Hi,
I love the new Dukes.
However I have issues with the slow propeller spool up of the engines on the turbine Duke during start up and the long delay during shut down after moving the condition levers to cutoff.
If the devs should read this forum, then please take another look at the above.
Thank you!
I suggest posting on the Just Flight forums. The developer responds there.
The exaggerated axis stabilities, make it too much “drive”, vs fly… almost a fly-by-wire feel… It really shows up when you try a textbook approach… power for vertical speed, pitch for airspeed. I made them close to what other aircraft of that size have.
I agree, curious how you made them “close” to other a/c. Which files and parameters were changed?
The “[FLIGHT TUNING]” paragraph, in the “flight_model.cfg” file…
These, for example, are the numbers for the default Baron:
pitch_stability = 0.125
roll_stability = 0.125
yaw_stability = 0.125
I changed it to “0.325” for the Duke.
Those parameters, and the MOIs (aircraft.cfg file), are where some developers get carried away, chasing the ever elusive, “realistic feel”… or attempting an unrealistic, “heft” to the feel.
With so many different control hardware setups, and different levels of ability (real piloting experience), it’s a tough thing for developers to accomplish. Usually ending up nearer an “easy” airplane to “drive” … vs … a more realistic flight-model to “fly”.
Do we know if their flight model has been verified by real Duke pilots, or if Black square has real Duke pilots on their team?
I don’t want to wade into those weeds… there are so many aspects to what would be “signed off on”, by a real pilot (it’s usually performance numbers, ala, cruise speed, climb rate, etc…)… but basic aerodynamics are universal.
Hi,
I only increased yaw stability and reduced rudder (yaw) effectivness.
Did not touch any other parameters.
The idea was to make the Duke more stable during taxi. Tried engine failures and Vmc to make sure that enough rudder authority was still present when flying on one engine. The airplane was still controllable on one engine down to the red line on the ASI, even with the critical engine failed.
BTW, I use HC Alpha yoke and Bravo throttles as well as MFG Crosswind pedals.
Don’t forget Blacksquare is only one man- Nick Cyganski
I would be amazed if he had a team of Duke pilots (not to mention that there really aren’t that many turbines in the world)
But what he has created with both the Blacksquare brand and the latest Dukes= mind blown 
Hi,
Just posted the same over there.
Thanks for the suggestion!
Shouldn’t the rudder be ineffective at taxi speeds?, i.e. to say that rudder effectiveness should not markedly influence taxi behavior. I found that the turbine Dukes (haven’t tried the other ones yet) is very hard keep on the centerline in crosswind takeoffs even in moderate (10-15kts) crosswinds, it weathervanes like crazy. Is that realistic?
An acid test… start with an airplane trimmed for level flight. If all you change is power, airspeed should remain constant (it will vary a bit but will return to equilibrium), and vertical speed will respond accordingly. Increase power, climb… decrease, descend… same airspeed.
This is very apparent, flying an approach. If axis stablities (especially pitch) (inflated MOIs can be even worse) are exaggerated, it throws the whole event out of whack … and you end up “pointing” or “driving” the airplane down the glide-path, vs flying it. It will also make trimming for a specific scenario (climb/descend/level) difficult, as the axis response is out of sync, if you will, with thrust/lift/drag.
For sure… and this airplane (piston, haven’t installed the turbine yet), is a stunning accomplishment.
As I indicated in other threads… flight modeling in MSFS, is a tall order… like several dogs chasing their, and each other’s tails… so many give-n-takes and compromises. I’ve spent days tweaking just the stuff mentioned here… because this model is so impressive, it’s worth the effort. And remember, the myriad of different control hardware… from cheap joysticks, to elaborate setups, complicates it all.
Hi,
Since the rudder is tied to the nose gear steering, it has a profound effect on how the airplane behaves during taxi. The rudder itself has very little influence at taxi speeds.
This should be a sticky, and it should be scripted to pop up every time we wade into the waters of discussing controllability. Without standardized, across the board force feedback, it’s nearly impossible to compare apples to apples.
For this very reason, I always say it flies like a plane, maybe not the plane. So as long as the performance numbers are close and there isn’t some crazy out of whack controllability/maneuverability issue, just take it as it comes, do the pilot thing, and fly the plane.
Is there any way for developers to interrogate what controller you have, and have custom settings that would better suit a given controller? Kind of like having many “if/then” statements in “flight_model.cfg”.
You might have the following files available:
flight_model_yoko.cfg
flight_model_xbox.cfg
flight_model_alpha.cfg
flight_model_fulcrum.cfg
And so on. Other code would then determine which of these files is the “flight_model.cfg” to use.
The Dukes definitely have more weathervaning behavior than any other plane I fly.
I had mentioned a few days ago that I just got Velocity One rudder pedals to replace my old Logitechs, and they sure make a difference for all phases. I was able to land in a decent crosswind last night with aileron into the wind with opposite rudder and keep it reasonably straight, though I need to practice that a lot more. I was all over the place with the old pedals, which had a twitchy potentiometer.
My opinion on the subject is that developers should develop their planes such that they fly as the real thing assuming that the user has full size actual controls, and then users with lesser controls can use various tricks up to them to make the plane suitable for them. Or, if possible, the developer can include a slider for users with controls with less travel, similar to what FlyInside does with their helicopters. But airplanes should never be dumbed done to please users with cheap controllers, in my opinion at least.
Standardizing to throw range is one necessary thing, but without standardizing the simulation of associated aerodynamic pressure on the controls (though various mechanisms), we’re still comparing apples and oranges.
Interesting discussion. How would they simulate aerodynamic forces on the controls (without a force feedback yoke)? I think they should simply simulate that the same deflection of the controls in the simulator has the same effect as the same amount of deflection in the real plane, assuming the user has full size controls. But if they used percentage of deflection thy could automatically accommodate all controller throw distances. But prolly it’s not as simple as that.
What I think distinguishes a good flight model from a bad one is that the good one accurately simulates the aerodynamic forces on the airframe. For example, one could investigate the amount of weathervaning if the controls are kept neutral as an aspect that can be studied without the apples and oranges comparison you mention.
Thanks! I need to experiment this!