Juno Ace Ultimate Free For All

[FahmiRBLX]

2 hours ago, Pds314 said:

Did you have trigger set?

Trigger set? Is it another mod or a BDAC Part?

EDIT : Through Wiki-ing, now I see what you meant. I'll try that out. Probably tomorrow.

Edited December 26, 2018 by FahmiRBLXian

[Alioth81]

I am surprised that the Piglet is doing that good.

I just checked my plane again and found its main gun (around which the whole plane is designed) to be significantly off axis... Kerbal engineering at its peek :-)

So on my PC Piglet Bs are killing their predecessors.

I have to say I am surprised how many 30mm hits a plane can take. I assumed that a single hit should be devastating so maybe Vulcans are the better choice.

 

I am eagerly waiting for super sonic plane designs to see how it is done.

I tried to design a fast plane before i submitted the piglet and although it made 400 m/s I just could not handle all the g-force issues that arise at that speed with my design. Many brave Kerbals either knocked themselves out due to G-force or were diving the plane into the ground because of lacking maneuverability.

Is there a way to limit/alter control surface response based on speed? AoA% limits its performance at low speed quite a lot. How do you do it?

As alternative how does the BDAc G-limit work? It does not seem to be active in a lot of situations (evasion and pull out below min alt at least).

 

Edited December 26, 2018 by Alioth81

[GDJ]

6 minutes ago, Alioth81 said:

I am eagerly waiting for super sonic plane designs to see how it is done.

I have one. I'll see how it does with some BDAC mods.

[Wanderfound]

50 minutes ago, Alioth81 said:

 

I am eagerly waiting for super sonic plane designs to see how it is done.

The Sparrowhawks will crack the sound barrier, but it takes them a while to get there.

It is possible to rip the wings off them, but you have to be trying hard; it takes a sudden 15g low-altitude turn at maximum speed to do it.

Edited December 27, 2018 by Wanderfound

[Alioth81]

I had some time to look at the AI of BDAc and although I am not a programmer I tried to make it easier so that the AI does not pull the plane into black out or disintegration at high speeds.

The idea was to leave everything as it was and just add a check to gradually reduce max control input above a certain speed. For this I added two more settings in the AI autopilot.

"Corner Speed" (for lack of a better name) and "Steer Limit At Max Speed"

Basically the behavior until "Corner Speed" is the same as before and it will fully pull the stick back

However between "Corner Speed" and "Max speed" it will reduce control input in a linear fashion until it reaches the value defined at "Steer Limit At Max Speed" at the speed "Max speed" (beyond that the reduction will continue in the same linear fashion until 0.05 which is the true minimum)

It is not a perfect solution but it works better for me than the limit G force slider.

The default setting for "Steer Limit At Max Speed" is 1 therefore no plane will be affected except if someone truly changes something.

Maybe someone like @SuicidalInsanity can integrate that into the beautiful custom version? (if others want to use it)

Of course if people have better ideas and want to improve it feel free to do whatever you want

I added the code I changed in BDmodulePilotAI.cs below
 

Spoiler

 

new settings:

        [KSPField(isPersistant = true, guiActive = true, guiActiveEditor = true, guiName = "Steer Limit At Max Speed"),
            UI_FloatRange(minValue = .1f, maxValue = 1f, stepIncrement = .05f, scene = UI_Scene.All)]
        public float maxSteerAtMaxSpeed = 1;
        [KSPField(isPersistant = true, guiActive = true, guiActiveEditor = true, guiName = "Corner Speed"),
            UI_FloatRange(minValue = 10f, maxValue = 500f, stepIncrement = 1.0f, scene = UI_Scene.All)]
        public float cornerSpeed = 200f;

new unclamped limits:

