FG's Cessna vs Real life.

[Megaf]

Hi all, so yesterday I piloted a 1977 Cessna 172N, upgraded to 172P (Engine replaced), in real life.

Important Observations

• About the Cessna 172 in real life. Man, in real life, it is an absolute pleasure to fly it!
• Below 90 knots, it is extremely forgiving, you can make huge inputs and the thing barely responds.
• A 60 knots climb, it's extremely sharp, and your nose will be pointing to the sky. A shallow climb will be at 85 knots.
• The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME. In fact, I didn't even notice that in real life.
• Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.
• Tyres don't squeak like that, like someone is raping a chicken.
• It accelerates a lot quicker than it does in FG and stalls at 45 knots.
• The controls are a lot less sensitive overall. I strongly advise to have a look at Octals/Joshs PA-28 on how to implement proper inputs from the joystick/yoke.

Below some photos of the aircraft I flew in.

The good in FG's Cessna.

• Start procedure is spot on, just like in real life.
• Engine sound is ok enough.
• Engine startup is alright, but the aircraft shakes too much in FG.
• General dynamics are ok too.

[wkitty42]

you do know that you can adjust the trim to counteract the left movement, right? there is a trim tab on the rudder... and the seats are adjustable... if you turn on the craft save state, the seat position will be retained for the next flight(s)...

[Megaf]

you do know that you can adjust the trim to counteract the left movement, right?

The rudder trim is optional, and the aircraft I flew in real life didn't have any sort of rudder trim.

and the seats are adjustable...

I am well aware of that and, in the Cessna, I need to adjust that all the time. Shouldn't the seat position be optimal by default? I can actually see a point where you need to adjust it once, in real life you do adjust the seat height once and then leave it there, assuming you are the only one flying the plane.

if you turn on the craft save state, the seat position will be retained for the next flight(s)...

Oh, I didn't know that! Thanks for sharing this information. Perhaps that should be the default?

Now, the seat adjustments, you mean the option in View > Adjust View?

[Megaf]

@wkitty42 I tried the "adjust view height" option and it doesn't persist between sessions. However, I just learned about the Save/Load slots for states, that works fines! Thanks!

[wkitty42]

Now, the seat adjustments, you mean the option in View > Adjust View?

grep -i -e "seat" c172p-keyboard.xml ;)

[Megaf]

reglnx@NeuX230:~/FGB/FlightGear-Common/Aircraft/c172p$ grep -i -e "seat" c172p-keyboard.xml
        <desc>Reset to saved seat view</desc>
        <desc>Seat Down</desc>
        <desc>Seat Down</desc>
        <desc>Seat Up</desc>
        <desc>Seat Up</desc>
        <desc>Seat Forward</desc>
        <desc>Seat Forward</desc>
        <desc>Seat Back</desc>
        <desc>Seat Back</desc>
        <desc>Seat Pitch Down</desc>
        <desc>Seat Pitch Down</desc>
        <desc>Seat Pitch Up</desc>
        <desc>Seat Pitch Up</desc>

[tonghuix]

I fly C172SP in real life, 180HP. I agree with you, the FDM of FG 172P is far more from the real life.

[wkitty42]

reglnx@NeuX230:~/FGB/FlightGear-Common/Aircraft/c172p$ grep -i -e "seat" c172p-keyboard.xml

the goal was to see and show you that the craft has seat positioning within it... you can see the defined seat movement keys, doubled for upper and lower, as well as one (lower 'q' IIRC) to reset the position to the current saved seat position... so you don't need to use the sim's view positioning but you can to augment if desired...

[wlbragg]

Thank you @Megaf for the details, it is really appreciated. I've been taking a break from development for the last month or two but just today did some work on the fg1000 variant and want to get some of the existing issues that are close to finished wrapped up, fg1000, variant, new kap140, compositor upgrades.

I will review your observations and get back to you, if possible, for more details as we try to address these issues.

One quick question about the screeching tires. I assume you talking about the screech that happens when you press the brake too long? Do you know if you can apply enough brake to lock the tires up on the c172p in real life? I imagine that would be an extremely rare event, but is it possible to put the craft into a skid by slamming on the brakes or the parking brake? The sound your hearing is the tires being locked up, as in a skid on pavement or dirt (different sound). The problem is there is no feedback in the sim and the brakes are not analog. So it is not a perfect simulation of how it would work in reality. You don't hear that sound unless you have the wheels not turning but the aircraft is still moving across the ground or pavement, in other words, you don't hear that sound until you hold the brakes down long enough to start skidding. So I submit, if you even can "lock the brakes up" in real life, you would most certainly hear that sound. In the sim, when you hear that sound, you are moving across a surface without the wheels turning.

