Jump to content

Recommended Posts

That wouldn't make a difference really in gameplay... its only that the animation is spinning backwards that is find offputting. How the plane brakes midair technically is unimportant for gameplay, considering this is a very simplified FM.

Having afterburner (provide thrust boost with alot of fuel consumption increase) would be cool however.

Share this post


Link to post
Share on other sites

Fixed-Wing devs, I've noticed this thread has come back to life. Thank you to those that have brought it back to life. I fly a 30,000 lbs (13,607kg) military turboprop for a living that comes with all of the latest bells and whistles. One of the biggest issues I see with the current ingame planes is the glide ratio. If I turn the engines off or get hit by AA and loose the engines the planes feel like they're full of rocks and sink way too fast. Anyway to fix this so planes feel more like planes?

Share this post


Link to post
Share on other sites
Fixed-Wing devs, I've noticed this thread has come back to life. Thank you to those that have brought it back to life. I fly a 30,000 lbs (13,607kg) military turboprop for a living that comes with all of the latest bells and whistles. One of the biggest issues I see with the current ingame planes is the glide ratio. If I turn the engines off or get hit by AA and loose the engines the planes feel like they're full of rocks and sink way too fast. Anyway to fix this so planes feel more like planes?

The planes just don't seem to actually fly like planes. The A-164 is closer to a helicopter than anything else - you can turn sharper than the airspeed of the aircraft should allow. At incredibly low airspeed the plane can be whipped around like it has rocket boosters on its control surfaces.

Not only do they sink like rocks, they lose a shocking amount of airspeed, even in a gentle, coordinated turn. I am using rudder pedals and a stick and I still cannot get the aircraft to make a flat turn at a sustainable rate without losing all of its airspeed. In fact, turning around by doing a full-on Immelmann is somehow more efficient than a coordinated flat turn in terms of loss of airspeed, especially since the gain in altitude allows you to accelerate once level again.

Share this post


Link to post
Share on other sites
Not only do they sink like rocks, they lose a shocking amount of airspeed, even in a gentle, coordinated turn.

draconic force seems too low. I increased it alot for my testairplane and i was able to do 700kph and a super hard banking turn and stay basically the same speed with barely any thrust envelope change over vanilla opfor cas.

And because of the very low lift they fall like rocks. Not only that, they also nose down very hard on low speed - bringing up the question if that is even normal, because i would assume the aircraft to be balanced around it's lift-force-center (or however it's called) which should mean that the aircraft would just lose altitude but not nose down or up.

It would be nice if oukej could answer this question:

Where is lift force applied/determined in regards to the model position? I have the center of weight (geometry LOD) of my plane at the origin of the model itself.

Wouldn't an ideally balanced plane stay level no matter how much lift you have? Can the balance even be influenced by model settings or is this some hardcoded stuff?

_____

If I turn the engines off or get hit by AA and loose the engines the planes feel like they're full of rocks and sink way too fast

If you are hit in the plane and lose engines your plane is already counted as dead - it will explode at any random time, for no logical reason whatsoever. How hard can it be to create a second hitpoint class "HitEngine" and disable thrust if damage value is at 1 in the gamecode ? Helicopter Rotor blades falling off and Shaft breaking but no Damage model at all for planes. The inconsistency is just infuriating.

Edited by Fennek

Share this post


Link to post
Share on other sites
draconic force seems too low. I increased it alot for my testairplane and i was able to do 700kph and a super hard banking turn and stay basically the same speed with barely any thrust envelope change over vanilla opfor cas.

And because of the very low lift they fall like rocks. Not only that, they also nose down very hard on low speed - bringing up the question if that is even normal, because i would assume the aircraft to be balanced around it's lift-force-center (or however it's called) which should mean that the aircraft would just lose altitude but not nose down or up.

It would be nice if oukej could answer this question:

_____

If you are hit in the plane and lose engines your plane is already counted as dead - it will explode at any random time, for no logical reason whatsoever. How hard can it be to create a second hitpoint class "HitEngine" and disable thrust if damage value is at 1 in the gamecode ? Helicopter Rotor blades falling off and Shaft breaking but no Damage model at all for planes. The inconsistency is just infuriating.