		Dictionary<string, float> altMaxValues = new Dictionary<string, float>
		{
			{ nameof(defaultAltitude), 100000f },
			{ nameof(minAltitude), 30000f },
			{ nameof(steerMult), 200f },
			{ nameof(steerKiAdjust), 20f },
			{ nameof(steerDamping), 100f },
            { nameof(cornerSpeed), 3000f },
            { nameof(maxSpeed), 3000f },
			{ nameof(takeOffSpeed), 2000f },
			{ nameof(minSpeed), 2000f },
			{ nameof(idleSpeed), 3000f },
			{ nameof(maxAllowedGForce), 1000f },
			{ nameof(maxAllowedAoA), 180f },
		};

new function:

        float GetSteerLimiterForDesignAndSpeed()
        {
            float speed = (float)vessel.srfSpeed;
            float designlimiter = 1;

            if (speed > cornerSpeed & maxSpeed > cornerSpeed & maxSteerAtMaxSpeed < 1)
            {
                designlimiter = 1- ((speed - cornerSpeed) / (maxSpeed - cornerSpeed) * (1f - maxSteerAtMaxSpeed)); // linear approximation to set max control input when above corner speed
                if (designlimiter < 0.1f)
                {
                    designlimiter = 0.05f; // added just in case to ensure some input is retained no matter what happens
                }
                debugString.Append($"Above Corner Speed limit to: {designlimiter}");
                debugString.Append(Environment.NewLine);
            }

            return Mathf.Clamp01(designlimiter);
        }

new check with the new function added in the flyToPosition function just above the finalMaxSteer debug text:

            float designlimiter = GetSteerLimiterForDesignAndSpeed();

            if (finalMaxSteer > designlimiter)
            {
                finalMaxSteer = designlimiter;
            }

			//test
            debugString.Append($"finalMaxSteer: {finalMaxSteer}");
            debugString.Append(Environment.NewLine);

 

 

 

 

 

 

 

Edited December 27, 2018 by Alioth81

[SuicidalInsanity]

On 12/26/2018 at 3:52 PM, Alioth81 said:

I am eagerly waiting for super sonic plane designs to see how it is done.

Good area ruling, proper wing shape (high swept, low aspect), and having more than one engine, basically.

@Pds314: Replacement for the Meteor -  IA-31 Fishtank.
3-ish tons, 2 engines, 2 AIM-9,s Vulcan, 3 flares, supercruise capable: 40 points

Edited December 28, 2018 by SuicidalInsanity

[Pds314]

On 12/27/2018 at 6:59 PM, SuicidalInsanity said:

Good area ruling, proper wing shape (high swept, low aspect), and having more than one engine, basically.

@Pds314: Replacement for the Meteor -  IA-31 Fishtank.
3-ish tons, 2 engines, 2 AIM-9,s Vulcan, 3 flares, supercruise capable: 40 points

You can definitely have a supersonic plane on one engine. I have one that will do 400 at sealevel, albeit its tendency to crash on takeoff, lack of a moving rudder, and questionable controls means it is less than ideal as a fighter.

 

Edited December 29, 2018 by Pds314

[SuicidalInsanity]

On 12/27/2018 at 3:39 PM, Alioth81 said:

Maybe someone like @SuicidalInsanity can integrate that into the beautiful custom version? (if others want to use it)

Of course if people have better ideas and want to improve it feel free to do whatever you want

I added the code I changed in BDmodulePilotAI.cs below

Easy enough.
Better question is what's pds314's stance on changing AI pilot behavior/available tuning widgets mid contest.
On one hand, if we all start teching into post-mach aircraft the ability to delineate sub/supersonic control regimes would be useful. The IA-32 could definitely use this...
On the other hand, if the change is made, then current entries and battles will have been done to and soon would be going against a different standard; they would either have to be re entered, redone, or become potentially invalidated in the face of the new AI tunings.

 

[Pds314]

1 hour ago, SuicidalInsanity said:

Easy enough.
Better question is what's pds314's stance on changing AI pilot behavior/available tuning widgets mid contest.
On one hand, if we all start teching into post-mach aircraft the ability to delineate sub/supersonic control regimes would be useful. The IA-32 could definitely use this...
On the other hand, if the change is made, then current entries and battles will have been done to and soon would be going against a different standard; they would either have to be re entered, redone, or become potentially invalidated in the face of the new AI tunings.

 

Well, the current leader, the Piglet, isn't going anywhere near the speed of sound unless you strap a RAPIER engine to it or something. It goes from efficient flyer to transonic brick at 250 m/s.

On 12/26/2018 at 3:52 PM, Alioth81 said:

I am surprised that the Piglet is doing that good.

I just checked my plane again and found its main gun (around which the whole plane is designed) to be significantly off axis... Kerbal engineering at its peek :-)

So on my PC Piglet Bs are killing their predecessors.

