Fixed Wing Flight Model (dev branch)

[oukej 2910]

I've done an inverted merge of this topic with phaorxs post by mistake. Sorry about that. Hopefully I haven't broken any links or anything.

[xxgetbuck123 945]

6 hours ago, dragon01 said:

Check out Caesar BTT from Apex (the racing version, especially). The DLC jets don't have the flight model fully configured, Caesar uses a number of config parameters that might give a more realistic turning behavior.

 

Yea the Caesar BTT definitely has a different flight model to the rest. Notably when you pull hard on the stick then let go the aircraft tries to go back to a natural position, unlike the jets. Also when you pull too hard on the stick you can feel the aircraft struggling to make the turn, which you cant feel in the jet. 

[venthorror 117]

6 hours ago, xxgetbuck123 said:

Notably when you pull hard on the stick then let go the aircraft tries to go back to a natural position

Exactly, It is trying to go back to its flight path vector and corrects it self to the relative wind. Had to test for my self to actually believe it.

Whether its real, or just a smart trick to make it behave like it should really does a great job.

 

Same goes for rudder. Stepping hard on the rudder and releasing makes that (flight sim like) sway left and right effect.

[pharohx 10]

Quote

 

OK, I will give that a look.

 

Thanks to all for your replies.

 

Was hoping though to have it corrected for all as default, BIS build, not as a MOD. Glad to know it's not a complete dead end though.!

 

Still concerned about the NWS system for navigating while on the deck....

Any tips on that?

[xon2 102]

I still have the problem that my Thrustmaster TWCS throttle responds when bound to the analogue thrust increase only past 50%...so i have only the upper half of the throttle range to use. This throttle axis is not bound to the air brake nor are the normal thrust increase/decrease keys bound to anything. The helo collective works without issues and gives me the full range. Is that an aknowledged bug which will be fixed or is it just me having this issue?

[oukej 2910]

43 minutes ago, xon2 said:

I still have the problem that my Thrustmaster TWCS throttle responds when bound to the analogue thrust increase only past 50%...so i have only the upper half of the throttle range to use. This throttle axis is not bound to the air brake nor are the normal thrust increase/decrease keys bound to anything. The helo collective works without issues and gives me the full range. Is that an aknowledged bug which will be fixed or is it just me having this issue?

You have to bind both halves of the axis to the action. We're sorry about this inconsistency and caused confusion. Newly this requirement should be hinted in the action tooltip.

[xon2 102]

haha nice. Now it works, taxing has become a joy with this. Never noticed this tooltip before, THX.

[anilbu 3]

I'm no expert in avitation/flight physics but, whatever you guys have done to older planes especially civilian plane and the buzzard, it definitely feels better. Is it planning to be done for jet planes in the DLC?

[dragon01 902]

@oukej, I found what is, perhaps, a root cause, or at least an important factor in the flight model feeling "off". Sadly, it's present even on Apex aircraft. I took a while to note the FPM position relative to the "wings level" marker and the wings' position with regards to horizon, and found that all planes magically generate significant lift at negative angle of attack. It's easy to see, just level off and observe the position of the FPM. In almost cases I tested, it was above the "wings" marker, where it should be a good bit below (and when it was not the case, it was right on top the "W"). This is a very incorrect behavior - wings' lift is a function of AoA, when it goes negative, so goes the lift. Most fighters have symmetrical airfoils, and even in case of planes where it's not the case (that is, on the Blackfish), camber and angle of incidence would not be enough to put the FPM above the "W", as is currently the case. Even the Xi'an is not exempt - at speed, Conada effect would almost completely nullify the fan thrust (and besides, to maintain stability, the fans would have to be disconnected when the engines are up).

 

On a side note, I noticed that both Xi'an and Blackfish have a really good turn rate when banked, about 2 degrees per second when banked to the "big notch" on the roll indicator (or to a point where it should be, in Blackfish). It should be like that on all aircraft.