The shaking "too much", is that the short shake at start-uo that was recently added?

The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME

Are you referring to the pull to the left on take off while still on the ground, or early on in the air, or always? Can you be really exact in the description of this effect or lack of?

[dany93]

Thanks @Megaf for this detailed report.

Part of your remarks are hard to fulfill, the aircraft's FDM is the result of compromises. Moreover, coupling between controls and effects often results in a change somewhere which breaks something elsewhere. However, if you find better compromises which does not break anything else which already works, please tell us.

JSBSim imperfections may also add to this, useless to say that this subject is very difficult and we (basic modellers) cannot change this. To submit improvements or remarks, we have to be certain that there is a flaw and accurately report it to the developers.

About the roll to the left: I think that you mean the roll at low speed and full throttle in the air, typically at climbing. Not "All the freaking time", as you wrote. Obviously, it is due to the propeller torque. I checked several times by power calculations, this torque is correct in JSBSim. It inevitably exists in real life and must be compensated. A difficulty is that is varies with on the propeller torque, which itself depends on the engine regime, throttle, aircraft airspeed. In real life, this trimming is either done by the the mechanics (preset compromise after tuning on the ground) or by a controllable aileron trim, if available. In FG (without aileron trim control), it is a compromise too. Currently in the c172p, this setting is such as the aircraft is approximately balanced close to 110 - 115 kt, 2300 - 2400 RPM (I don't remember exactly, but in typical cruise conditions). It can be changed on the fly in the Internal properties, but not saved. If you want to change this trim preset value and keep it saved, you can in c172p-main.xml, line about 594:

    <controls>
        <flight>
            <aileron-trim type="double">0.022</aileron-trim>

About the general behavior (control sensitivity): often reported in simulation, mostly for light aircraft, very hard to fix. Part of it is the result of a feeling due to our joysticks. Springs are often harder than in real life, and we have no force feedback. For example, if we abruptly loosen the JS, it too quickly reaches back to zero. This gives a false feeling. Also, one should keep in mind that the JS sensitivity for each axis must be such as it can reach the max deflection (full effect), e.g. the ailerons must be powerful enough to hold a forward slip or a landing with strong crosswind (crossed controls); the elevator must enable to reach the stall AoA at landing. This for a given stick deflection (0 - 100%), often mechanically short. In the FDMs, these controls are generally linear. They can be made softer near the center by a quadratic or power 3 response in the JS binding file, but at the expense of a steeper response at high deflections. This can also be done and saved in the FDM file, but not so easily. I do not like this too much because it is a kind of trick, a workaround.

Once more, if you find better compromises which do not break any thing else, please submit your proposals such we can test them. You are in the best position to test this if you pilot a c172.

For the brakes do you have rudder pedals (progressive effect) or do you use the keyboard by short touching?

[Megaf]

@wlbragg welcome back :)

The shaking "too much", is that the short shake at start-uo that was recently added?

When you start the engines, in real life it is much smoother, the aircraft moves just a little opposite to the spin of the engine. It's not your head that shakes, it is the aircraft that rolls slightly, very slightly.

I assume you talking about the screech that happens when you press the brake too long?

Yep, and when turning sharply. In all fairness, you might be correct here in leaving that as feedback. In real life, you have a very different feeling. And the brakes are digital, so you have a valid point.

Are you referring to the pull to the left on take off while still on the ground, or early on in the air, or always? Can you be really exact in the description of this effect or lack of?

In sim the Cessna has, in flight, a constant roll tendency cause by the rotation of the propeller, rolling in the opposite direction the propeller is turning. To counteract that, we use the rudder. In sim, this is too strong, so strong that even on the ground it makes the aircraft turns. I surely don't have a lot of experience, I have exactly one flight in real life, but I didn't notice any turning on the ground nor in the air. I flew it myself, with full controls, for around 40 minutes and barely had to touch the rudder pedals. @Octal450 your input would be very much appreciated in here, about constant roll tendency, I understand the PA-28 has a similar engine and propeller.

@dany93 Hi :)

Obviously, it is due to the propeller torque. I checked several times by power calculations, this torque is correct in JSBSim. It inevitably exists in real life and must be compensated. A difficulty is that is varies with on the propeller torque, which itself depends on the engine regime, throttle, aircraft airspeed.

I understand that. The thing is, in sim, without using the rudders, FULL AILERON might not be able to compensate for it and, that is just plain wrong. In real life, not only the effect is much weaker, but a lot less input is required. In the sim often times I see myself applying nearly half rudder trim to compensate. IMHO, if the effect was reduced to half of what it is now, it would be better already.

