Fixed Wing Flight Model (dev branch)

[dragon01 902]

I don't think that IL-2 style FM is needed. However, a few things should definitely be added for it to feel "right":
1. Proper AoA. At landing speeds, it's usually about 8 degrees (see the pic above). Less if you're going fast, but it never goes into negative in level flight.

2. Turning behavior, especially when banked. It improved a lot lately, but the actual turns are still too slow.

3. Stall behavior. While autopilot and other aids can help it, it could be a bit more refined. This mostly applies to DLC jets and VTOLs (to a lesser extent), which should have significantly stall and post-stall characteristics from the other planes.

[Strike_NOR 898]

2 hours ago, oukej said:

This can be fixed by thrust configuration and we're just doing that atm :)
 

Doesn't sound modest at all :O ;)

 

Hey! Awesome news Oukej! I can't wait to see how jets act with a proper thrust configuration :)

 

Oh maybe I'll throw in a kidney too for good measure ;)

[CaptainAzimuth 714]

Flew the Xian recently just to get a feel for the flight models better banking and it's something the jets are missing. Feels better, but it would need to be tweaked right to feel good on the agile fighters.

[teabagginpeople 398]

On 02/06/2017 at 11:13 AM, oukej said:

This can be fixed by thrust configuration and we're just doing that atm :)
 

Doesn't sound modest at all :O ;)

Awesome news. Got my ass handed to me by a ka because I was in stall limbo.  I wasn't even pushing it hard. 

 

From the ka view it may have seemed I was hanging there on purpose saying "come on that all you got". When in reality I was full thrust saying to my jet "don't you do this to me you bitch". Finger hovering over eject. 

[slowrider8 41]

It appears that the Buzzard has had a Flight Model Overhaul on the Dev Branch. It feels a lot harder to stall now which is nice and it seems that the wings actually produce lift now and you can't fly completely inverted without being pulled down. Also seems like it has some sort of auto trim that forces the plane to fly level, doesn't feel exactly great to be fighting with the plane constantly so that might have to be toned down a bit. Can we assume that the Buzzard is a testbed for these new changes and we will see them spread to the other planes later on?

[oukej 2910]

54 minutes ago, slowrider8 said:

Can we assume that the Buzzard is a testbed for these new changes and we will see them spread to the other planes later on?

Exactly that :)

[Imperator[TFD] 444]

Loving how the Buzzard now noses down when you roll.  Agree with Slowrider8 though on the auto-level, feels a bit janky and too sudden.  Looking forward to more developments.

[oukej 2910]

Is what you're talking about the effect of the airplane aligning back from sideslip?
As for auto-trim - we don't have such feature but we're trying to hack adjust the forces in a way that the airplane stays leveled throughout most of it's operational speed range and it doesn't pitch up in higher speeds as much and require trimming. Not ideal but necessary for controlability via kb+m.

[dragon01 902]

I noticed the "auto-level" effect, but it's not always present. It doesn't seem drop the nose when flying upright at high speed, but something like that definitely happens at low speeds (though it's only the nose dropping, it doesn't rise when in a dive). There is definitely something odd going on with that. Also, the "fake AoI" isn't gone, AoA behavior is still somewhat unrealistic, especially at high speeds. 

 

That said, turning and lift behavior is much improved. I even measured the turn rate with a watch and found it's pretty much correct. :) Stall behavior is also nice, very gentle, like I would expect from a modified trainer.

[a_killer_wombat 120]

Also feels like control surface compression is now felt (or at least amplified) when trying to manoeuvre at high air speeds which is an improvement. 

[slowrider8 41]

15 minutes ago, oukej said:

Is what you're talking about the effect of the airplane aligning back from sideslip?
As for auto-trim - we don't have such feature but we're trying to hack adjust the forces in a way that the airplane stays leveled throughout most of it's operational speed range and it doesn't pitch up in higher speeds as much and require trimming. Not ideal but necessary for controlability via kb+m.

I'm no expert on plane aerodynamics but what I'm referring to is how when you pitch either up or down if you don't hold it there by pushing the mouse continuously it will pitch back down or up so that the plane is level. It makes using the plane for CAS pretty difficult as you're constantly fighting the plane just to keep the nose down for an attack run. I suppose this might be realistic or necessary but if the effect could be toned down a bit it would be appreciated.

[KiTooN 63]

I noticed something weird since starting flying jet again, banking is waaaay less effective to turn now. Am I the only one ?

Here's a vid, a whole 3-minutes for a 360 turn at 45° :

 

Here's a steep bank in a real plane :

 

Slowing down doesn't help, the only influence you can have on your bank is using the rudder.

Is this intentional ? Looks like a bug to me (or poorly configured asset maybe).

[Sniperwolf572 758]

3 hours ago, slowrider8 said:

I'm no expert on plane aerodynamics but what I'm referring to is how when you pitch either up or down if you don't hold it there by pushing the mouse continuously it will pitch back down or up so that the plane is level. It makes using the plane for CAS pretty difficult as you're constantly fighting the plane just to keep the nose down for an attack run. I suppose this might be realistic or necessary but if the effect could be toned down a bit it would be appreciated.

 

As another KB+M user with no clue about aerodynamics, this "correcting" when you tap pitch up or down feels horrible when trying to aim at any weapon that needs aiming.

 

And if I hold the pitch up or down, the plane will start pitching up or down, start feeling like the pitch rate is progressively being slowed down by a rubber band, the rubber band then snaps and pitch resumes at full speed again, until I release the pitch key, then the process repeats itself. This happens if you towards any direction, vertically, horizontally or diagonally. It feels as if I'm fighting against the auto-pitch and it's trying to correct me, then gives up and lets me do whatever I want as long as I don't release the button.