[x3kj 1247]

On 30.4.2017 at 11:45 AM, dragon01 said:

I took a while to note the FPM position relative to the "wings level" marker and the wings' position with regards to horizon, and found that all planes magically generate significant lift at negative angle of attack. It's easy to see, just level off and observe the position of the FPM. In almost cases I tested, it was above the "wings" marker, where it should be a good bit below (and when it was not the case, it was right on top the "W"). This is a very incorrect behavior - wings' lift is a function of AoA, when it goes negative, so goes the lift. Most fighters have symmetrical airfoils, and even in case of planes where it's not the case (that is, on the Blackfish), camber and angle of incidence would not be enough to put the FPM above the "W", as is currently the case.

Many A3 planes seem to have 3° angle of incidence. They fly perfectly straight when oriented to -3° (csat vtol for example)

 

You wonder whats wrong with the FM? Look at this.

Plane is rolled to the side and has almost no airspeed -> falls sideways. Sidewaysacceleration shows 8.5-9 m/s². Sideways velocity barely increases. (and yes Gryphon can fly also backwards currently)

Edited May 13, 2017 by x3kj

[dragon01 902]

That was supposed to be fixed by Apex improvements, especially the "flying backwards" thing. It boggles the mind that we're so close to release and the DLC jets still aren't fully configured (either that or they aren't configured very well). 

 

Yeah, the artificial "angle of incidence" seems to be the case. I suppose it would make sense for Blackfish (which does seem to have some AOI), but the others have wings level with the ground. If BIS could remove this behavior, it'd be a noticeable improvement.

[x3kj 1247]

Quote

Added: Air friction parameters for planes (gearsUpFrictionCoef, airBrakeFrictionCoef, airFrictionCoefs2, airFrictionCoefs1, airFrictionCoefs0) 

I'm most curious about more details on this

[Beagle 684]

The DLC is out but still the FM is not only arcadish it is worse than arcade ... it is plain wrong.

You can still bank all jets to 110° without any change in direction or even a drop of the nose, It loooks like the lift vector of wings is not modelled at all. Another evidence for that is the fact that even with flaps down the planes do not takle off by semselfs at any speed...you always have ti yank on the stick for any change in attitude or direction.

 

When even such basic manouvers like a standard turn or canopy rol are not possible, its not even arcade FM but NO FM. at all.

 

https://en.wikipedia.org/wiki/Standard_rate_turn

 

https://en.wikipedia.org/wiki/Barrel_roll

[oukej 2910]

We are really sorry. We didn't make it all in time :(

[dragon01 902]

Think you can make it for the post-Jets patch?

[Beagle 684]

1 hour ago, oukej said:

We are really sorry. We didn't make it all in time :(

Does that mean you're at least considering a more decent FM?

That would be fine, after all im used to waiting.

[a_killer_wombat 120]

After reading up on how the AOA Indexer is supposed to work, which is used in the real-life F18 Hornet as well as Arma's fictional Black Wasp, I believe the way it's been implemented in Arma is incorrect. 

 

 

As seen here, the red light is supposed to indicate that you are currently too fast so you need to decrease throttle and pitch nose up while the green light is supposed to indicate that you are currently too slow so you need increase throttle and nose down. The amber/yellow light is supposed to indicate that you currently have the correct speed and nose pitch (AOA).

 

However, from my experience, the way the AOA Indexer is currently portrayed in Arma seems to indicate the exact opposite of the above, where red light is lit up when I'm too slow and green light is lit up when I'm too fast.

 

I'm no expert on this so if anyone else has tested this or has more knowledge on the real-life counter-part, feel free to correct me.

[CaptainAzimuth 714]

26 minutes ago, a_killer_wombat said:

Snip

I think that belongs in the Sensor thread. 

[dragon01 902]

Why? AoA indicator has nothing to do with sensors and everything with FM.

[x3kj 1247]

with a FM that is as borked as it is, that AoA not beeing precise would be the least of my concerns. Especially since the optimal landing parameters need to be known for this to be of any value.

[MK84 42]

3 hours ago, a_killer_wombat said:

After reading up on how the AOA Indexer is supposed to work, which is used in the real-life F18 Hornet as well as Arma's fictional Black Wasp, I believe the way it's been implemented in Arma is incorrect. 