I have to say I am surprised how many 30mm hits a plane can take. I assumed that a single hit should be devastating so maybe Vulcans are the better choice.

 

I am eagerly waiting for super sonic plane designs to see how it is done.

I tried to design a fast plane before i submitted the piglet and although it made 400 m/s I just could not handle all the g-force issues that arise at that speed with my design. Many brave Kerbals either knocked themselves out due to G-force or were diving the plane into the ground because of lacking maneuverability.

Is there a way to limit/alter control surface response based on speed? AoA% limits its performance at low speed quite a lot. How do you do it?

As alternative how does the BDAc G-limit work? It does not seem to be active in a lot of situations (evasion and pull out below min alt at least).

 

My tests and battles seem to show that 20 mm = kind like 50s were in WWII. You need an awful lot concentrated on critical parts to stop a plane from flying, especially a bigger aircraft, whereas a single burst from a single 30mm is enough to do very serious damage, like 20mm in WWII. In testing, a 1-Vulcan craft going head-on with a Piglet is usually gonna take significantly more damage than it or even die.

One particularly-nasty thing I've seen 30mm do is just in a split-second rip off the rear control surfaces indendently of each other and all four engines, such that a fairly big plane spins out of control with no power and crashes.

Edited December 29, 2018 by Pds314

[Pds314]

Regarding control authority vs speed, I'm actually having the opposite issue. My plane bleeds energy a lot at low speed and I'd rather it didn't try to turn much at all below 200 m/s and was still aggressive at 300-350 m/s.

This is actually one way I've found that the Piglet can catch an aircraft going almost mach 1. AIM-9s and long-range gunfire force the plane to turn and it will lose energy. Often enough to put it in hit range or into a turning fight it really should not try to enter because it can't maintain energy trying to out-turn a piglet.

I've found actually the main way my plane wins is if it ends up taking the fight into the vertical. Quite simply, pulling a zoom climb while dodging attacks is not what the piglet is ideal at.

Edited December 29, 2018 by Pds314

[Triop]

I could enter my mini MiG 15, but then you have to replace the Juno first because I gave it a little more "oempf" in the cfg file...

[sturmhauke]

I either completely suck at designing FAR jets, or there's something wrong with my installation. As evidence of the latter, I downloaded @SuicidalInsanity's Fishtank, and it can't even take off under AI control without putting a crater in the runway. Mine gets airborne, but then tumbles out of control when trying to maneuver.

[Pds314]

On 12/29/2018 at 11:26 AM, Triop said:

I could enter my mini MiG 15, but then you have to replace the Juno first because I gave it a little more "oempf" in the cfg file...

Whoaaaaa. Okay that thing is nuts. IRL MiG-15bis has 26.5 kN of thrust and a top speed in the 300-310 m/s sort of range.

It also has an empty weight of about 3600 kg and a length and wingspan of about 10 m, so other than the length that's actually not too "mini" from the looks of it. Unless those wings are really clipped into the fuselage.

My 'Canyon Chaser' aircraft is, except for the mass, pretty similar in size to a MiG-15 and has within a few m/s the same top speed. It was likewise not designed with proper area-ruling.

People do often compare the MiG-15 and F-86 and not without reason but the F-86 had a much larger empty weight. Like the weight of a sedan larger.

Edited January 2, 2019 by Pds314

[aleksey444]

13 hours ago, sturmhauke said:

I either completely suck at designing FAR jets, or there's something wrong with my installation. As evidence of the latter, I downloaded @SuicidalInsanity's Fishtank, and it can't even take off under AI control without putting a crater in the runway....

Same. I can't get the Fishtank to get off the ground.

13 hours ago, sturmhauke said:

Mine gets airborne, but then tumbles out of control when trying to maneuver.

Check FAR stability derivatives.  Make sure everything is green.

Edited January 2, 2019 by aleksey444

[sturmhauke]

16 minutes ago, aleksey444 said:

Check FAR stability derivatives.  Make sure everything is green.

Yeah that's the part I don't get. There can be nothing but green and white at flight speeds, but my plane still stalls out asymmetrically on a turn and can't recover. My BAD-T IV entry finished in the middle of the pack, so I thought I had a handle on the basics. I've tried using a larger empennage, and both fixed + edge control surfaces and all-moving control surfaces, but I haven't really got anywhere. I dunno, maybe my wing root is too wide and I need narrower swept wings instead of a largish delta shape. My transonic stats are kind of a mess, but I'm not really trying to go supersonic so I don't know if that matters much. I'll post some pics when I get home.