About the general behavior (control sensitivity)...

I agree! It's very hard to adjust that. I just got a new joystick today, a T16000M. Before, I was trying in the Extreme 3D Pro. Will do more testing and get back to you.

@tonghuix Do you fly the Cessna 172P in real life?

[dany93]

In sim the Cessna has, in flight, a constant roll tendency cause by the rotation of the propeller, rolling in the opposite direction the propeller is turning. To counteract that, we use the rudder.

No. Propeller torque (roll tendency) is to be counteracted by the ailerons (or an aileron offset). The rudder is for other effects (see below). And, as I wrote in my previous message, the aircraft roll tendency is neutral at cruise speed.

In sim, this is too strong, so strong that even on the ground it makes the aircraft turns.

This is normal and well known. I remember on an ultralight IRL, at taxing before takeoff, we would have been on the left out of the runway with no rudder correction.

Asymmetric Thrust Explained Left-Turning Tendencies Explained: Why Your Plane Pulls Left During Takeoff

.... I didn't notice any turning on the ground nor in the air. I flew it myself, with full controls, for around 40 minutes and barely had to touch the rudder pedals

You might well be the only one to notice no turning tendencies on a single-engine light aircraft....

The thing is, in sim, without using the rudders, FULL AILERON might not be able to compensate for it and, that is just plain wrong. In real life, not only the effect is much weaker, but a lot less input is required. In the sim often times I see myself applying nearly half rudder trim to compensate.

I did tests (once more...)

Wind = 0 (or at least no crosswind), 180 hp, Pilot alone, Taxiing, then climbing 70-80 kt (full throttle). 1) Taxiing before takeoff: rudder 0.15 max (decreasing with airspeed) 2) Climbing after takeoff:

• No rudder correction (rudder = 0, bad way, slip ball off center!) : aileron 0.15 - 0.2 (0.24 briefly).

• Rudder 0.08 (for slip ball centered): aileron 0.05 - 0.1.

  Far from full aileron or half rudder. Even for the ailerons with the "bad" way of climbing with no rudder correction. I don't know where the problem is, but something seems not correct in your controls, your observations?

Which does not mean that the FDM accurately renders the real behavior, but the effects are realistic, observed and reported in real life. The only thing, I admit that the controls probably need a bit more sensitivity than the real one. But I did my best and couldn't find a better compromise. Either that's proper to simulation, FG?... difficult to fix. Or that's me, but please in this case, I suggest to those not pleased (@tonghuix ?) to propose a better FDM.

[wkitty42]

.... I didn't notice any turning on the ground nor in the air. I flew it myself, with full controls, for around 40 minutes and barely had to touch the rudder pedals

You might well be the only one to notice no turning tendencies on a single-engine light aircraft....

this is probably due to the craft being already well tuned...

[Octal450]

Propeller torque (roll tendency) is to be counteracted by the ailerons (or an aileron offset)

Just pointing out, in the real plane, the aileron yoke is neutral or near neutral. I do not really ever recall needing separate aileron input. Simply pushing right rudder 's roll moment is enough to counteract MOST of the tendency.

Kind Regards, Josh

[Megaf]

@dany93 Are we flying the same airplane? You are the first ever I see who says the roll is not strong, and the first ever I see who says the tendency is neutral in cruise.

I've seen lots of people giving up on the cessna because of what I stated.

[tonghuix]

@Megaf Yep, I plan to write a test report recently. We, with my instructor, tested this project compared with other FS platform (P3D and X-Plane), and also real C172.

[wlbragg]

There is something really strange happening here. For kicks I decided to try using only the optional rudder trim and no other input to take off down the runway. Setting the rudder trim to .120 allowed me to roll down the runway at full throttle. At the lower speed it started drifting left slightly and then as the speed increased it started drifting to the right. Point being, it took literally a tiny amount of rudder to counteract the full throttle torque. So little in fact I was able to use only rudder trim. So I submit, there is a big discrepancy in descriptions of how this issue affects users, why? From my perspective and on my hardware, there is no issue at all.

@tonghuix I will be really interested to see your report.

[Megaf]

@tonghuix Thanks so much for that. That will greatly help with the clarifications. May I suggest you fly in the sim the same engine (120 vs 180) you fly in real life and the same location. I've been testing with the same engine, same airport and same weather conditions.

@wlbragg I will make a video and upload it. Will use fair weather which isn't supposed to have strong winds.

[Megaf]

@wlbragg @dany93 @tonghuix https://youtu.be/drLXaidstww

So you can clearly see the amount of right aileron I need to make the Cessna stop rolling. You can also see that a bit of right rudder will stop it rolling.

