My impressions with the CAP 10
I noticed that the couples were not really modeled. With the CAP 10 propeller turning clockwise (seen from the pilot seat) in real life we would have
the effect of the helical breath: on takeoff the aircraft should board on the left. In flight at low speed and high engine speed, without countering, the ball should start to the right (yaw to the left)
the effect due to the gyroscopic torque of the propeller: on take-off, going from the 3-point position to the 2-point position, the airplane should, in addition, yaw to the left. In flight during a reversal to the left, at the moment of the rotation, the airplane should have a pitching movement that we normally counter by putting the elevator forward (and also the warping to the right to counter the induced yaw ). In reversal to the right it is the opposite!
These couples can explain unfortunate runway excursions on take-off of more powerful single-engine propeller planes and conventional gear. The tricyle train is a great invention.
During stalls and spins the steering does not seem really efficient!
In addition, when the plane is vertical, the dashboard is so dark that you can’t read the joke (I couldn’t find the lighting).
Having a few dozen hours of aerobatics on CAP 10 and a lot more on other planes, I find it much easier in real life.
Sensations and peripheral vision help!
For me, the aerobatics airplane flight models of Extra, Pitts and Cap10 are just plain wrong. There are limitations of the simulator and the Extra flight performance does IMO exceed the simulator capabilities, but still you can make it as realistic as FS2020 can do.
I am an armchair pilot - this is bad. But I know how to change the FS2020 flight model - this is good. REAL pilots with aerobatics experience please check my aeobatics airplane modifications at MSFS 2020 sailplane and give me feedback.
For Paul Falke
I downloaded and tried a bit. your modified CAP 10
I find it very unstable on the ailerons (excessive roll rate with very little stick steering) it cannot be controlled if you do not put a large dead zone in the sensitivity of the joystick.
Strangely enough, if we move the ailerons quickly the plane does not move and then leaves on the other side.
I tried the spins: you can’t do them using the rudder alone. But the spin starts well if you use the ailerons.
I also tried with the “historical” flight model. There it goes better with the ailerons, on the other hand, the deflections at the elevator (pitching) its much too weak: impossible to make a back ascent!
I check. Maybe I can only select “fast roll, but unstable” or “slow roll, but stable”.
I check. But maybe there is a bug in the “numerical integration beast” that I can not change by configuration files editing. Then you can only write a good report: What is wrong, how to test it, what you expect.
I can set a stall speed, but nothing else for stalls and spins. I see that the stall and spin improves, but maybe in this area FS2020 will never ever be very realistic. I think they are on the way to “a little realistic”.
There is - sadly - no future in “classic” flight model. The bad news: Both models use the same configuration parameters, but they have different interpretation. Say, in the one model I need the value 2.2 for a specific behavior, in the other flight model I need a 3.0 for the same behavior.
With the large number of controllers, and everyone setting their controller sensitivities differently, I don’t expect accurate aerodynamic modeling, which is impossible to model anyway, especially on a simple home PC.
I’d like to see better overall dynamics, and controller profiles that match specific controllers to specific aircraft. Only then will we have a system that comes closer to matching MSFS to real-world dynamics. Unfortunately, I don’t ever expect to see this in a hundred dollar video game.
If we are trying to create a simulation here and not just a game, would it not make more sense to concentrate on developing solid aerodynamics and a flight model that performs as close to reality as the “simple home PC” will allow, using baseline control settings? The various control configurations out there all produce the same analog min/max. Windows sees to that. The only difference from the very best to the old cheap (like my MS Sidewinder Precision Pro) is the resolution and stability. The plane will fly exactly the same for the guy with $3000 worth of controls as me. I would like to have an airplane that performs as accurately as possible. Non-pilots that want to soften the controls to make it easier can do as they wish, but for me, let’s push the envelope on what this simple pc can actually “SIMULATE”.
The reason was that I used the “classic” flight model. As told before, the same value has different effect in “modern” flight model to classic. I “tuned” the Cap10 now to modern flight model.
Please check again. I tuned a confguration parameter, but if it is a bug in the formulas, this did not help.
The rudder has now much more effect. Please check again.
Summary: This process needs several iterations. Are you with me? And can you please tell me how you do your things? Like how you use rudder to start spin.
The new version 1.1 Cap10 X is on my website. The same download button will give you the new version. Please overwrite the old version.
In roll it is now more realistic, and I can now set the joystick dead zone to 1.
the spins do not start by engaging the joystick fully backward and at the same time engaging the rudder fully.
For the spin to start on the right side, you have to engage the rudder first and then the joystick fully astern.
The rudder is certainly more efficient but the effect produced should be a yaw movement which increases the lift of the wing on the outboard side and makes the inner wing stall.
if we move the ailerons quickly and fully, the reaction is now normal.
in final, to land with reduced engine, the speed decreases too slowly (it looks like a competition glider)!
On a single-engine airplane with propeller and tail wheel, it is certainly very difficult to model: there is a lot of asymmetric forces which vary continuously according to the speed of the airplane and the engine speed (power). This is therefore particularly sensitive to take-off and stall turn.
I agree, even with power off during the entire final, speed on point and exact height over the THR for the flare, the thing floats for minutes eating up half of the runway. Something’s wrong with lift/drag
Okay, this demands changes in parameters cruise_lift_scalar, parasite_drag_scalar, induced_drag_scalar. Asobo did set them to default values of 1 and I did not change them. But, yes, even I did wonder about this good glide behavior.
I will call the new mod version 1.1.1, because changing the lift and drag is only a little change. By the way: Do you think that there is a polar curve for the Cap10?
As a first guess I will copy the Cessna C152 lift and drag values.
I like to make the flight model as realistic as possible. Can you please tell me in detail what you have to do in real live to make these stalls to happen?
You obviously are way deeper into flight models etc. Please explain me, if I watch the “feature discovery series” about aerodynamics I get the impression that the flight model is physics based. But you are talking about lines in the aircraft .cfg file. If the flight model is really physics based I would think it’s not possible to alter the flight model with a few lines in the .cfg file except if it comes to flight control deflections, aerofoil shapes etc.
The environment is physics based (athmosphere, wind movement, temp, density, humidity, etc).
However each aircraft is entirely different with different aerodynamic profiles and shapes, hence each one interacts differently with the environment.
Not saying this works entirely realistically in MSFS (because I don’t know), but how an aircraft flies will always be based on both environment and aircraft characteristics
As I understand X-plane for example predicts aircraft characteristics and behaviour based on the actual model and not by lines of code in some cfg file. I was getting the impression FS2020 works the same way when watching the aerodynamics video…