I am actually really interested in how this person was able to lose engine power without dying?

and I too am completely infuriated by the blatant lack of attention toward fixed wings on BI's end. Even before the helicopter DLC, we were seeing actual simulation of antitorque rotor, engine loss, rotor loss, etc, and yet we have a binary damage model for the fixed wings - either you're perfectly fine, or you explode and die. There isn't even an indication of damage besides the cracked windshield and smoking engines. In fact, the ability to eject is completely useless, because there is no real reason to eject. Either you are alive and can return to base in perfectly good condition, or you are already dead. No warning sounds, no master caution, nothing.

Go into the editor, place yourself as an A-164, and put in the init "this setDamage 0.999;". this is 0.001 health. But you can still fly just fine.

At the very least, have a threshold of damage where one (or multiple) of several things could happen:

  • You lose all control of the plane, its control surfaces seize and you have as much time as it takes to hit the ground to bail out. This is the most simple approach.
  • The plane catches fire, a master caution warning tone plays, and after x amount of seconds the jet explodes.
  • You lose all engine power and have to either do a dead stick landing (unlikely, since the jets can't even fly straight on their own, let alone glide home), or forcing you to eject
  • Your systems all die - all MFCDs, the HUD, and all weapon systems (can't fire anything).

None of these things would be very difficult to implement. And each and every single one would add a massive amount to the damage model, at least you have the ability to do something if you get damaged (besides just die). Search and Rescue for anything but heli pilots pretty much doesn't exist.

Of course it would be nice to see not only HULL but also RENG, LENG, INST, FUEL, and more, but at the very least tell us when we are about to die so that we have a chance.

Share this post


Link to post
Share on other sites
I am actually really interested in how this person was able to lose engine power without dying?

Get hit by one missile and survive. That's it, the jet simply doesn't put out as much thrust as it did when fully operationally mint condition. This can cause you alot of pains and even death if you don't treat the plane right afterwards, compromising a situation, and even the entire mission. lmost happened to be in that one Youtube video with the To-199 at night. One missile made me play it safe the rest of the round.

Share this post


Link to post
Share on other sites
Get hit by one missile and survive. That's it, the jet simply doesn't put out as much thrust as it did when fully operationally mint condition. This can cause you alot of pains and even death if you don't treat the plane right afterwards, compromising a situation, and even the entire mission. lmost happened to be in that one Youtube video with the To-199 at night. One missile made me play it safe the rest of the round.

Interesting, I haven't noticed much of a performance drop from getting hit.

But this actually makes me even more angry, if they went to the effort to make you lose some engine power when you're damaged, why did they not make you lose all engine power when you're extremely damaged?

Share this post


Link to post
Share on other sites
Interesting, I haven't noticed much of a performance drop from getting hit.

But this actually makes me even more angry, if they went to the effort to make you lose some engine power when you're damaged, why did they not make you lose all engine power when you're extremely damaged?

Because they want to see you explode when you try landing at base. Lol, no idk.

Share this post


Link to post
Share on other sites
I don't even want to talk about stalling right now - it's so nonexistent that I don't think it's worth worrying about, it's going to be a while until we see that drastic a change to the fixed wing flight model. The aircraft don't stall. They only stall when you are going slowly. You can't actually stall (lose effective list) while going at high speed (which you should be able to). It's not even remotely simulated.

Well I actually did stall earlier today when doing FCLP drills in John_Spartan's Super Hornet... so a stall is possible, its just not as common as usual.

But the issue I've noticed is that there's too much lift at lower speeds as opposed to losing lift at low speeds (as it should be). Sometimes I have to push down on the stick either to land on the runway or deck (regardless of what aircraft), so it goes along with what you're saying, which is the problem.

However I would like to see (maybe it's just my stick settings) a proper throttle, re-learning the Su-33 in DCS jacks me up a bit as they model a real throttle setup...

Edited by EricJ

Share this post


Link to post
Share on other sites
to see not only HULL but also RENG, LENG, INST, FUEL

Well L Eng and R Eng wouldnt have much purpose, because the simulation forces of thrust are not position dependant (i.e. you dont define a memorypoint in the memory LOD of the model where the engines sit). All forces on the plane are applied at a center point wherever that might be. So it doesnt matter if L Eng or R eng would be red - there is no torque by disbalanced thrust, the FM doesnt support it. So a single Engine Hitpoint would be sufficient.