[wlbragg]

Wow, I don't get that at all. I get slight left pulling on the ground until the airspeed is enough to allow the rudder to correct. I need to hook up my peddles or use my desktop though to really test this more. I'm on my new laptop and I have no independent rudder control.

OK, using auto-coordination I exhibited something a little closer to what your seeing, but not near as much.

I'll try to record this to show you what I get. But currently my controls are in disarray and it is difficult to cleanly control the aircraft.

Are you using auto coordination by chance?

[Megaf]

Are you using auto coordination by chance?

Nope. And again, I'm not the only one experiencing that. And just to confirm, I used "Fair Weather" with "Basic Weather".

[dany93]

I don't have to do such large actions, neither. It seems that, some times, you counteract too lately, then have to over react.

One thing: if you have the "Rudder Trim Option" activated, the rudder trim is at 0 per default (you have to adjust it yourself). I set it at 0.02 in c172p-main.xml, but it is at this value only if you start with "Rudder Trim Option" OFF. Important for the following.

I checked the balance at (some) cruise airspeed. 180 hp, default load (pilot alone, 1876 lbs), Trims: aileron-trim 0.022 (= default) rudder-trim 0.02 (= default, only if started with "Rudder Trim Option" OFF). Or set it in the Internal Properties.

Balanced (at least well enough) for: rpm 2250, airspeed-kt 104, elevator-trim 0.064

[wlbragg]

@Megaf after much more testing (while I try to get the JS configured), I am in agreement with @dany93 It is definitely related to your rudder trim and for sure you need that 0.02 or more to make it behave more closely to what your expecting. Try experimenting with the optional rudder trim and see if that make sense to you.

[tonghuix]

When we test it, we forgot to record a video, but good news I keep the written record.

[Megaf]

@wlbragg rudder trim is 0... I tested in 3 different systems and 4 versions of FlightGear, all show the same behaviour ¯\_(ツ)_/¯

[wlbragg]

What versions? I just checked 2020.4 and starting without rudder trim option active /surface-position/surface-pos-norm is set to 0.0199... As soon as the rudder trim option is activated /surface-position/surface-pos-norm is set to 0.0. Regardless, does setting /surface-position/surface-pos-norm to 0.2 help make it feel more like the real thing to you?

[wlbragg]

I'm checking now, but I think any released fgdata version will not have this logic programmed? I just checked 2020.3.9 and it had only .02 regardless of trim option being active. This c172p version is whatever is in fgdata next.

[dany93]

@wlbragg wrote

rudder trim and for sure you need that 0.02 or more to make it behave more closely to what your expecting.

0.02 (preset in c172p-main.xml) is a compromise between pilot alone and pilot + copilot. Yes, 0.02 to 0.03 or 0.04 is more appropriate with the pilot alone (which is default load).

But this must not be an issue. For tests, it can easily been tuned with the Rudder trim option ON, and with this option OFF, it can be changed in c172p-main.xml at preference for a default value at start. Also changeable in the Internal Properties in-sim.

[wlbragg]

But this must not be an issue. For tests, it can easily been tuned with the Rudder trim option ON

Yeah. that's why I want to make sure @Megaf test with the trim option. Because, as he stated

rudder trim is 0... I tested in 3 different systems and 4 versions of FlightGear, all show the same behaviour

So that does explain why it's pulling so much for him, if he is not changing that value.

[Octal450]

Hello please. see the changes by @HHS81 to the c182s. They have made the aircraft FAR more realistic to fly and better, perhaps the same can be done to the c172, it seems they use a similar type of aero modelling.

Kind Regards, Josh

[HHS81]

Not only that- while investigating the c182s yaw issue I found c172 specific coefficients from the same source as the c182s uses. I can bring them up here on weekend.

[wlbragg]

@dany93 maybe you can take the lead on this?

[wlbragg]

see the changes by @HHS81 to the c182s.

What specifically? That could be a pretty broad review.

[HHS81]

@wlbragg specifically a lot 😄 Give me some time on weekend

[dany93]

Thank you @HHS81 for your contribution.

I don't intend to do a broad refactoring, but I think this comparison can be instructive and useful.

To help you focusing on the important issues, without spending too much of your time: (the left-over can be tackled gradually if necessary).

I briefly tested the C182S and had a look at the FDM. @Octal450 wrote

They have made the aircraft FAR more realistic to fly and better

Aren't you exaggerating? Qualitativeley and subjectively (not easy...),I don't find the behavior is that far from the C172's. Please, can you explain with more details the main differences between both, and issues, from your point of view?

