The 5th Generation Fighter challenge [FAR]

[CrisK]

I was inspired by Halsfury' flanker, and I decided to create two variations of my "Silver".

#1: USAF Lockheed style:

Javascript is disabled. View full album

#2: Flanker Sukhoi style:

Javascript is disabled. View full album

The USAF version is more compact, shorter, and more aerodynamically stable. I designed it with a focus on "stealth". The Flanker variant is less stable, a little bit longer, the body's wider, and the engines are spaced further apart.

Both are capable of 17g turns, vertical climbs, supercruising, and supermaneuverability. SAS optional.

[TheHengeProphet]

That is the trend, yes, but does that mean it has to be? I feel that so long as the plane has the given requirements to be 4.5gen or higher, it should be fine. For example, unless you're specifically banning day-fighters, a radar datalink or radome should be required.

Hmm... now that I think about it, that's the only thing that really makes a difference here. There isn't a lot of Gen5 tech available, which means you're otherwise operating purely on the concept of aerobatic performance and lifting load.

Managing to get a fighter that can compete which weighs under 10t would be an interesting challenge in itself...

Sorry for constantly questioning your rules. I really need to get out of that habit.

Nice planes, CrisK. Any chance we can manage copies of those things?

Edited August 31, 2015 by TheHengeProphet

[Halsfury]

That is the trend, yes, but does that mean it has to be? I feel that so long as the plane has the given requirements to be 4.5gen or higher, it should be fine. For example, unless you're specifically banning day-fighters, a radar datalink or radome should be required.

Hmm... now that I think about it, that's the only thing that really makes a difference here. There isn't a lot of Gen5 tech available, which means you're otherwise operating purely on the concept of aerobatic performance and lifting load.

Managing to get a fighter that can compete which weighs under 10t would be an interesting challenge in itself...

Sorry for constantly questioning your rules. I really need to get out of that habit.

Nice planes, CrisK. Any chance we can manage copies of those things?

Second that, Crisk's planes are nice.

I suppose that an aircraft under 10 tons which can compete in a dogfight is possible, but unlikely considering all the conflicting requirements of modern warfare.

My rules are designed to make a middle of the road kind of fighter, rather than have diverging designs like in the previous iteration of the challenge.

That being said there are all kinds of aircraft in use today, with all kinds of different specs.

[CrisK]

Nice planes, CrisK. Any chance we can manage copies of those things?

Thanks! Sure thing. I'll create copies without the internal infernal robotics parts tomorrow to reduce the mod count. Other than IR, they both use BDArmory, b9 wings, and adjustable landing gear.

Second that, Crisk's planes are nice.

I suppose that an aircraft under 10 tons which can compete in a dogfight is possible, but unlikely considering all the conflicting requirements of modern warfare.

My rules are designed to make a middle of the road kind of fighter, rather than have diverging designs like in the previous iteration of the challenge.

That being said there are all kinds of aircraft in use today, with all kinds of different specs.

Thanks man! I like the idea that you proposed a while back of there being different categories of planes. Multirole, air superiority, bomber, close air support, electronic warfare, utility, etc.

PS: BDArmory was just updated again with some goodies.

[TheHengeProphet]

BDA updated again? OOH, I'll have to go check that out!

Yeah, the only mod among those I do not use is IR, heh. For what do you use IR on them?

Actually, having multiple categories would be interesting, and allow for being able to tell how good of a multirole fighter a given plane could be by how they place in each category. The main downside to this, is it means more work.

[Halsfury]

Thanks man! I like the idea that you proposed a while back of there being different categories of planes. Multirole, air superiority, bomber, close air support, electronic warfare, utility, etc.

PS: BDArmory was just updated again with some goodies.

Now that would really require a 2.0 version of the challenge, but I like the idea.

That would be even better if we had a BD expansion pack with some proper missiles, like the AIM-54 Phoenix, and the AA-6 "Acrid", with mach 4-6 speed and long range ability.

Also all the missiles in BD could use a significant buff in performance.

I will consider rules for interceptors, and ground attack but somebody else has to make the ground attacker obstacle course and put it in a save file.

I'm thinking of having 6 installations for an attacker to take on to prove combat worthiness

Ignore all that, I propose a mini challenge!

