FAR/RP-0 Plane: Sonic-X1, aka "how to go supersonic after over 9000 attempts"

[SyzygyΣE]

After the not-so-good performance of an earlier plane—the Sequoia, I decided against patching it up and instead chose to retire it from service. Several helpful users explained to me what features I had overlooked in that vehicle, and now, I opted to create something smaller, simpler and lighter, hopefully able to carry the first Kerbal beyond Mach 1. After an uncountable number of tests and tweaking, I present forward the Sonic-X1—the first of my supersonic aircraft family. It is by no means the best, and there are likely still some areas I could further enhance. Given its stable performance so far I am quite satisfied with the result. But, if I find methods to improve its capabilities, I am sure to implement them.

If you have any suggestions as to how to improve on this design or designs for the future, please leave a message. I will appreciate any input.

[chadgaskerman]

Good craft reminiscent of the Bell x 1 with the cockpit, also

Edited July 14, 2016 by chadgaskerman
Beware the grammar police

[Van Disaster]

Well done only thing I can see is that when you're supersonic you might be using a bit too much pitch control deflection ( I'm guessing based on Mw from the Mach 1 derivatives ) to keep level, so either move fuel back ( next time use multiple tanks ) or shift CoL forwards. Centre of Lift moves rearwards as part of the process of going supersonic, the longer your wing chord is the more it moves.

[SyzygyΣE]

7 minutes ago, chadgaskerman said:

Good craft reminiscent of the Bell x 1 with the cockpit, also

With the RP-0 install I have, I can actually create the real Bell X-1. There are components to do so. But the parts designed for it isn't as efficient, cheap or light as procedural parts, so I too decided against it. It never hurts to innovate a design that differs from real life counterparts. 

12 minutes ago, Van Disaster said:

Well done only thing I can see is that when you're supersonic you might be using a bit too much pitch control deflection ( I'm guessing based on Mw from the Mach 1 derivatives ) to keep level, so either move fuel back ( next time use multiple tanks ) or shift CoL forwards. Centre of Lift moves rearwards as part of the process of going supersonic, the longer your wing chord is the more it moves.

By multiple tanks, do you mean several fuselages on the body of the plane or the fuel distributed in the wings? I actually don't put fuel on the body of the plane; I use a procedural structure that doesn't actually hold fuel. The fuel I store in the wings. Is that how its supposed to be done? I wonder how you can judge if the deflection is not enough or too much from just scanning over the derivatives. Is it just experience? I wish I knew what each of those values meant. Mousing over it gives some explanation but I'm not sure I understand it. For me, it's just the old "green = good, red = bad." 

And also, the CoL moves? I never knew that. I always thought the CoM and Dry CoM were the ones to pay attention to. As a result I have always abided by the rule "CoL slightly behind the CoM for stability." So that means it's ok to have it on the CoM or slightly ahead of it if I'm aiming for supersonic speeds?

[Van Disaster]

If you can keep it stable, you can have the CoL wherever you like using wet wings is fine, you just might want a little ballast tank in the craft somewhere you can pump some fuel to trim. On the other hand if you're not finding the craft flying around supersonic with the stick pulled back just to keep the nose up, no worries.

CoL moves even subsonic, it'll gradually creep backwards until nearly supersonic, and then there's a funny region where it all goes a bit strange & you might find CoL moving forwards. Once you're supersonic, thanks to the process of supersonic lift you'll find it's moved quite a long way aft. Beware the CoL ball in FAR, it's only really any use for getting things roughly in place.

Most of the derivatives you don't really need to worry about at this point; Mw is important because it's an indicator of pitch stability, like the yellow line on the graph. If it's red it means that any change in pitch will increase the pitch force in that direction - ie you're naturally unstable in pitch and you're likely to backflip without anything correcting it. Not terminal, it's one thing the AoA on the FAR flight control setup is for, but not a good thing if you don't know how to manage it. If it's green and fairly large it means the craft is trying very hard to fly prograde ( to borrow a space term ) which isn't much use if you need some AoA to stay in the air - so you'll wind on a lot of pitch deflection & cause drag. There's some info on the little Wiki on Ferram's github.

[SyzygyΣE]

Oh right, I see. So it's about being stable but not being stable to the point where the plane doesn't even want to move from prograde.

While you're here, there is something else about wing design that leaves me confused. So, in B9 Procedural Wings, which I am using to create a wide variety of wing shapes, you can set the thickness of the wing at a particular cross section, as well as the length and thickness of the leading and trailing edges. I always turn the length of the edges down a little and make my wings thick at the root and thin at the tip. Given your expertise in this subject, what does modifying these parameters affect? It only seems to ever so slightly shift the CoL as I tweak them and makes no significant difference when flying so I'm not too sure what principles I should follow when deciding on these characteristics of a wing. There is an image on the mod thread that illustrates what I'm suggesting at.

Edited July 14, 2016 by SyzygyΣE

[Van Disaster]

Wing cross-section affects drag and mass, and I think from vague memory of poking my nose in the plugin code, joint strength. It should also affect fuel capacity ( obviously, as you're changing volume ) but there's been some funny issues with RF/MFT & those, so if you notice any fuel & mass quirks please drop a note in the thread. Also you might want to try the shader I patched up a bit - I passed it over to the maintainer but he's been busy elsewhere ( I should have done it properly via git if it would ever let me... ). I generally just have equal thickness wings, usually as thin as required tankage & structural considerations will allow. I use KJR which autostruts everything so joint strength is rather less of an issue.