 

 

As seen here, the red light is supposed to indicate that you are currently too fast so you need to decrease throttle and pitch nose up while the green light is supposed to indicate that you are currently too slow so you need increase throttle and nose down. The amber/yellow light is supposed to indicate that you currently have the correct speed and nose pitch (AOA).

 

However, from my experience, the way the AOA Indexer is currently portrayed in Arma seems to indicate the exact opposite of the above, where red light is lit up when I'm too slow and green light is lit up when I'm too fast.

 

I'm no expert on this so if anyone else has tested this or has more knowledge on the real-life counter-part, feel free to correct me.

Yup it is reversed.

[a_killer_wombat 120]

1 hour ago, x3kj said:

with a FM that is as borked as it is, that AoA not beeing precise would be the least of my concerns. Especially since the optimal landing parameters need to be known for this to be of any value.

 

With a working landing approach HUD indicator, an AOA indexer which is working (except for the fact the lights seem to be reversed), and a landing autopilot, I would say it's fair to say that the optimal landing parameters are known. While there are obviously more glaring issues with the flight model which are of a higher priority, it's still important to notify the minor issues as well. Especially when the minor issues typically require far less time and resources to fix than the major ones.

[dragon01 902]

The problem is, ArmA planes have three degress of "fake" AoI, which in turn means that AoA is positive only at extremely low speeds. You can actually see planes fly at negative AoA because of this. In general, I think that this needs to be fixed before relevant indicators are adjusted.

[Strike_NOR 898]

12 hours ago, dragon01 said:

The problem is...

 

From my point of view it seems that a lot of the arma fixed wing flight models use "shortcuts" to mimic real flight behavior. Sometimes I feel like I am flying an ArmA 3 missile. It feels like the aircraft do not have weight or inertia, and there's absolutely no aerodynamic stall. The stall appears to simply "lock controls and lose altitude" in ArmA. 

 

My other major gripe is the way thrust is handled. A jet or propeller aircraft will always produce thrust as long as the engine is running, in fact, it produces more the slower you are going because the speed of the air pushed behind the aircraft is much faster than the aircraft speed (a plane in level flight can't make itself go faster than the air it is pushing backwards). However, I often get a super-weird phenomena in ArmA where I can enter stall speed (red airspeed text) with a slight nose up and even at full throttle, it just hangs there. Like magic. Jet engines on fighters may easily produce 7 tons of thrust without AfterBurner, maybe 12+ tons with AfterBurner. Even if the aircraft was "hanging" in the air or falling straight down like a brick, it should create a forward thrust vector and start accelerating. This actual thrust vector does not appear to exist in ArmA. Instead, it seems that engine power just directly influences airspeed, and is prohibited when the aircraft has entered the "stall" area. So you are absolutely screwed if you enter this "stall area" and can't get your nose down.

 

I would give an arm and a leg to see a flight model similar to the first IL-2 Sturmovik game. It had good stall mechanics and also some decent low-speed flight modeling. Also, I'm not sure if it is modeled yet, but I think wind and gusts should effect aircraft in ArmA.

 

A final comment that I'd like to make is that if we assume most jets in ArmA have flight control computers, then many of the stall effects are reduced to a point where the aircraft will appear to just lose control authority. The computer prevents events such as asymmetric stalls and flatspins, but it may induce a form of "rocking stall" where the aircraft oscillates in pitch axis while falling straight down like a brick.

 

I hope that we can get some proper flight models into the game. Keep them super simple, but slightly more advanced than today's models. It needs to have a few vectors modeled, the minimum being "Lift, Weight, Drag and Thrust".

[oukej 2910]

2 hours ago, Strike_NOR said:

However, I often get a super-weird phenomena in ArmA where I can enter stall speed (red airspeed text) with a slight nose up and even at full throttle, it just hangs there. Like magic.

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

2 hours ago, Strike_NOR said:

I would give an arm and a leg to see a flight model similar to the first IL-2 Sturmovik game.

Doesn't sound modest at all :O ;)