Create a long range Air to Air missile for your aircraft from scratch (inside KSP), take BD armoury's explosive warhead and the missile guidance system, set the guidance to A to A

Use the radar data receiver to keep it in touch with the aircraft as it flies.

Good luck!

Edited September 1, 2015 by Halsfury

[TheHengeProphet]

I have yet to figure out how to even launch these things, hah.

[CrisK]

Now that would really require a 2.0 version of the challenge, but I like the idea.

That would be even better if we had a BD expansion pack with some proper missiles, like the AIM-54 Phoenix, and the AA-6 "Acrid", with mach 4-6 speed and long range ability.

Also all the missiles in BD could use a significant buff in performance.

I agree with you that the missiles are heavily under-performing right now. I launched 6 AMRAAMs and 4 sidewinders at a target directly in front of me less than 1km away and they all missed. I was locked on!

@LORDPrometheus created some nice long range air to air missiles in his P.E.W. add-on. There's the AIM-54 and the R-37, plus others that I don't really know anything about.

[Halsfury]

I have yet to figure out how to even launch these things, hah.

Yeah I've been messing around with modular missiles and they are sorely lacking, First off my test missile is a truly crude contraption, being basically a T-10 Hammer booster (yeah the biggest A to A missile ever yet created) with 4 fins and a fairing nosecone with a big wad of C4 in it.

The missiles don't actually need a radio data receiver, but the guidance system is so horrible, plus the missile sometimes starts controlling the aircraft, so I'm going to see how it works as a Ground to Air missile maybe that will tell me something about how they work!

- - - Updated - - -

I agree with you that the missiles are heavily under-performing right now. I launched 6 AMRAAMs and 4 sidewinders at a target directly in front of me less than 1km away and they all missed. I was locked on!

@LORDPrometheus created some nice long range air to air missiles in his P.E.W. add-on. There's the AIM-54 and the R-37, plus others that I don't really know anything about.

Yeah but PEW is not compatible or I might have downloaded it wrong. do you just drop PEW into the GameData? cause I dropped the parts into the BD armoury file.

EDIT: I tried PEW again but it's behaving badly still. I'll have to keep trying to get KSP's parts to turn into a successful A to A missile.

I've found that missile making is just about as hard as making the plane. I've been firing them at a subsonic flying wing fighter I made a while back (the Bhlom und Voss P209) Which is successfully evading my very large missile every time. The missile AI is quite stupid about identifying lead and trying to intercept the target (which is surprising considering BD armoury can calculate lead for other projectiles just fine). I've identified that to solve the problem I need to make the missile more maneuverable, but so doing has lead to a lot of flying backwards.

In an even more annoying twist it seems totally impossible to use the missile manager in conjunction with a player controlled aircraft, this will usually result in the missile AI taking over the plane's control.

Edited September 1, 2015 by Halsfury

[Exploro]

There's an intrinsic problem to this challenge: what constitutes a 5th gen fighter? A larger multirole fighter like the Eurofighter or Rafale, or a something a bit smaller like the F-35 or Saab Gripen?

What makes a 5th generation fighter is not so much how large a fighter is. The F-35 is a much smaller airframe compared to the F-22, the Chinese J-20, or the Russian T-50. Yet the Lightning II is a 5th generation fighter.

The key areas that define a 5th generation fighter are:

• Low observability techniques used to drastically reduce a warplanes radar, infrared, acoustic, electronic and visual signatures
  
• Sensors and on board computer systems that provide the pilot with unparalleled situational awareness.
  

My thoughts from the beginning of reading this thread is that the challenge of designing a true facsimile of a fifth generation plane on KSP is more or less impossible. There are no means gauge signature reductions. There are no means to replicate the onboard sensory and computer processing capabilities of these planes on KSP (which is why I suppose Halsfury has focused on performance and design metrics rather than what defines a 5th generation plane in real life).

[TheHengeProphet]

Yeah, the missiles are a bit ridiculous to try to build...

And as for that 5th Gen stuff, I believe the best we can actually approximate is making the best 4.5 Gen we can, and attempt to make it "look" as stealthy as possible.

[CrisK]

Yeah I've been messing around with modular missiles and they are sorely lacking, First off my test missile is a truly crude contraption, being basically a T-10 Hammer booster (yeah the biggest A to A missile ever yet created) with 4 fins and a fairing nosecone with a big wad of C4 in it.