@Octal450 and @HHS81 I've mainly noticed two points in C182S's FDM:

• The adverse yaw is inverted. This probably makes the aircraft easier but I think this is a loss of realism. Its amplitude can be discussed (it has already been in the past for the C172P) but the adverse yaw is real on every such aircraft.
• The empty CoG is behind the AERORP, differently from the c172p. c182s: empty CoG 40.31 in, AERORP 37.37 in. c172p (160 hp): empty CoG 39.06 in, AERORP 40.6 in. Changing this in the c172p would probably allow to decrease the necessary elevator authority, but I don't see an objective reason to change these values. I'm fiddling to make the elevator softer by other means, not sure....

Your aircraft is pleasant and stable, at least with the default load. However, it begins to be slightly unstable when loaded such as the CoG is at 45 inches (which might be normal).

After more observation, I think I might be able to decrease a bit the aileron authority (it's still at Megginson's values). But the rudder and elevator authority are needed for reaching stall AoA and, at 15 kt crosswind landing, getting aircraft axis alignment with the runway .

Useless to say that the main FDM coefficients initially come from Megginson's table (original author and pilot). And I made my "stall and spin" model trying to respect as much as possible these values (I'd say more: I initially kept all of them). Since that, I had to change a few of them, but I don't like this kind of pragmatism without understanding. @okroth also helped for this: Pitch axis too sensitive #1047

[Octal450]

Aren't you exaggerating? Qualitativeley and subjectively (not easy...),I don't find the behavior is that far from the C172's. Please, can you explain with more details the main differences between both, and issues, from your point of view?

No, its the little things that change the flyability as your model is not "totally wrong". Its good but has a few issues. Its been a few months now since I've tested this so you'll need to bare with me until I can find time to re-test.

As I recall it the 2 biggest problems were:

• Far too much adverse yaw. Flying the aircraft without a rudder was nearly impossible (and I am comparing to the "jammed rudder" training I did in the real world, the aircraft is still flyable even though a little less coordinated) I found the c182 (and 172) were extremely difficult to almost impossible in some weather conditions to land properly without the nose swinging all over the place.
• Inability to correct for roll and yaw through the rudder alone (in the real plane, ailerons can be pretty much neutral with the rudder use to correct)

Yes the elevator is too sensitive I agree but that problem is not the biggest problems.

The adverse yaw is inverted.

I didn't notice this, but perhaps something changed since I tested. Adverse yaw should definitely be there (I'm all for realism), but as of now there is WAY to much.

I don't intend to do a broad refactoring

Well I am looking at the stance of this from being a realistic sim. If you aren't willing to consider changes based on feedback from multiple people who have flown the REAL aircraft, then I don't know what to tell you.

Kind Regards, Josh

[wlbragg]

Inability to correct for roll and yaw through the rudder alone (in the real plane, ailerons can be pretty much neutral with the rudder use to correct)

I actually tried this in the sim a long while back when the subject first came up. I didn't find it impossible, in fact I was able to complete a circuit and land. Although it was quite difficult. But I felt the difficulty was in power management and lack of experience.

Well I am looking at the stance of this from being a realistic sim. If you aren't willing to consider changes based on feedback from multiple people who have flown the REAL aircraft, then I don't know what to tell you.

I've been working with @dany93 for quite some time and I think all he is saying in regards to "refactoring" is he is confident enough in his thorough past review and collaboration that it isn't needed or appropriate to make "major" changes. We have heard many real life pilots make their comments about this FDM and none of them ever seem to agree on everything. I know @dany93 will listen to any actionable suggestions backed by data and make improvements to this aircraft if appropriate. Please don't forget he has a lot of skin in the game with this aircraft. He is the lead on the FDM and I support his reasoning. Furthermore, every ounce of data and programming that has gone into this aircraft since I started working on it has been under the doctrine of "realism first". Including the existing FDM.

[wlbragg]

I don't want to loose the valuable critique in the fray so I think we need to concentrate on each item, line by line and come to some sort of resolution.

Below 90 knots, it is extremely forgiving, you can make huge inputs and the thing barely responds.

Is this something that can be softened in the JS settings. I know in MS2020 there is some outstanding JS tuning algorithms. (I cant think of the right word for it)? Keeping in mind for the crosswind stuff the max inputs are needed.

A 60 knots climb, it's extremely sharp, and your nose will be pointing to the sky. A shallow climb will be at 85 knots.

Have we tested against this statement?

The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME. In fact, I didn't even notice that in real life.

We have discussed this at nauseum. Part of the miss perception in my opinion is rudder trim. This has been discussed as well. We have added a convenient way to test that.

