Surefoot

Members
  • Content Count

    367
  • Joined

  • Last visited

Posts posted by Surefoot


  1. First make sure your control surfaces, and wing shape/surface are similar to the real thing:

    a5431318-105-Su27%20Top%20View%20-%20Air

    Also make sure you try and keep the proportions right (and thus the CoL vs CoM) because that's a key element. Also note how the wing root extensions are done and merging with the fuselage, and how they cover the intakes (which are not too far behind the cockpit actually).

    I know first hand it's a lot of hours of work doing all that fine detail, but Sukhoi are into fine details that are kinda important to aerodynamics...

    Once you have a really close model, you'll find out it's completely unstable and will flip over immediately. You need the AtmosphereAutopilot mod that simulates the fly by wire system of modern jets. Just set the limits in AA to decent realistic values, and you'll have an SU 27 that flies almost like the real thing :)

    (edit) from what i see in DCS (and i suppose it's like the real one) the edge slats are working in counter-AoA mode.


  2. 12 hours ago, AeroGav said:

    I'm hoping this career game produces more realistic looking aircraft than my efforts in stock aero.     

    It does mostly, and if you stick to (proven) realistic designs they usually work very well. That's the main reason why i love FAR so much and spent hundreds of hours tinkering with aerodynamic designs. It's like "build-your-own-X-Plane" simulator :P Fighter jets usually fly pretty much like in the best flight sims out there, which is amazing.


  3. They seem to work, at least the flight model of my KU-57 seems realistic enough, although i have limited elements of comparison (SU-33 simulation in DCS or other flight sims). I know the SU-33 (it's basically an SU-27 with canards) is not the same design as SU-57 but it takes from the same bag of tricks and is also naturally unstable. On my flight tests with FAR i notice the leading edge AoA slats will definitely improve handling at high AoA and delay stall. With thrust vectoring i can even throw the plane around at such high G's i'll just disintegrate the airframe from the stress. Putting a G limiter on AA makes it fly mostly like an agile SU-27. The real world SU-27 has also very wide flight characteristics and is able to pull damaging G's if you override the joystick limiter, to pull aerobatic figures like the "cobra".

     

    Quote

    Seems like the usual trick is to set them up to -AoA% in the right click menu(so they would be at a smaller AoA compared to the surface behind)

    Yeah that turns out to be similar to how they work on real models, you can watch them in DCS for example on SU-27 you'll see the leading edge slats countering AoA. Same goes for the F-15 "side pods" extensions, they do move counter to AoA (although in a very limited manner). And yeah on that F-22 photo you can see them clearly too.

    (edit) on that very nice SU-57 photo they are also quite obvious:

    3c129bfa4d22d4dd1867a9d1241363a8_0.jpg?i


  4. 4 hours ago, Schwarz said:

    It does feel a bit like a yaw control issue, if lucky I might be able to solve the problem right there :)
    Quick question: Should elevators be excluded from yaw control?

    Yes, if possible you should have elevators near the center, and flaperons as far as possible on each side (see how the B-2 achieves it). It's not the guarantee of a perfectly stable aircraft though, if your CoL shifts too much when opening the bay (and that pressure curve doesnt look good...) that might be another cause for instability. I'd try to add more yaw authority first, in any case.


  5. Others will comment better than i do, but i give a few hints: I see an aera ruling issue here, coupled with a distinct lack of vertical surfaces. You should first try and reduce the area behind the bay, or give extra width to the area around the bay (or move it backwards, that would reduce the impact). See that yellow line, it should be as straight as possible.

    Then you seem to lack control surfaces, are you using flaperons for yaw control ? If not, try either adding control surfaces used as assymetric airbrakes for yaw control (there's a specific option given by FAR to allow this) or increase substantially your vertical tail fins, current ones are way too small, and way too tilted so they'll produce pitch/roll coupled with the yaw and have a very weak yaw lever.

    (edit) also where is your CoM in relation with your CoL, what's happening probably is open bay doors add a hint of lift and move the CoL forward which would bring more instability.


  6. The problem with canards and forward swept wings, is the bad re-entry heat/mechanical resistance. Canards will take a lot of heat being out of the shock cone, and fw swept wings will take a huge load out of the slightest AoA bump. I'd rather stick to standard tailless delta, even though these tend to have lawn dart behaviour...


  7. I used the diagrams i posted above, which i believe represent the real F-15. Thing is the MK2 body is a bit wider, otherwise i tried to respect proportions, also you might be induced into thinking that because of the stubbier nose (nothing i can do there, i have to do with the MK2 parts i have). Like i said, it's quickly slapped together, not up to the level of detail of my KU-57. It still flies rather well and has fighter-like agility.