Hitmodel wishes (priority 1 high to 3 low):

1 - HitEngine; Reduces Thrust. ThrustNew(x)= Thrust(x) - (Thrust(x) * DamageHitEngine) // Thrust=f(x) x= maxspeed, thrust coef, ...

2 - HitRudder; Reduces Ruddereffectivenes. rudderinfluence= rudderInfluence - (rudderInfluence* DamageHitRudder)

HitAileron same just with aileronSensitivity

HitElevator same just with elevatorSensitivity

3- HitWing; Reduces Lift. LiftNew(y)= Lift(y) - (Lift(y) * DamageHitWing) // Lift=f(x) x= maxspeed, envelope, ... you get the idea right?

If we already have reduced thrust output (i trust DarkSideSixOfficial's statement) with increased HitHull, decoupling it to a new Hitpoint class should be less then 15min work... Theres no 3rd party involved that could prevent addition to the damage model/ behaviour (compared to physx and rotorlib).

Having this feature would allow way more detailed mod planes. And adding those few simple hitpoints to vanilla models (3 models!) and configs would take 1h at most.

Instruments and Fuel ... i'm not sure if fuel makes sense. Arent currentday fueltanks self-sealing? HitFuel in cars just makes them explode michael bay style. Instruments are more complicated.

Share this post


Link to post
Share on other sites
Well L Eng and R Eng wouldnt have much purpose, because the simulation forces of thrust are not position dependant (i.e. you dont define a memorypoint in the memory LOD of the model where the engines sit). All forces on the plane are applied at a center point wherever that might be. So it doesnt matter if L Eng or R eng would be red - there is no torque by disbalanced thrust, the FM doesnt support it. So a single Engine Hitpoint would be sufficient.

Hitmodel wishes (priority 1 high to 3 low):

1 - HitEngine; Reduces Thrust. ThrustNew(x)= Thrust(x) - (Thrust(x) * DamageHitEngine) // Thrust=f(x) x= maxspeed, thrust coef, ...

2 - HitRudder; Reduces Ruddereffectivenes. rudderinfluence= rudderInfluence - (rudderInfluence* DamageHitRudder)

HitAileron same just with aileronSensitivity

HitElevator same just with elevatorSensitivity

3- HitWing; Reduces Lift. LiftNew(y)= Lift(y) - (Lift(y) * DamageHitWing) // Lift=f(x) x= maxspeed, envelope, ... you get the idea right?

If we already have reduced thrust output (i trust DarkSideSixOfficial's statement) with increased HitHull, decoupling it to a new Hitpoint class should be less then 15min work... Theres no 3rd party involved that could prevent addition to the damage model/ behaviour (compared to physx and rotorlib).

Having this feature would allow way more detailed mod planes. And adding those few simple hitpoints to vanilla models (3 models!) and configs would take 1h at most.

Instruments and Fuel ... i'm not sure if fuel makes sense. Arent currentday fueltanks self-sealing? HitFuel in cars just makes them explode michael bay style. Instruments are more complicated.

That's pretty unfortunate, it's one of the more interesting things that can happen to a damaged aircraft. It can't sort of force the aircraft to yaw in one direction slightly? Similar to how trim would be implemented?

I do agree with your priority list, although INST needs to be in there somewhere. I'm not sure if many of you guys are into this stuff, but if you've ever played DCS A-10 you'll know that 80% of the time, the very first thing to happen when you get hit with machine gun fire is for some sort of electronic system to fail. It's typically CICU - one or both MFCDs go out. Also often is for EGI (GPS navigation and targeting system) and CDU (navigation computer). This is often accompanied by at least some amount of loss of engine power (since they are large and get hit often).

Getting hit by missiles is a different story, depending on how hard you are hit. A near miss will have the same effects as machine gun fire (as stated above). A light hit will cause the same issues as the machine gun fire, and also will usually knock out one full engine, or at least cause a severe drop in power from one or both. A heavy hit will set one or both engines on fire. Because the AC generators from the engines cease to function, almost all electronic power is lost, and because hydraulic power also comes from the engines, there is a significant drop in power of the controls. Wings being torn off is also not unheard of, especially when the missile is a larger one (such as a radar guided SAM).

There are two possible ways BI could implement better damage. The first one is if they remain adamant about not implementing any better hitboxes.