[SuicidalInsanity]

Huh. I'll take another look at the Fishtank - is it a decide to veer off to the right and roll over instead of taking off issue? I thought I fixed that - could be I uploaded the wrong version of the plane by mistake.
 

Edited January 3, 2019 by SuicidalInsanity

[GDJ]

Managed to get this little beast past 400 m/s (Mach 1.3 roughly).



Airplane Plus cockpit, SXT rear tank, Twin Juno's in a vertical arrangement.

It's a work in progress......

Edited January 3, 2019 by GDJ

[sturmhauke]

2 hours ago, SuicidalInsanity said:

Huh. I'll take another look at the Fishtank - is it a decide to veer off to the right and roll over instead of taking off issue? I thought I fixed that - could be I uploaded the wrong version of the plane by mistake.

Yeah, that's exactly what it's doing.

[SuicidalInsanity]

Yep, veering off the runway issue. Needed bigger vert stabilizer, go figure. Fixed the DL with the corrected version of the plane that can takeoff.

That said, if you're trying to get design tips, the Fishtank could stand more as an example of what not to do; short enpennage, so needs a huge tail, if any of the aero surfaces are even slightly out of position of where they currently are, the plane has issues and trouble taking off, the overall airframe would be better suited for engines that weighed more, so the CoM is a bit further forward than ideal, etc.
It was mainly an attempt to resurrect rebuild from scratch an old design that sort of ended up morphing into something that looks an English Electric Lightning II, as built by Sukhoi out of spare MiG 21 parts following a game of Telephone for the plans

Edited January 3, 2019 by SuicidalInsanity

[SuicidalInsanity]

So. Was running a test battle to do some craft tuning, and apparently the AI will not target opposing craft with GLOC'd pilots, which may be an issue if the plane was, say, doing a banking turn it could continue and orbit indefinably.

[sturmhauke]

36 minutes ago, SuicidalInsanity said:

That said, if you're trying to get design tips, the Fishtank could stand more as an example of what not to do

Hmm, well mine does look kinda like yours. I was originally going for something like the F4 Phantom. My daughter said I should make it pink and purple, so here it is:

Quote

It was mainly an attempt to resurrect rebuild from scratch an old design that sort of ended up morphing into something that looks an English Electric Lightning II, as built by Sukhoi out of spare MiG 21 parts following a game of Telephone for the plans

LOL

[SuicidalInsanity]

In general, the Fishtank's a not a terrible design, and there are only so many ways of designing a superimposed twin engine airframe; it was more the specific 'why this, instead of something perhaps more sensible? 'design choices present in the Fishtank that I was referring to.

Edited January 3, 2019 by SuicidalInsanity

[sturmhauke]

I messed with my design more and made something less terrible, but I'm kinda frustrated with the whole thing. Maybe I'll take a crack at it again later, but I'm gonna bow out for now.

[SuicidalInsanity]

On 1/2/2019 at 8:51 PM, SuicidalInsanity said:

So. Was running a test battle to do some craft tuning, and apparently the AI will not target opposing craft with GLOC'd pilots, which may be an issue if the plane was, say, doing a banking turn it could continue and orbit indefinably.

Did some poking at the AI and managed to fix this so the AI will still engage planes with KO'ed pilots. -BDAc 1.2.3.5.SI. Compiled for 1.5.1, since that's what the JAU FFA is using, but will work in 1.6 as well. I also implemented and expanded upon the AI tweaks aleksey444 proposed:

...I may have gone a little overboard. All advanced settings are disabled and hidden by default, so they shouldn't impact any currently competing craft, but they're there for personal use outside the context of this competition by those who want to.

[aleksey444]

9 hours ago, SuicidalInsanity said:

I also implemented and expanded upon the AI tweaks aleksey444 proposed...

I proposed something?

I don't remember, but you made something cool, so that's great.  That Speed Adjusted Steer Limiter looks useful - maybe it'll make my plane not tear itself apart at 300m/s.

If I may propose something, please document your changes.  In fact, you seem to know a lot about BDA, maybe document the old tweaks as well .  I still don't really understand what pitch Ki does, for example.