The missiles don't actually need a radio data receiver, but the guidance system is so horrible, plus the missile sometimes starts controlling the aircraft, so I'm going to see how it works as a Ground to Air missile maybe that will tell me something about how they work!

- - - Updated - - -

BDA updated again? OOH, I'll have to go check that out!

Yeah, the only mod among those I do not use is IR, heh. For what do you use IR on them?

Actually, having multiple categories would be interesting, and allow for being able to tell how good of a multirole fighter a given plane could be by how they place in each category. The main downside to this, is it means more work.

I use IR to build missile rails. Have you ever experienced BDArmory's missiles colliding with your plane and blowing it up? Well, when they're attached to IR parts that doesn't happen. It's also great for building internal missile racks that extend out of the plane with an action group, fire, then draw back into the plane. This process takes about 2 seconds.

Yeah but PEW is not compatible or I might have downloaded it wrong. do you just drop PEW into the GameData? cause I dropped the parts into the BD armoury file.

EDIT: I tried PEW again but it's behaving badly still. I'll have to keep trying to get KSP's parts to turn into a successful A to A missile.

I've found that missile making is just about as hard as making the plane. I've been firing them at a subsonic flying wing fighter I made a while back (the Bhlom und Voss P209) Which is successfully evading my very large missile every time. The missile AI is quite stupid about identifying lead and trying to intercept the target (which is surprising considering BD armoury can calculate lead for other projectiles just fine). I've identified that to solve the problem I need to make the missile more maneuverable, but so doing has lead to a lot of flying backwards.

In an even more annoying twist it seems totally impossible to use the missile manager in conjunction with a player controlled aircraft, this will usually result in the missile AI taking over the plane's control.

Hmm. P.E.W. works fine for me. Did you download the updated version from @LORDPrometheus's Dropbox account?

I'm fairly certain that the modular missiles are meant to be long-range cruise missiles. They're not really designed for AA.

Here are craft files.

1. USAF variant armed with IR.

2. USAF variant unarmed with minimal mods.

3. Flanker variant with all mods including IR.

4. Flanker variant unarmed with minimal mods.

These planes are somewhat challenging to fly. I recommend using a soft touch when flying them at low altitudes (under 5k). Pilot assistant's throttle control is very helpful. I put fuel in both the tank and the wings. Empty the forward tank if you want to be more acrobatic at the cost of losing half your flying time. I limited the thrust to 70, but it should fly nicely at anywhere above 45. 70 seems to be a sweet spot.

For the USAF variant: keep the AOA below 60. It's stable up to around 70, and it won't stall until you pass around 75. It's the more stable design, so it can't pull stunts as easily unless the forward tank is emptied.

For the flanker variant: Keep the AOA below 70. This variant is more acrobatic and can/will stall if you go above an AOA of 75. It's capable of really weird and fun stunts, at the expense of burning speed. The cockpit is angled downwards, so there's more drag and the navball is a bit off. This gives a great IVA view though, and it's really cool for dogfighting in the IVA view. It's also amazing for IVA landings because you can see below you as you land.

Neither need airbrakes to land. Hold the S key when you touch down and the rear fins will slow you down. It's easy to land on a very short strip (including Eskandare's aircraft carrier) with these planes, even with IVA. The USAF is especially forgiving with landings.

Edited September 3, 2015 by CrisK

[TheHengeProphet]

Stable up to 70 degrees AoA? Wow, that's impressive. What's it's stall angle?

I've been trying to figure out why my plane stalls asymmetrically at high AoA, but it's a bit of a mystery. I might have to build the wing as a single main piece and make my own leading edges to see if that fixes anything.

[CrisK]

Stable up to 70 degrees AoA? Wow, that's impressive. What's it's stall angle?

I've been trying to figure out why my plane stalls asymmetrically at high AoA, but it's a bit of a mystery. I might have to build the wing as a single main piece and make my own leading edges to see if that fixes anything.

I'll check when I get a free moment to play KSP again!

To be honest, the high performance was half voodoo/luck and half science. Every time I made a change to the plane I checked FAR's analysis window to see if the AOA increased or decreased. If the performance went down, I loaded an older save. By doing that I found 3 different wing configurations that performed well enough.