Effects of damage would be seen at different points of health loss:

30% damage (70% HP): slight loss of engine power, light electronics issues (perhaps inaccurate altitude and speed information on MFCDs).

50% damage: more significant loss of engine power, noticeable decrease in responsiveness of control surfaces, loss of either one of the MFCDs

70% damage: significant loss of engine power, maneuvering becomes even more sluggish and difficult.

80% damage: even greater loss of engine power, complete loss of electronic systems. all weapons are locked and unusable. extremely poor maneuverability (simulates manual reversion mode)

90% damage: engines die completely.

(alternatively, if the plane was made so that it didn't explode at 100% damage, that could be when the engines die completely).

The other way to do this would be to implement some hotboxes as we described earlier.

It may be wishful thinking, but it would be nice to have some more specific ones (even if they are not displayed in the typical HULL/etc area).

These might include the gun, and the targeting pod (no more shooting missiles) among others. Having specific wing pylons get hit and become "hung" (marked as red on DSMS, the left MFCD) and therefore unusable might be wishful thinking but would be great to see.

Share this post


Link to post
Share on other sites
That's pretty unfortunate, it's one of the more interesting things that can happen to a damaged aircraft. It can't sort of force the aircraft to yaw in one direction slightly? Similar to how trim would be implemented?

I do agree with your priority list, although INST needs to be in there somewhere. I'm not sure if many of you guys are into this stuff, but if you've ever played DCS A-10 you'll know that 80% of the time, the very first thing to happen when you get hit with machine gun fire is for some sort of electronic system to fail. It's typically CICU - one or both MFCDs go out. Also often is for EGI (GPS navigation and targeting system) and CDU (navigation computer). This is often accompanied by at least some amount of loss of engine power (since they are large and get hit often).

Getting hit by missiles is a different story, depending on how hard you are hit. A near miss will have the same effects as machine gun fire (as stated above). A light hit will cause the same issues as the machine gun fire, and also will usually knock out one full engine, or at least cause a severe drop in power from one or both. A heavy hit will set one or both engines on fire. Because the AC generators from the engines cease to function, almost all electronic power is lost, and because hydraulic power also comes from the engines, there is a significant drop in power of the controls. Wings being torn off is also not unheard of, especially when the missile is a larger one (such as a radar guided SAM).

There are two possible ways BI could implement better damage. The first one is if they remain adamant about not implementing any better hitboxes.

Effects of damage would be seen at different points of health loss:

30% damage (70% HP): slight loss of engine power, light electronics issues (perhaps inaccurate altitude and speed information on MFCDs).

50% damage: more significant loss of engine power, noticeable decrease in responsiveness of control surfaces, loss of either one of the MFCDs

70% damage: significant loss of engine power, maneuvering becomes even more sluggish and difficult.

80% damage: even greater loss of engine power, complete loss of electronic systems. all weapons are locked and unusable. extremely poor maneuverability (simulates manual reversion mode)

90% damage: engines die completely.

(alternatively, if the plane was made so that it didn't explode at 100% damage, that could be when the engines die completely).

The other way to do this would be to implement some hotboxes as we described earlier.

It may be wishful thinking, but it would be nice to have some more specific ones (even if they are not displayed in the typical HULL/etc area).

These might include the gun, and the targeting pod (no more shooting missiles) among others. Having specific wing pylons get hit and become "hung" (marked as red on DSMS, the left MFCD) and therefore unusable might be wishful thinking but would be great to see.

The only problem with this is 9 times out of nine you either have an engine or you don't. The only thing that could possibly limit power is either a faulty valve or the power control lever/linkage gets stuck and then it is stuck at one setting; in which case you would shut the effected engine down...

Share this post


Link to post
Share on other sites
One of the biggest issues I see with the current ingame planes is the glide ratio. If I turn the engines off or get hit by AA and loose the engines the planes feel like they're full of rocks and sink way too fast. Anyway to fix this so planes feel more like planes?

I made a few flights with the engine off and it was extremely difficulty to land. Only manage to land one time, issue that i had:

1 - the plane looks like a car, every turn it looses speed (ground speed), we cant glide. Not to mention high angle banking turn, don't know if it is related to air friction.