 

If there is a system that I need to keep me level, I'd rather turn it on myself when I need it (rest, targeting pod usage) rather than fight against it when it's not helpful (at all other times).

[x3kj 1247]

5 hours ago, KiTooN said:

I noticed something weird since starting flying jet again, banking is waaaay less effective to turn now. Am I the only one ?

Certainly not, other people and myself have mentioned multiple times in the past. The truth is that the FM in Arma feels very lackluster, because apart from flying straight, pitching up with elevators and rolling, it does not behave at all like a normal plane would. There must be something seriously wrong or missing in the flightmechanic, otherwise tweaking config values would be able to fix most of this. Without access to the source we will never be able to tell what that issue is...

 

I played around with the addforce commands on planes and by doing that, the hack nature of the FM really comes to light. If you catapult a plane high up into the air (from normal parking position), it will fall down slowly like a feather. This is what i also observed when "stopping" a jet midair and have it falling sideways (video some pages back). I have mentioned the same case many many months back as well (and was "rejected" because my custom plane i used to demonstrate it was vtol... skip forward to now. We have vtol, we have jet dlc, same exact problem remains.)

Not only are the planes affected, but their wrecks as well. And this looks especially silly. Shooting down an incoming plane while it dives on you for a gun run.  Wreck crashes into the ground, bounces off like a rubber ball, and falls back down slowly like a feather. Ridiculous.

 

At least we can take matters into our own hands now...

https://www.youtube.com/watch?v=Ebp8-MUwRbI

 

[venthorror 117]

On 6/5/2017 at 4:19 PM, slowrider8 said:

It appears that the Buzzard has had a Flight Model Overhaul on the Dev Branch......Can we assume that the Buzzard is a testbed for these new changes and we will see them spread to the other planes later on?

 

23 hours ago, oukej said:

Exactly that :)

 

Again, I had to test it to actually believe it. It feels so much better. And (just as the overhauled FM on Caesar BTT) stepping hard on rudder and releasing it does that (flight sim like) sway left and right effect.

 

Does this mean that Buzzard will get some more (much needed) love?

Sorry for going off topic a bit, but it does seem buzzard never had serious attention.

[dragon01 902]

On 6/5/2017 at 7:31 PM, KiTooN said:

I noticed something weird since starting flying jet again, banking is waaaay less effective to turn now. Am I the only one ?

Here's a vid, a whole 3-minutes for a 360 turn at 45° :

You're flying the Wipeout, the Buzzard is the one with updated FM. Wipeout still has the "Apex" FM.

 

As for the old FM, bank used not to cause turning at all. So even this is an improvement.

[CaptainAzimuth 714]

As of the recent dev update for the Buzzard... Feels Good. Feels Real Good. Only concern for me is the super sensitivity at low airspeed. Other than that, it feels great. Can't wait to feel this update on a much more powerful jet.

 

Landings finally felt good with the rudder issue solved, just the sensitivity of pitching when at slow speeds is a concern.

 

Edit: After much testing, this is exactly what the doctor ordered.

[dragon01 902]

Your enthusiasm is permature. It does feel a lot better, but it's far from done. Both Buzzard and Shikra display the same, fundamental flaw: incorrect AoA behavior.

This is an airspeed vs. AoA graph, for an RV-6 general aviation aircraft. It would have a different speed scale for ArmA planes, but the physics are the same. Note how AoA changes with airspeed.

 

Not so in ArmA. There (tested on Shikra), it shoots up rapidly form about 15 degrees at around stall speed to 0 degrees (on other planes, into negatives!) at about 200km/h above stall speed, then stays there for the rest of the flight regime. This is blatantly unrealistic.

[x3kj 1247]

speed- aoa-curve is not something you can define as a parameter. What this graph is representing is "required AoA" to keep a given heigth, depending on speed.  What actually needs to be changed to achieve a change in speed-aoa curve is the Speed-Lift-curve for given AoA's. More speed, more lift (generally, within normal flight parameters) at a given AoA. Means less AoA required to keep same heigth...

[dragon01 902]

Yes, but in absence of weight changes, AoA is directly proportional to lift. Besides banking, I don't think there's any other way of affecting lift in ArmA3.

[Grumpy Old Man 3545]

Any chance to remove or disable this newly added anti crash prevention when opening a dialog or going into the menu?

This stuff also makes the plane move on the ground with engines on.

After all the authenticity arma is striving for it's basically a punch to the face that the aircraft auto levels when mid-flight and opening a menu/dialog...

 

Cheers

[oukej 2910]

4 hours ago, x3kj said:

Speed-Lift-curve for given AoA's. More speed, more lift (generally, within normal flight parameters) at a given AoA.

That can be adjusted via envelope property.

[x3kj 1247]

Just now, oukej said:

That can be adjusted via envelope property.

i know, i was just "translating".

[dragon01 902]

Looks like that's what needs adjusting, then. Ideally, it would include the transsonic spike (both in lift and drag). If that's how it's configured, you'll want the curve to resemble this graph:
http://www.feretich.com/rocketry/TotallyStable/Images/LiftCoef2.pdf

This is probably the point where the problem is. Looking at the examples (dunno how old they are), they're completely out of whack.

[gatordev 219]

Something else to keep in mind with AoA indicators is that they're actually not indicating degrees of angles.  Most often, they're indicating "units" of AoA.  That graph that dragon01 posted seems to be saying that in some sort by indicating "normalized AoA," but it's important to understand that in different aircraft (or different aircraft publications), AoA doesn't always mean the same thing.