If I increased the wing's length, the AOA stability/performance decreased but lift increased. There's a healthy balance between lift and AOA performance. You can make the plane stable at an AOA of 90+ on the analysis graph, but it won't have enough lift at higher altitudes so the entire plane will stall.

Tweaking the control surfaces is also very important. For example, on the flanker variant you can increase the deflection of all of the control surfaces to 35 or so, and add forward-facing control surfaces that are set to affect pitch. That will allow you to turn 90% pretty much instantly. The downside of this is that you lose a huge percentage of your speed. The plane will also snap in two at low altitudes (below 5000) if you do this. The real Russian planes do this, but it leaves them sitting ducks for a bit while they pick up speed. Imagine a plane going from Mach 0.9 to Mach 0.5 in the span of a few seconds. It looks cool, but what practical purpose does it serve?

I've also played around with BahamutoD's thrust vectoring engine. It makes just about any plane supermaneuverable, but every turn bleeds speed.

[TheHengeProphet]

So, I tested out your USAF variant (had trouble downloading the other one), and that thing is ridiculous! Only weighs 10.5t to boot, which is pretty darn light, considering how much fuel it's carrying. I checked, and it looks like your stall angle is just around 30 degrees, which is pretty darn good. I noticed you had torque on and 70% thrust, so I turned off torque and replaced the engines with ones that I managed to finagle to 50% and save as a sub-assembly. Your wave-drag is under 1, which is nice, and the mass allocation helps it be quite stable (the AI still managed a flat spin...). The AI managed to repeatedly rip the thing apart, and I noticed that your mass value on the wings was at 1, so I bumped it up to 1.5, and it stopped disassembling. Once those adjustments were made, it was impressive to watch it outmaneuver and destroy everything else it could.

On the other hand, I appear to have built one of the most confounding craft FAR has ever seen. I get drastically different numbers from when I adjust something, then save and reload, and then the numbers go back to what they were before if I lower and raise the landing gear. I'm not at all sure which is the real number, and I don't think ferram4 wants to do anything about it considering he's going to do a wing overhaul in the next version of FAR. When I get home, I'm going to try a simplified wing design, and see if that helps.

Amusingly, if I adjust the wings to get the optimal lift/stall balance for their configuration, it shifts the center of lift so far forward that I can't get enough mass up there to cover it properly, making it hyper-maneuverable again, keeping the AI from handling it.

This is one of the reasons I like this challenge so much! It continues to press me to make this thing better, and when I do, someone comes out with something that makes me have to continue to redesign!

[CrisK]

So, I tested out your USAF variant (had trouble downloading the other one), and that thing is ridiculous! Only weighs 10.5t to boot, which is pretty darn light, considering how much fuel it's carrying. I checked, and it looks like your stall angle is just around 30 degrees, which is pretty darn good. I noticed you had torque on and 70% thrust, so I turned off torque and replaced the engines with ones that I managed to finagle to 50% and save as a sub-assembly. Your wave-drag is under 1, which is nice, and the mass allocation helps it be quite stable (the AI still managed a flat spin...). The AI managed to repeatedly rip the thing apart, and I noticed that your mass value on the wings was at 1, so I bumped it up to 1.5, and it stopped disassembling. Once those adjustments were made, it was impressive to watch it outmaneuver and destroy everything else it could.

On the other hand, I appear to have built one of the most confounding craft FAR has ever seen. I get drastically different numbers from when I adjust something, then save and reload, and then the numbers go back to what they were before if I lower and raise the landing gear. I'm not at all sure which is the real number, and I don't think ferram4 wants to do anything about it considering he's going to do a wing overhaul in the next version of FAR. When I get home, I'm going to try a simplified wing design, and see if that helps.

Amusingly, if I adjust the wings to get the optimal lift/stall balance for their configuration, it shifts the center of lift so far forward that I can't get enough mass up there to cover it properly, making it hyper-maneuverable again, keeping the AI from handling it.

This is one of the reasons I like this challenge so much! It continues to press me to make this thing better, and when I do, someone comes out with something that makes me have to continue to redesign!

I'm really glad that you had fun playing around with the USAF variant! I spent more time on that one as compared to the flanker. You're right that cranking up the wing strength dramatically improves the durability, but I'm not sure if that's considered cheating for this challenge. Even setting it to 1.25 is a huge improvement.