2 - In stall condition cant nose up for landing

Share this post


Link to post
Share on other sites

@oukej &Dr. Hladik here is the explanation why aircrafts turn when banked

Forget all the centriwhateverforce explanations that are all over the place. They start on the assumption that the aircraft already turns, you can't explain an effect by assuming the existance of the effect. I had the same thoughts like you, it didnt make sense to me either, so i asked some people and here is the explanation from what i understood (and to me is very logic).

Thoughtexperiment: What happens if you have an airflow directly from the side of the aircraft? It will get pushed to the side because of the force, BUT, as the surface area is not balanced in regards to the center of mass you also get a torque. The vertical stabilizer at the rear of the aircraft has more surface area then the front of the aircraft, and also sits farther back.This example might be extreme, but it illustrates it very well i think^^

Now, what happens if we have the airflow directly from above. The wings are at the center of mass, therefore they do not produce any torque (they do produce a lot of force though). The horizontal stabilizer, just like in the first case, has more surface area and(or) sits further back then the part of the aircrafts body in front of the wings/Center of Mass. Similarly, the aircraft will be pushed down by the force, but at the same time receive a torque that turns the nose into the airflow.

What happens in a banked turn then? (here 45°)

planefm1tquiq.jpgplanefm2nekjz.jpg

The Lift vector (green) produces 2 components: A vertical component (blue) that only partially counters the force of gravity (red), and a horizontal component (yellow).

This results in 2 things if the pilot doesn't do anything further: 1-The force of gravity is greater then the vertical lift component. Therefore a resultant force shows downwards, the aircraft is accelerated towards the ground. 2- The horizontal lift force component to the side has no counter yet, therefore the aircraft is accelerated towards the side.

As the vertical speed downwards increases, we get an increasingly strong "airstream component" in the opposite direction. Same with the horizontal speed to the side. This results in the 2 thoughtexperiment cases i described above. The difference is that, that the vertical and horizontal stabilizers are angled towards the airstreams in this case, and that the airstream in vertical direction comes from the bottom. Therefore without counteraction, the aircraft will turn, but also slowly nose downwards. This is why you have to "coordinate a turn" with slight elevator input to generate more liftforce (so vertical lift component and gravity cancel each other) if you want to prevent losing altitude in the turn.

In the end, the aircraft receives 2 torques, one that is similar to a rudder movement by the vertical stabilizer. And one that is similar to an elevator movement by the horizontal stabilizer.

Kind of like an Autorudder and Autoelevator, hahahaha :yay: (just without moving controll surfaces)

The turning of an aircraft can only be explained by this aerodynamic effect. Because in space/vacuum, turning the lift vector (replaced by a thruster) while keeping the gravitational vector the same just results in the aircraft accelerating into another direction, but not cause a turn. The torque created by this effect causes the turn. This generates a pseudo force called "centrifugal force" which results from the bodies inertia/resistance to acceleration. The centrifugal force brings balance to the force equation (yellow component is now opposed by it)

Edited by Fennek

Share this post


Link to post
Share on other sites

I have tested draconic force and torque values... and for me the only conclusion is, that the something in the force calculation is fucked up. It just doesnt work correctly. Something prevents correct levels of acceleration even when a force is present for certain directions.

Example:

Custom aircraft with fairly high thrust. Notable setting: both draconic torque's set to very low value (0.01), X and Z draconic force to high

Now watch this

config here

So what can you see here? First i decelerate to a stall to provoke free-fall. At some point the aircraft tips over to the left, player inputs don't do anything anymore (nothing wrong here). Because of the high draconic force however, the aircraft falls really slow. According to our information we got from Dr. Hladik/samples the draconic force is what centers it to it's speed vector. But it doesnt do that. For some reason it keeps the aircraft in the air instead by some magical means / prevents true freefall. Pay attention to the First value in velocity (x direction, sidewards velocity) during the "fall" states. It increases only very slightly. The value is somewhat around 7 m/s in sidewards direction during the timeframe i gave it.

And you see something else, the aircraft doesnt go down nose first anymore, because the draconic torque doesnt rotate the aircraft. In another test i increased the Torque really high, to 30. It works completely wrong and is not at all what you would expect. If you fly level, the Y value applies elevator downwards so you always turn into the ground. If you bank 90° it stops and if you rotate 180° it's full torque again. It's basically an "invisible" autoelevator into the ground. The X value is the same, except that it works like a rudder force. Fly level, no effect. Turn 90° maximum effect. It does nothing else. What the hell?