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Always was my thought that I was way too low, but I had no real life experience. So the solution was to add persistent adjustable seats. The highest settings lets you "see the full nose of the aircraft."

Tyres don't squeak like that, like someone is raping a chicken.

I discussed this and it is a fact if you lock up your brakes on any moving vehicle you will screech your tires. The issue is that the digital V analog aspects of the brake make it lock up far to easily or quickly. I think there is room for programming some kind of filter or code to deal with this and make it still possible but not so easily. There are things we need to know however that may prove difficult to ascertain. Such as at 30 knots does setting the parking brake or slamming on the brakes lock the wheels? The sound you hear in sim is only there if the aircraft wheels are touching the ground while the aircraft is moving and they are not turning. The cause and effect are totally realistic, but may lack a proper implementation.

It accelerates a lot quicker than it does in FG and stalls at 45 knots.

That is something to look at in my opinion. We need hard numbers though, not perceptions for the acceleration. And for the stall, I thought we were in POH numbers.

The controls are a lot less sensitive overall. I strongly advise to have a look at Octals/Joshs PA-28 on how to implement proper inputs from the joystick/yoke.

Again, "I know in MS2020 there is some outstanding JS tuning algorithms.", this might apply here. Also, lose the sensitivity and maybe give up something else like the crosswind stuff? Is it FDM or JS and control inputs. Where should we tune this? For sure lets look at the PA-28, or better yet please explain the difference! Is the PA-28 capable of the same crosswinds we have tuned for.

Keep in mind when I talk about the FDM tuning and crosswind stuff I am not an expert in flight maneuvers or FDM so I may use the wrong language. I hope I at least get the point across that I am trying to make. That point being the settings that exist allow for certain flight maneuvers. loose those settings and do we loose the maneuvers? Does the PA-28 still retain the same maneuver capability?

[HHS81]

@dany93 @dany93 I didn´t change anything since we closed the issue. Are you sure you use the right version of the c182s?

I´m not a fan of creating fdms just by "feeling". Rather more of numbers for given control inputs. No one of us here has flight controls at home which simulates the forces created by the airflow, mechanic and everthing else involved, but is responsible for the "feel" of the aircraft. Or do you use hardware which has the same travel way like the real controls? When I started to develope aircraft I was told that the fdm should work and be correct in use with hardware which simulates all this. Travel way, forces etc...

"The adverse yaw is inverted. " With the first values the aircraft swung heavily from one side to another. I checked more then once several different values to decrease this unwanted effect but keeping the non-pedal use on turns. I just checked it and I didn´t updated my cessna 182s version since the last major update some months ago, and I´m sure I can see stil adverse yaw. It is much less pronounced- but it is there. At speeds above 100ktn adverse yaw get very slightly inverted- hardly noticeable. The question is how strong the advery yaw effect is on cruise speeds above 100ktn?

" Your aircraft is pleasant and stable, at least with the default load. However, it begins to be slightly unstable when loaded such as the CoG is at 45 inches (which might be normal). "

What do you mean with "unstable"? I didn`t see anything unstable.

[Octal450]

The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME. In fact, I didn't even notice that in real life.

Well it does roll left mostly. But all I need to stop that in the real plane under cruise power is just resting my foot on the right rudder, barely any force needed. Only during takeoffs and such did I need a little more pressure but usually not to the point of the FG's one. I'm comparing rough deflection amount.

Below 90 knots, it is extremely forgiving, you can make huge inputs and the thing barely responds.

Well.... I don't know about that. I'd say that behavior sounds more like on approach slow speed. That's when it get more looser. at 90kts there is some responsiveness and especially above that.

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Agreed. It was too low in the 182 also iirc but I always ctrl drag it up a bit.

Is this something that can be softened in the JS settings.

No. This is something that should be analyzed by rough deflections of the control surfaces versus aircraft responses. Everyone joystick difference you shouldn't point why an FDM is wrong or right thru joystick settings. This is something glaringly wrong with FSX default aircraft for example.

Is the PA-28 capable of the same crosswinds we have tuned for.

What did you tune for? I recall testing this quite a bit. I can do some more analysis.

The sound you hear in sim is only there if the aircraft wheels are touching the ground while the aircraft is moving and they are not turning. digital V analog aspects of the brake make it lock up far to easily or quickly

Digital input is not a reason, another thing bad from default FSX planes -> the wheels just "lock" when you push your breaks. JSBsim by default seems to have this fixed, I've never really had to play with brakes and my wheels don't lock. I just adjust the friction values to get the desired breaking effect. So yes this noise should not be there unless you implement mechanics for actually locking tires, and just using the digital input to do that is a horrible idea.... Nobody can brake the thing right unless they have expensive pedals then! There is no sense in that.