The AI rips the plane apart on my install too.

I actually re-built the body of the flanker this evening to lower the part count, clean up the lines, fix the z-fighting, and make it more durable. I sort of rushed the original body of the flanker - I built it on a whim after playing around with Halsfury's flanker. I'll update my older post and add the flanker craft file to this post.

Edit: Here's a zip file with the craft.

In terms of FAR giving you drastically different numbers - that has happened to me too. I think that it's a bug. I save scum to get around that. Vis-à-vis the lift bug: what FAR is doing is actually shifting the lift to the root of the plane. It's a weird bug and the only fix that I've found is to reload from an older save. It completely screws up the craft file.

PS: The flanker version is stable at 80% AOA and can turn on a dime.

Edited September 3, 2015 by CrisK

[TheHengeProphet]

Peculiar, the link takes me to the text in the file, where the link to the USAF allows me to download the file. Dropbox can be a pain sometimes...

Interesting, I'll have to see if I can get around this bug somehow.

Edit:

Yup, moving to a single-piece wing fixed much of the problem. For some reason, FAR will only stall out wing sections, so having a mono-piece cuts out that stall area. I wonder if he's going to fix that in the wing overhaul.

Tested out your Flanker style, and I must say I actually like it more than the USAF one. Cute how you hid that GAU-8 on the nose. The extra visibility out of the cockpit is really nice, even with that GAU in the way, haha. I noticed that you still have the trailing edges on those wings, which looks quite weird when rolling. I like to cut the trailing edge down to 0.08 and make it a rounded edge so I don't get that gap that forms when actuating surfaces. Not sure what's up with that drone core in the center, because it doesn't look like it provides much use other than its phantom torque.

I further modified it and reverted the intakes to their upright positions (due to lack of real effect, and for aesthetic purposes), and increased its overall ease of control by altering your control surface settings. It'll still stall out around 40°, but you're going to basically hit stall speed before you hit stall angle. When I'm done adjusting it, I'll put it back up here for you, if you like.

I'm going to go ask on the FAR thread which numbers to trust.

Oh, also, thanks for zipping that for me!

Alright, update: ferram4 says that the numbers you'll get after actuating landing gear open and closed again are more likely to be accurate than the ones given on load.

Edited September 3, 2015 by TheHengeProphet

[CrisK]

Peculiar, the link takes me to the text in the file, where the link to the USAF allows me to download the file. Dropbox can be a pain sometimes...

Interesting, I'll have to see if I can get around this bug somehow.

Edit:

Yup, moving to a single-piece wing fixed much of the problem. For some reason, FAR will only stall out wing sections, so having a mono-piece cuts out that stall area. I wonder if he's going to fix that in the wing overhaul.

Tested out your Flanker style, and I must say I actually like it more than the USAF one. Cute how you hid that GAU-8 on the nose. The extra visibility out of the cockpit is really nice, even with that GAU in the way, haha. I noticed that you still have the trailing edges on those wings, which looks quite weird when rolling. I like to cut the trailing edge down to 0.08 and make it a rounded edge so I don't get that gap that forms when actuating surfaces. Not sure what's up with that drone core in the center, because it doesn't look like it provides much use other than its phantom torque.

I further modified it and reverted the intakes to their upright positions (due to lack of real effect, and for aesthetic purposes), and increased its overall ease of control by altering your control surface settings. It'll still stall out around 40°, but you're going to basically hit stall speed before you hit stall angle. When I'm done adjusting it, I'll put it back up here for you, if you like.

I'm going to go ask on the FAR thread which numbers to trust.

Oh, also, thanks for zipping that for me!

Alright, update: ferram4 says that the numbers you'll get after actuating landing gear open and closed again are more likely to be accurate than the ones given on load.

Thanks! I updated the flanker's body slightly again. It has improved performance, and I fixed a weird bug where the IVA would have reduced visibility due to the engine nacelle bugging out.

I kept the trailing edge because it weirdly improves AOA performance. Not sure why.

The drone core in the center acts as a stored battery - it holds 250 electricity and never overheats. Regular batteries tend to explode at random.

If you reduce the control surfaces to around 30 or even 25 then it'll be easier to fly, but that extra control is needed above 10000m. I'd rather have it be difficult to fly below 5k than give up maneuverability over 10000m.