And here the Error of the applied "forces" becomes extremely obvious: draconic forces fairly low (2.5 all), no drac.torque.

Compare the sidewards X velocity from the video (very slow acceleration, ~ 7m/s for draconic X and Z = 30) with the values in the pictures, which where almost immediately at around 27m/s.

2014-10-13_00001spuhh.jpg2014-10-13_00002ogu3w.jpg2014-10-13_000044muh1.jpg

The second value in the Force row - ~ 9.8 m/s² * 1kg . Earth gravity*normalized mass. Now i tilt the aircraft, and earth gravity is applied at the X axis (to the side), first number. Watch what happens between second and third picture. Force stays the same. However, velocity in X direction also stays the same. And there is almost 0 acceleration in that direction. This can't be, it's impossible to have a constant force downwards but travel at a constant speed in that direction! Therefore Force calculation must have an error (by accident or by wrong design) somewhere.

In essence: If the aircrafts nose is not pointed downwards it can't accelerate beyond a finite speed limit. Amongst other things this causes what someone else noted, that aircrafts feel like its impossible to increase sink rate without bringing the nose down (and therefore accelerating), because it seems like there is a finite vertical and horizontal (sidewards) speed possible, which is affected my draconic Force.

Edited by Fennek

Share this post


Link to post
Share on other sites
I have tested draconic force and torque values... and for me the only conclusion is, that the entire force calculation is fucked up somewhere. It just doesnt work correctly.

Example:

Custom aircraft with fairly high thrust. Notable setting: both draconic torque's set to very low value (0.01), X and Z draconic force to high

Now watch this

config here

So what can you see here? First i decelerate to a stall. At some point the aircraft tips over to the left, player inputs don't do anything anymore. Because of the high draconic force however, the aircraft falls really slow. According to our information we got from oukej/samples the draconic force is what centers it to it's speed vector. But it doesnt do that. For some reason it keeps the aircraft in the air instead by some magical means.

And you see something else, the aircraft doesnt go down nose first anymore, because the draconic torque doesnt rotate the aircraft to it's speed vector. If you fly level, the Y value applies elevator downwards so you always turn into the ground. If you bank 90° it stops and if you rotate 180° it's full torque again. It's basically an "invisible" autoelevator into the ground. The X value is the same, except that it works like a rudder force. Fly level, no effect. Turn 90° maximum effect. It does nothing else. What the hell?

And here the Error becomes extremely obvious: draconic forces fairly low (2.5 all)

http://abload.de/thumb/2014-10-13_00001spuhh.jpghttp://abload.de/thumb/2014-10-13_00002ogu3w.jpghttp://abload.de/thumb/2014-10-13_000044muh1.jpg

The second value in the Force row - ~ 9.8 m/s² . Earth gravity. Now i tilt the aircraft, and earth gravity is applied at the X axis (to the side), first number. Watch what happens between second and third picture. Force stays the same. However, velocity in X direction also stays the same. And there is almost 0 acceleration in that direction. This can't be, it's impossible to have a constant force downwards but travel at a constant speed in that direction!

Wholly cow... Those there are some dirty calculations...

Share this post


Link to post
Share on other sites

@Fennek, geometry LOD setup [mass distribution/weight] also plays a mayor role in the FM behaviour. I understand what you are trying to archive, we had made similar mistake. SU35 FM, we are still working on, is based on carefully distributed mass [tail/engine part of the jet are the heaviest] with such setup we managed to archive something closer to FM we would like to see.

Share this post


Link to post
Share on other sites

The center of mass of my geometryLOD is precisely at 0/0/0, it's as balanced as it can get. No, mass imbalance is ruled out. The imbalance also wouldn't prevent sidewards acceleration.

An imbalance in mass would just cause a constant torque (depending on where the COM is in regards to the center of turn). Trying to achieve curve flight by making imbalanced COM is wrong. It's like applying constant autoelevation that you have to counteract in levelflight all the time. The ideal center of mass of a normal airplane is the center of lift (and i assume this is 0/0/0 in the model because i have no other information).

I have not tried to achieve any particular flightmodel here. I only tested extreme values to see what effects produce what and what could be wrong. If it was just a config error that prevents all planes from acting like planes, i'd think other people or myself would have found it by now.