Given that JSBsim does that right, its something programmed in the c172 specifically that is causing that problem and the complaint is 100% valid.

Kind Regards, Josh

[HHS81]

Below 90 knots, it is extremely forgiving, you can make huge inputs and the thing barely responds.

Is this something that can be softened in the JS settings. I know in MS2020 there is some outstanding JS tuning algorithms. (I cant think of the right word for it)? Keeping in mind for the crosswind stuff the max inputs are needed.

Nope. The 15ktn crosswind demosntration is just for certification. The real aircraft is able to do much more crosswind (30ktn+)

A 60 knots climb, it's extremely sharp, and your nose will be pointing to the sky. A shallow climb will be at 85 knots.

Have we tested against this statement?

It is just Vx (best angle) and Vy (best climb)

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Always was my thought that I was way too low, but I had no real life experience. So the solution was to add persistent adjustable seats. The highest settings lets you "see the full nose of the aircraft."

We had the same discussion on the Cessna 182S- I found out how to solve. Put on a headtracker or use NoFacetrackIR- change the field of view to something about 94-104deg. Suddenly you will see the "full nose". It is a problem of the perception by not-correct simulated head pivot, which can´t be simulated by FlightGear. As shown here: https://www.flightsim.com/vbfs/showthread.php?315593-Flight-gear-18-3-2-initial-set-up&p=2075607#post2075607 and https://www.flightsim.com/vbfs/showthread.php?315593-Flight-gear-18-3-2-initial-set-up&p=2075663#post2075663

Tyres don't squeak like that, like someone is raping a chicken.

I discussed this and it is a fact if you lock up your brakes on any moving vehicle you will screech your tires. The issue is that the digital V analog aspects of the brake make it lock up far to easily or quickly. I think there is room for programming some kind of filter or code to deal with this and make it still possible but not so easily.

Use the one from Cessna 182S, it sound like the real ones.

The controls are a lot less sensitive overall. I strongly advise to have a look at Octals/Joshs PA-28 on how to implement proper inputs from the joystick/yoke.

I don`t see any difference in modelling the inputs. Just using other aerodynamic coeffecients.

[HHS81]

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Agreed. It was too low in the 182 also iirc but I always ctrl drag it up a bit.

Oh- you guys all have the same height? As I know the seat height can´t be increased in the cessna?

[tonghuix]

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Agreed. It was too low in the 182 also iirc but I always ctrl drag it up a bit.

Oh- you guys all have the same height? As I know the seat height can´t be increased in the cessna?

No, it is can be increased by a handle under the seat.

[HHS81]

Seating position is way higher in real life, in FG it seems like you are hiding behind the instruments panel, in real life you can see almost the full nose of the aircraft.

Agreed. It was too low in the 182 also iirc but I always ctrl drag it up a bit.

Oh- you guys all have the same height? As I know the seat height can´t be increased in the cessna?

No, it is can be increased by a handle under the seat.

Ah, o.k. Anyway you can do it in FGFS by yourself. I discussed this issue here: https://www.flightsim.com/vbfs/showthread.php?315593-Flight-gear-18-3-2-initial-set-up&p=2075607#post2075607 and what´s the real issue behind

[Octal450]

I don`t see any difference in modelling the inputs. Just using other aerodynamic coeffecients.

This. The only difference I have is I cancel potentiometer noise in all my aircraft with a lag filter.

Kind Regards, Josh

[HHS81]

Btw. to consider: @dany93 @wlbragg @Octal450

The adverse yaw is minized on most aircraft by using several design aspects: 1.) Differential aileron deflection 2.) Frise ailerons

Cessnna makes use of 1.), but not of 2.). But it should already minimize the Adverse Yaw effect.

[HHS81]

@wlbragg https://groups.google.com/g/sci.aeronautics/c/69ECYNDKIJA

According to a table in one of Roskam's books the following hold for
the Cessna 172:

Condition: cruise
Altitude: 5000 feet
Speed: 219 ft/sec
CG = 25% MAC
Sref = 174 feet^2
Span = 35.8 feet
MAC = 4.9 feet
GW = 2645 Lbs
Ixx = 948 slug-feet^2
Iyy = 1346 slug-feet^2
Izz = 1967 slug-feet^2
Ixz = 0 slug-feet^2
CL = .31
CD = .031
CT = .031