Edit: here are a couple of screenshots of the Flanker's updated body. As you can see, the lines are much cleaner now.

It's also really fun to fly IVA because of the increased viewing area. The GAU-8 has been shifted so that it's parallel with the horizon, and it's pretty fun to dogfight in IVA now.

Edited September 3, 2015 by CrisK

[FourGreenFields]

That looks pretty good.

Have you tried increasing the length of the leading edge? Should reduce drag without worsening handling if I am not mistaken (but haven't downloaded the craft, so haven't tried).

[TheHengeProphet]

I really like the new design. I find it much more enjoyable to fly than the previous iteration.

I adjusted the controls a bit, and blunted that trailing edge on your wing (improved the AoA by almost 5°). I also disabled the torque in the probe and cockpit, and reduced the engine power to 50%. I felt no particular need to adjust anything else on this version, so kudos on that!

Flanker HVar

I've been messing around with my plane a bit, and I feel I've improved it substantially! Here's a version I deem worthy to post:

MRF-4 Devil Ray

Now, I took the suggestion that a fighter capable of meeting all requirements would weigh over 10t anyways as a challenge.

I present the MRF-5 Spite! This puppy weighs 8.4t, and is capable of filling every requirement short of the required minimum mass. The required mass makes it nearly impossible to make a single-engine design that is at all usable. The cockpit is from QuizTech Aerospace. I have yet to arm it with a gun, but I'll enjoy seeing it dogfight once I do.

Javascript is disabled. View full album

Update on the spite: Something odd happened with FAR, and it's stated AoA values went crazy low, but the AoA values for everything else went up. I think it's time for a fresh install...

Everything is wonky now. Looks like whatever happened messed up everything, so... here, have this. MRF-5 Spite

Update: Added a couple more images to the Spite album to highlight the bizarre change in AoA behaviour. Interestingly, if I load the jet, it screws up all of my things, so I'm going to pull the file for now, so that link will currently take you nowhere.

Also, I made a super slick jet that is capable of high maneuverability at high speeds. It has a wave drag area under 0.3. Let me know if you want a copy.

Javascript is disabled. View full album

Edit: Here is this one: MRF-6 Spectre

Edit: Whatever this glitch is, it has tainted everything on my end. I've turned it into ferram4, but I'm struggling to figure out how to reproduce the glitch. If you folks have encountered anything similar, head over to the FAR thread and let's see if we can squash this.

The gunned version of the Spite weighs 9t even and still meets all of the requirements, save weight. I still haven't figured out this odd bug, but the plane still handles ridiculously well, despite stalling out at a much shallower angle than before (or rather, being able to stall out, period). I think I still have some work ahead of me to try to get the wave drag area down a bit further. It's currently just below 0.4 m², and I think I might be able to work it down closer to 0.3.

Edited September 4, 2015 by TheHengeProphet

[CrisK]

Great. Now my system is glitched too. I took a few screenshots and posted them in the FAR thread.

[TheHengeProphet]

Amusingly, only one plane has not been affected in my plane inventory: your flanker. I'm going to just proceed as normal, since it's all broken, and re-establish that link to the Spite. Oh, also, I did ask Quiznos323 if they'd make a cockpits only version of the pack, and it's being considered, but for now, the other parts ARE pretty cool.

I'm going to install the Kerbinside air races and see how quickly I can do them in the various planes. Should be fun.

I THINK I downloaded the Dev FAR, which would make sense, considering there was only one plane I could fly after I installed it that would not rip itself to shreds: the Spite... I thought that would have been the only one to actually do so, but it was fine with pulling 20g+ turns, while others would fly apart at under 15g, even with 2x mass strength on the wings. Considering whatever glitch is causing the odd behaviour in the graphs is part of the legacy wings, which is getting overhauled, I think I'm just going to go back to standard Haack, and hope for the best.

Edited September 6, 2015 by TheHengeProphet

[CrisK]

I built two more flanker designs, one large and one small. Here's a few shots of the small one flying:

Javascript is disabled. View full album

Here's the large one in front of the small one:

[TheHengeProphet]

Those are looking good! Did you use KerbPaint for changing the colours?

[Dwerto]

I've been working on a basic jet entry, and have been trying to do it with a single engine.

This is my best speed so far, not quite mach 2, but getting there.