  8. If you want to do a F-15 that works you have to pay attention to all the aerodynamic details (wing surface ratio, thickness, anhedral, sweep, etc.), all proper control surfaces sized like the real one, and the CoM should be really near the CoL almost on top of it. Fighter jets are very twitchy by nature. Just look at the KU-57 example i posted on previous page (page 47) here, it flies very well but i had to go into a lot of details to be close to the original.

    If i look at that blueprint:

    macdonnell-douglas-f-15c-eagle.png

    I can see that the wings have a specific shape, slight anhedral, are top mounted, and pitch ailerons are quite offset to the rear possibly for increasing their authority.  The wing root starts with a weird oblong shape that's on the side of each intake, i suppose that's for area ruling (which is also a thing in FAR so we have to pay attention to it), i also notice the "thin waist" and on photos of the real thing the rounded shapes for each engine pod. Notice each control surface size and shape, and the big fixed twin vertical tails. Taper is also taken into account by FAR somehow and it's noticeable on that blueprint.

    The F-15 has no leading edge slat (notice the difference between control surfaces and leading edge of each wing) so it should be a relatively straightforward design to do in KSP. Intake pods cannot be done exactly like on the F-15 but we can do something close.

    (edit) slapped something together very quickly:

    A3AACA89DE9EA8238B4CC6183013E9D3D6EB6567

    Yep, it works :) It's not as unstable or twitchy as my KU-57 but flies still very nice.

    D33C2CD9A969953191554C0F528EFD3F48BC0D47

    Of course there's no double cockpit length wise in KSP so the front is a bit shorter here.

    EEE1ABCB3AAD0B2FB9CFFED50A50F8C0CC55C5F0

    I think i got the system to make that F-15 apart from the intakes i'm not happy with what i did there. I also could elongate the cockpit by adding a fuselage section behind it. But that was a quick job :) and the CoM is really well where it is now.

    So yeah, it flies, it rolls pretty well and has enough pitch authority to destroy itself from G forces and incoming airflow. Also it's a stable design, so SAS is enough to fly it.


  9. On 8/14/2017 at 9:46 PM, DaMichel said:

    Few days ago I watched a documentary on the Pak-FA fighter. And soon I felt like building planes in KSP again ... :P

    If you want that one (now called KU-57 :)) i can send you the .craft file

    CD2AA73CE2CF4BBE537E07D3E04EF7199C1A37DD

    Flies very well, definitely needs Atmosphere Autopilot  as it's an unstable design, and of course it's made with FAR in mind.

    Love your mod :) Simple, clean, effective, makes flying planes and spaceplanes a lot more pleasing.


  10. The Rafale is also built for high AoA, so far in combat exercices it's superior to most alternatives, be it BVR or dogfight. The Russians may be thinking in similar ways, that once the BVR exchange of formalities is done, it's high G time, and there whoever can manoeuver the best has a much easier time. If you play a bit on DCS World you can see the strengths of SU-27 vs F-15 for example, it's quite interesting.


  11. 1 hour ago, Van Disaster said:

     Not sure what else the variable leading edge next to the inlet on the SU-57 actually does ( presumably vortex generation similar to close-coupled canards is one thing, but Sukhoi are far more clever than I am... ).

    The amount of trickery on Sukhoi planes is insane :P The shapes they use, the sleek profile of engine nacelles (thanks to clever gear box positioning... specific to russian jets), the small aerodynamic details on every surface.. I had a lot of fun reading litterature on the SU-27. They are on top of the game in terms of aerodynamics as of today, IMHO.

    At least those inlet "canards" on SU-57 from what i read on various military publications are pure AoA coupled, where i'm a bit fuzzier is the wing edge slats and the inner control surfaces which i've assigned to flaps, given their positioning not far from CoM, and the slats went to AoA but that's pure guesswork here. In any case with FAR this is a beast in flight, it doesnt need so much control authority to be thrown around in high G's and does your exact bidding from any flight mode, and this for a quite heavy plane (we're talking MK2 base here, not a light MK1). Makes me think the real thing will be quite fearsome since they also seemed to improve a lot upon SU-27 family avionics.