Longitudinal: Lateral:
Cmu = 0 Clbeta = -.089
Cmalpha = -.89 Clp = -.47
Cmalphadot = -5.2 Clr = .096
Cmq = -12.4 CldeltaAileron = .178
CmTu = 0 CldeltaRudder = .0147
CmTalpha = 0 Cnbeta = .065
CLu = 0 Cnp = -.03
CLalpha = 4.6 Cnr = -.099
CLalphadot = 1.7 CndeltaAileron = -.053
CLq = 3.9 Cndeltarudder = -.0657
CDalpha = .13 CYbeta = -.31
CDu = 0 Cyp = -.037
CTXu = -.093 Cyr = .21
CLdeltaElevator = .43 CydeltaAileron = 0
CDdeltaElevator = .06 CydeltaRudder = .187
CmdeltaElevator = -1.28

Only Cruise, but should give you an idea.

[HHS81]

@dany93 @Octal450 @wlbragg https://www.youtube.com/watch?v=ewHL8-agBWI might gives more idea how the real thing acts regarding adverse yaw and rudder input

[wlbragg]

OK, seat position as far as I am concerned is totally adjustable both as a seat adjustment or using sim controls so is not an issue. If you have advanced equipment then it sounds like it is easily adjustable to taste. I could ad FOV to the seat controls but that is not keeping with the reality of what a seat adjustment can do. The user can change the FOV as needed. So the seat view in a non-issue.

Now about something I feel qualified to talk about, tire screech. I understand what is wrong with the simulation of the tires and braking. What I meant by digital VS analog is that with a digital input control you get on or off with analog you can apply various degrees of force and the program code can be programmed to reflect that so you can hold the brakes down a portion of the way and not lock the brakes up. As it is programmed right now it is a timed braking (closer to simulating a digital response), hold it down long enough no matter the force and it will screech. I agree it needs work, But the fact remains, you don't get a screech until a wheel quits turning at the same rate that the aircraft is moving. It is correct. Don't lock the wheels up and it wont screech. If you remove the sound then you have no idea that you are actually sliding and not rolling, that is not realistic. One way, don't lock up the brakes, realistic. The other way, allow the brakes to lock and not provide any feedback that they are locked, not realistic. The reason I implemented this in the first place is because we have a ski variant, 2 different bush variants, pontoons, amphibious and the default tire variants. and also logic to detect surfaces including ice, snow depth and soft and hard ground. We have variants that require that ground detection logic and work appropriately. I am proud and please with that system and the flexibility and realism it provides. In my world of flying it is as much about the landing surfaces as the flying. So someday I do plan to look at the brake code and see if I can't come up with a better way, my brake inputs are analog as far as I am aware and I plan to tune a set of brakes that can take advantage of that type of controller. As far as the sound file, by all means, if we have a better GPL sound, I am all for it. Also it is easy enough to turn that one sound off in the sound code file if t really bothers someone that has no need for that feedback. We could add a GUI option to toggle it on/off, but we have many who resist clutter in the GUI. Maybe a quick key to turn it off? But I wont take up my time to research the availability of a key for something I think will eventually be removed once the brake logic has been modified.

As for the FDM stuff, it all over my pay grade and I have to defer to @dany93 I merely brought up a couple reasons that it is difficult to agree on and quantify what should be. Unfortunately I relied on my limited vocabulary of this subject to try to explain my understanding of why the controls are tuned the way they are. Specifically "max inputs are needed", not the right language. I meant tuned the way it is in order to be able to perform certain flight maneuvers.

I don't know if anyone caught this but we already have a disagreement between to people that have flown the c172p in real life, about one of the issues that is front an center.

The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME. In fact, I didn't even notice that in real life.

VS

Well it does roll left mostly. But all I need to stop that in the real plane under cruise power is just resting my foot on the right rudder, barely any force needed. Only during takeoffs and such did I need a little more pressure but usually not to the point of the FG's one. I'm comparing rough deflection amount.

[wlbragg]

@HHS81 nice videos, really punches it home.

[HHS81]

@wlbragg @dany93 @Octal450 @Megaf

The Cessna DOES NOT ROLL TO THE LEFT ALL THE FREAKING TIME. In fact, I didn't even notice that in real life.

VS

Well it does roll left mostly. But all I need to stop that in the real plane under cruise power is just resting my foot on the right rudder, barely any force needed. Only during takeoffs and such did I need a little more pressure but usually not to the point of the FG's one. I'm comparing rough deflection amount.

What we can see here might can be explained:

1.) FGFS has only one pilot on the left seat per default. Question: megaf flew alone? or with an instructor? 2.) just resting the foot on the right pedal stops left roll tendency- it is barely noticeable according Octal450- so megaf might not noticed that

That´s the problem. Subjective views where we need objective numbers