Edited by Fennek

Share this post


Link to post
Share on other sites

@Fennek, I cant really say that I am an expert on RL flight models so I wont argue on that. But for 0/0/0 being the centre of lift in RV engine we can disagree. Look up F35 by Chortles [released A3 mod, based on A2 sample models], and then download and compare this F35 [unfinished by me and Saul] - difference is centre of mass shifted to the rear of jet and we have added extra weight on wingtips [all config values are the same 1:1 by BI A2 FM]. Behaviour completely different, one spins around its axis and misbehaves other is more controllable in flight. How realistic, again I won't argue since I am not an expert bro, but its fun and OK for a game. I would vote for more authentic [don't confuse with realistic] FM/simulation airplane with emphasis on functionality of jet, to get player more involved [rather then TAB/fire].

I agree that more intel on what and how affects the FM so we can do some more authentic representation of our mods would be great.

Share this post


Link to post
Share on other sites

If even the most important fundamental of aircraft flight (banking turn) is not achievable without sideeffects, gravitational free fall doesnt work like intended, and other problems that result from the wrong calculation, i can't call this a functioning flightmodel and it's not OK. It's bugged, and just because it has been in this state for propably a long time is no excuse to not at least try and fix the problems in the system. Why work with forces and torques if you don't apply them correctly? If Autorudder is deemed wrong and removed forever, then i expect the bugs to be fixed.

Edit:

Here is what needs to be done.

1- The draconic torque needs to be changed to create a torque whenever the speedvector does not match the orienation of the airplane, so the airplane ends up aligned with the speedvector. This automatically accounts for the single use that the torques currently have, which is to drop the nose when the aircraft loses lift and is accelerated downwards. This change would also automatically account for a banked turn, IF the forces (see 2) are fixed. Additionally it would make planes flying backwards extremely difficult, if not impossible, as the angle difference between aircraft orientation and the speedvector would be extremely large, therefore a large torque is produced, causing the aircraft to flip around rapidly. That's at least 3 fixes with one stone.

The draconic torque you set will controll how strong the aircraft will react to speedvector changes. The torque should be proportional to the difference in angle (torque per angle degree would make sense i guess).

Maybe 1 already works like described and just number 2 is broken in some way which causes 1 to fail as well? I can't determine that. Atm it seems both torques only pull the aircraft down. No matter what.

2- Something prevents free acceleration in directions other then Z direction (thrust direction) even though forces are applied. This needs to be looked at. Do not prevent acceleration in any direction. With 1 implemented, the aircraft will automatically turn into the correct direction. Maybe draconic forces (not torques) need to be removed or reworked, i'm not entirely sure about them, as it's hard to get a feeling for what exactly they do and what they might not do/fail to do.

Edited by Fennek

Share this post


Link to post
Share on other sites

Awesome This is the due diligence that can help see the flight model for fixed wing aircraft fixed. Hope bis takes your findings on board and helps a have that's getting better ask the time, get better in this part (fixed wing) of the game as well. At least to getting the important fundamentals right

Share this post


Link to post
Share on other sites
Wouldn't an ideally balanced plane stay level no matter how much lift you have? Can the balance even be influenced by model settings or is this some hardcoded stuff?

I wouldn't say so. Depends on the airplane's longitudinal stability, but generally I wouldn't like to sit in a plane, that doesn't tend to dive after loosing speed. We don't have trimming (or autotrim.) for airplanes, so you have to constantly adjust and maintain your input depending on the speed (and configured envelope). On a standard airplane the lower airspeed the more input on elevators is needed to keep the plane leveled (up to stalling).

The balance - as already noted by John Spartan - is dependable on the model. (you can also check http://www.armaholic.com/page.php?id=3157 )

The inconsistency is just infuriating.

You may say so and it wasn't an easy decision for the project. We won't try to hide the fact that airplanes have become a poorer neighbor of the helicopters (those damn rotoflipperz even have a lift force simulated on each individual surface now ;) :P). We were given the chance to utilize the Take On legacy and experience and we thought something relatively immediate for all of us to enjoy is better than any uncertainties in a far future. Even though we had been aware that it would have highlighted the shortcomings elsewhere. The other thing is how it revolves around the combined arms combat - to be sorely honest, there will always be more cannonfodder on the ground or close to it than the elite hawks high in the sky. The love a feature has - as democratic as it gets. (And I am saying that as a total sim-freak). We are still gonna try to do our best to improve the system and fix the obvious issues.