  12. On 10/12/2017 at 2:11 AM, Van Disaster said:

    Shadowing would be a reason to have the tail *below* the wing - otherwise you end up with conditions like deep stall. Somewhat awkward to arrange, sure ( although a lot of combat aircraft have quite high wings & do this - obvious examples, Tornado, F-15 ) but it'll only get blanketed at high negative AoA ( when are you going to do that? ) and the higher the AoA, the less likely it is to be in turbulent air.

    I suppose that's mostly due to air combat manoeuvers involving high positive G's and almost never negative G's which are way more difficult to counter for the pilot (the pressure suit will work on legs, but not on the head). So the aircraft will be mostly encountering positive AoA situations, makes sense to have this high wing / low tail design, also structurally speaking it makes it easier to design for resistance against high positive G's.

    Also on newer designs (such as the SU-57 i copied here) they use moving slats or canards to keep the wing from stalling at high AoA. Look how the horizontal tail surfaces are in line with the main wing on the SU-57 for example, i think the front slats (and those "canards" above intakes) allow for this to happen. The Rafale is using canards for the same reason.

    Well at least on the KU-57 it works :) It can manoeuver fine with 90° AoA and recover from a stall without any issue. I set up the slats/canards in counter-AoA mode, i dont know exactly how they are set up on the real thing but on KSP+FAR they do their job.

     

    (edit) VTOL done :)

    E97D327B10CEEE13201836C5E620408F8ECAD256

     

    About to land on top of KSC

    0E1F92B0851959C2A619520483E88B33917FCEB2

     

    Mach 1.7 low altitude flying

    F6D38BEA059377B8797B6FF2C784C27935F897E8

    Of course i'm using TCA :) (edit: used to have trouble with horizontal flight transition, looks like i missed i can assign an action group to both TCA on/off and engines toggle, also 'cruise' mode seems to be the right one for controlling main engines..)

    That aircraft flies nice, is mostly stable actually, i added front slats so i can do high AoA stuff with the vector thrust engines... it's not as nimble as the KU-57 but it will do well for tasks needing a VTOL !


  13. Yeah i have 700 hours of FAR KSP under my belt, i used to nail tailless delta SSTOs but i seem to have lost my mojo :D I stopped playing for about a year and just started again, exactly thanks to Atmosphere Autopilot which allows exploring new flight characteristics. I do have enough pitch authority on those delta wings, but the plane wants to fly at a ridiculous 20°AoA where my usual working designs were between 2° and 4° under nominal conditions (flying straight on, cruise altitude). Probably CoM issues, or wing loading, or subtle angles i missed. I'm still fighting with the editor to get correct wing profiles, on that fighter it was "easy" because wings are mostly flat (well they are polyhedral but cross section is flat), whereas on a delta we should put some camber somewhere in there, i have to remember my bag of tricks :wink:

    Some Kronal Vessel Viewer views:

    Huk18A1.png   auqBxtH.png

    That one has a near perfect CoM / CoL alignment (with the CoM a few centimeters ahead), i'll balance for empty tanks later but it has plenty of room for tweaking.

    (edit) oh, missed that:

    Quote

    In a tail-less design, those control surfaces are in-line with the wing, which means there's a greater tendency for the two force vectors to line-up

    It's not the issue, it's about having a lever against the CoM and CoL so the greater the distance between your elevon / elevator and the pivot axis, the greater the moment (as in rotational force) it will have. The real issue of having a tailless design and why in many designs tails are elevated is shadowing, as in the wing in front will separate the airflow so as the elevator / elevon will not get any (or just turbulence) and thus will lose its acting power. Delta wings are supposed to mitigate that effect, but given the CoM is rearwards (due to the nature of the design itself) you need bigger surfaces to get the same force.


  14. I'm having a lot of fun with the new version of FAR and that AA mod (fly by wire) that allows unstable designs to be flown !

    I'm still strugging with delta / elevon designs (cant get a decent level flight AoA for some reason, i'll look into that later) but that unstable fighter design is a total success:

    CD2AA73CE2CF4BBE537E07D3E04EF7199C1A37DD

    I named it KU-57.

    A8AE6DF8FBE9E0DA0DE7166163373CC9C986EC57

    It flies beautifully, it's a bit twitchy on roll (should tweak deflections here and there), can manoeuver at high AoA / low speed.

    3C4F229A43E769E6DD614AFC6AFB90C48DAF27DB

    Took me about 5 hours of work, the original design is full of tricks, what with the anhedral lifting body and leading edge AoA elevons. The wing design is very intricate and everything was done using B9 procedural wings.

    I doubt this can fly with stock SAS, it wont react fast enough.

    This will be my kerbin atmo research platform, next project is a VTOL...