conrad May 27 '18 at 5:52
  • 1
    $\begingroup$ @gilberto.agostinho.f Goodwill alone is not enough. Imagine a reverse case: you're a pilot and you've started writing a simulator - but you have no idea about programming at all so you come asking on stackoverflow on how are functions written. You'd get same answer - sent to "programming for beginners" course. Harper is 100% right on this: First you need to know what you want to achieve, then start going there, not the other way around. Most of programming is getting the requirements right. $\endgroup$ – Agent_L May 27 '18 at 6:16
9
$\begingroup$

With several hundred hours logged on multiple variants of the 172, I would give the relative answer of ‘not much’. In my experience a C-172 tracks pretty straight with only slight right rudder pressure needed on the ground roll. It is much less than what is required for a C-182, 206, 210 or other high performance single engine Cessnas and much less that a lot of tailwheel airplanes I’ve flown.

It seems like in a 172 when you are brakes off and rolling, I usually rest my heels on the cockpit floor and apply a slight pressure on the base of the right rudder pedal with the ball of my right foot. It is necessary as the plane will waltz to the left during the takeoff roll if you don’t. But the force applied and the deflection of the right rudder pedal is light and limited, if you do it right.

A recommendation might be to send a developer down to a local flight school operating the C-172 and arrange to practice takeoffs and landings with a CFI so they can get an empirical sense for what rudder input is needed on a takeoff roll.

$\endgroup$
  • $\begingroup$ "Not much", true, yet when I first flew C172 after hours of initial training in C152, I thought: wow! that's a lot of rudder! $\endgroup$ – Zeus May 30 '18 at 8:06
3
$\begingroup$

Wayne Conrad made a very important point in the comments which deserves serious consideration. Namely, flying is done by force, not by position. 'Apply that much pressure' rather than 'move by that much'. (This, in fact, applies to most well-designed controls of such kind, for example, brakes in your car).

This is beneficial for dynamic control because we have a better force feedback in our muscles than a position feedback.

But this raises two problems for simulation. First, a certain amount of pressure (which normally provides a certain amount of net effect) translates to different amount of movement in different conditions, particularly at different speeds.

For takeoff roll this is even more complicated because the nose wheel works proportionally to position (and even this is not quite true because the linkages have springs), while the rudder works proportionally to force, and one gives way to another during the roll. In practice, I'd say, you need to gradually increase the pressure while the amount of actual deflection reduces!

Second, all stock PC gaming controls work by position and cannot re-create a realistic force feedback. So unless you are building a simulator with specialised control loading, you can give up any hope of creating a simulator that feels even remotely correct at the controls. This is, perhaps, the source of much disagreement between pilots when they are asked to evaluate such a simulator.

For this reason, it may be better not to pursue 'true' control inputs for a PC simulator. It may be unflyable. For example, if I say that C172 requires about 10% or rudder deflection during takeoff (which is not too far from the truth) and you implement it thusly, I bet I myself would find it difficult to keep the model on the centreline. The truth is, what I remember is not the 10% deflection but rather (say) a couple of kg of force that is required. Yet this force will push most PC pedals to the limit, or at least much farther than 10%. And doing 10% corrections would be difficult, because the force required is so light. So in the end it may feel more real, or at least more flyable, if you make it 50% rather than the 'true' 10%: it will be closer to the true force, which is far more important.

But then, how do you simulate full deflection? Well, you cheat. You add significant non-linearities (S-curves) to the controls, you treat on-the-ground conditions differently to flight, etc. This may sound anathema, but it may be a better compromise and may create a better simulator.

$\endgroup$
  • $\begingroup$ Thank you for turning my comment into an answer, and much better than I could have. $\endgroup$ – Wayne Conrad Jun 2 '18 at 13:21
1
$\begingroup$

it's difficult to answer this question. depends on the engine, how you apply your power, the wind etc. there is no definitive answer as to how much, you just give the airplane what it needs at that given moment. even if there is no wind, the Left turning tendency of the airplane will still require some right rudder to be applied.

$\endgroup$
  • 1
    $\begingroup$ As I wrote in my comment above, I completely understand on how difficult this question is to be answered, but we are looking for some general opinions at least. I am assuming no wind (as written in my question) and we are modelling the model P which has a 160hp Lycoming O-320-D2J engine. I was just hoping I could get some sort of figure such as 'I would guess I apply around 15-20% rudder deflection during the take off roll' for instance. $\endgroup$ – gilbertohasnofb May 26 '18 at 19:55
  • 3
    $\begingroup$ But there's the problem: I wouldn't know what 15-20% rudder is, I just keep the nose pointed down the runway, which is a matter of feedback between the eyes and the feet rather than thinking "I need this much rudder". And of course the force varies during the takeoff roll. Seems like (as a programmer myself), you would do better computing needed rudder deflection from a mathematical model of the forces involved. $\endgroup$ – jamesqf May 27 '18 at 16:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.