@oukej &Dr. Hladik here is the explanation why aircrafts turn when banked

Noone ever doubted ;)

Custom aircraft with fairly high thrust. Notable setting: both draconic torque's set to very low value (0.01), X and Z draconic force to high

...

config here

.

Thanks a lot for the solid elaboration! I'll go thru it in detail as soon as I can.

Just by looking at it - I wouldn't go into the extreme numbers (30 or 0.0x) when configuring the forces and torques. Ideally start just with 1.0 and build up. Conversely you may want to try higher TorqueCoefs. All these X-ZCoefs are applied in the relation to the model world. Everything goes together or against each other with envelope and control surfaces' coefs and relatively to the maxSpeed.

Also I've noted the use of VTOL - it hasn't been revisited in A3. It should work, but no assurance.

---

I would vote for more authentic [don't confuse with realistic] FM/simulation airplane with emphasis on functionality of jet, to get player more involved [rather then TAB/fire].

I support that.

As for the moar info - we may try to put some more into the modding samples.

Edited by oukej

Share this post


Link to post
Share on other sites
I wouldn't say so. Depends on the airplane's longitudinal stability, but generally I wouldn't like to sit in a plane, that doesn't tend to dive after loosing speed.

Yes, agreed, once you apply the aerodynamic effect of the stabilizers it becomes obvious that it has to nose downwards if gravity is stronger then lift. Because you accelerate downwards, and the aerodynamic effect of the stabilizer automatically turns the nose into the velocity vector.

Noone ever doubted

Ok then i confused it, but Dr. Hladik wrote in one of the tickets that he would appreciate a proper explanation because he didn't believe the centrifugal force explanations of other people (rightfully so).

_______

I did more testing, because i hadn't done tests with draconic forces below 2.5 yet.

First Test - draconic forces 1, torques 0.01. Extreme loss of speed during turns, after changing direction you glide for some time. More extreme then with 2.5 obviously. Maximum Falling velocity in X (sidewards) direction is 40m/s.

Fourth Test - draconic forces 1 and torques 1. Like First Test, but additional magic torque turns you into the ground, no matter how you angle your plane.

Second Test - draconic force and torque 0.01 on all acounts. It feels like flying in space. Lift was generated (countering gravity in level flight), but you continued traveling downwards until you killed all downwards velocity by pointing the thrust vector in the opposite direction.

Third Test - draconic forces 0.01 and torques 1. Flying in space like before, but not only do you get pulled down by gravity, you also get rotated towards the ground by magictorque. Banking does not change the torque to a positive one.

Observation:With high draconic forces (30) i was only able to get the AOA value to 0.2, with low draconic forces (1) only to 0.5. I'm not sure what unit the AOA value is in or if it is relative to something? It would be good to have input on what this value represents.

Conclusion: Values in draconic torque&forces ranging from 0 up to 30 do not produce believable flightbehaviour / solve the problems i mentioned earlier.

Also I've noted the use of VTOL - it hasn't been revisited in A3. It should work, but no assurance.

It does, otherwise my plane wouldnt be airbone - it has no wheels, only skids. I also tried using the config on a vanilla plane but with vtol=0 to check if vtol was the problem - it wasnt.

Edited by Fennek

Share this post


Link to post
Share on other sites
Y

Observation:With high draconic forces (30) i was only able to get the AOA value to 0.2, with low draconic forces (1) only to 0.5. I'm not sure what unit the AOA value is in or if it is relative to something? It would be good to have input on what this value represents.

Conclusion: Values in draconic torque&forces ranging from 0 up to 30 do not produce believable flightbehaviour / solve the problems i mentioned earlier.

/// angle of incidence - difference between forward and airfold chord line - def. val is 3*H_PI/180;

angleOfIndicence = 0.05235987; (3º degrees)

This value is in radians:

- 0.2 rads / 11.5º

- 0.5 rads / 28.6º

I`ve already ask, how AoA is related to Arma Flight Model if the lift is constant.

index.php?action=dlattach;topic=423.0;attach=248

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now

×