As for CoL/CoM balance, how I usually get people to picture it is:

* The craft is a rod.
* There's a fulcrum under CoM
* There's a spring ( adjustable force ) attached to the craft at CoL, pulling up
* There's adjustable force springs pulling up and down through the CoL of the pitch controls

Now you can picture what forces have to be in balance for the craft to stay level. If you go dig up a basic physics book or some school reference on mechanics, you'll find that's pretty much how people teach Moments ( well, they might use weights instead of springs but it's the same principle ) - and everything can be taken back to your craft. The pitch controls are applying torque; you can see that the torque is the distance from CoM * the lift the surface is producing, remembering lift is proportional to area * airspeed * AoA, so the further away from CoM the pitch controls are the smaller they can be. CoL ( the sum of static lift ) is also producing torque in exactly the same way. If you ever get into building aerobatic aircraft you'll want to consider moments of inertia ( which is just how the forces respond to the craft rotating ) but that's not something to generally worry about.

Why is CoL behind CoM naturally stable? because the craft has to fly at an angle to the air to keep altitude, and you can see from the forces that what it really wants to do is put it's nose down, which is fine - nose down means more airspeed, which means more lift. If the nose starts rising then lift increases, but lift is all on the same side of the fulcrum so torque will rotate the nose down - and as airspeed builds up so does the downward force on the tail, so there is a negative feedback loop. If CoL is in front of CoM, then an increase in AoA means increase in torque on either side - and if CoL is too far in front then the torque from the tail won't rise as fast as the torque from CoL, and you lose speed and stall ( or even just flip ) unless your dynamic lift ( control surfaces, airbrakes ) or something odd like body lift ( if you happen to be flying something much wider at the rear ) counteracts it. If you get the CoL/CoM balance right you need very little force to cause a change in direction - you just give it an initial push, the feedback loop keeps it turning and all you have to do is control the feedback loop - which is why combat aircraft are generally dynamically unstable.

If you think about canards, you'll go "hey, they aren't naturally stable either", and you're right - it's just for fixed canards most of the lift is still behind CoM. What you don't get is the tailplane increasing in lift at higher AoA & pushing the nose down - remember it's on a long lever arm - instead you get the main wings on a short lever arm trying to push the nose down, vs the little canards - which don't forget are your control surfaces - trying to push the nose *up* as AoA goes up. You want either very small canards, which means your CoL has to be pretty close to CoM because there's not much force from them, your canards to be at a bigger angle of incidence to the main wing so they stall first - this does mean you literally can't fully stall your plane but stalled surfaces add *massive drag* - your control setup to actively compensate which is how I prefer it, or I think you can do something with wing shape I've forgotten about at the moment.

Now move where the CoL spring is attached & look at the torque forces again. To get back to equilibrium you either need to apply more spring force at the pitch controls ( I guess you could use reaction wheels if you have stock ones, I hate doing that though ), or move CoM, and pretty much the only way of doing that is pumping fuel around. That leads on to a little trick - instead of applying a constant force through the pitch controls to stay in normal equilibrium, you can just angle the tail down ( or forward fixed canards up ) by a much smaller amount because they're usually much bigger area than the control surfaces. Now you've built in torque to counteract the nose-down torque from CoL being offset. The other thing to watch - and this has caused IRL crashes - is engines being offset from CoM, because they'll produce a pitching torque too.

Finally for that stuff, CoM and CoL will quite often move relative to each other as you pitch up or down - especially if one is offset vertically from the other. High AoA will cause a lot of body lift which isn't usually apparent and that might move CoL too. If you go back to the rod and attach a spring at each end: if CoM is in the middle then there's equal length rod on each side of the fulcrum, so if you apply equal force in the body lift springs you'll get equal torque each side & thus equilibrium. If CoM is towards the rear then when you apply equal force to the body lift springs, the one furthest from CoM - the front - will apply more torque, because once again it's force * distance. That means there's additional lift from the nose, so CoL is going to move forwards, and you might be in trouble. It's equally true if you're pitching down and quite importantly equally true if the nose drifts off the direction of travel - can be a big problem for shuttle-style spaceplanes with long bodies & all the engine mass at the rear.

Er, wall of text, oops tl;dr, dig out a physics book & read about Moments again. Lift devices are the forces, distance of lift devices from CoM is the lever. Normal body parts at non-zero AoA become lift devices.

Edited July 14, 2016 by Van Disaster

[SyzygyΣE]

Haha, thanks for taking your time to write this. I'll try and do some research. If there is still something I don't understand, I'll leave you another message. So, if wing thickness doesn't have much of an effect except on fuel capacity and joint strength, does the same apply for control surface length? With a longer trailing edge on control surfaces I'd expect them to simply provide mode force—so say, if you want to pitch, the longer the control surface, the greater the effect on pitching?

Edited July 15, 2016 by SyzygyΣE

[Van Disaster]

Control surfaces are just little wings, you change their AoA and their lift force changes just like changes in craft AoA affect the main wings - so they obey all the same laws, more area means more lift & drag. More lift means more torque if the surface hinge hasn't moved relative to CoM.

Wing thickness affects cross section area, don't forget.

[SyzygyΣE]

@Van Disaster

Understood. I'll keep that in mind. So, while I was designing this plane I sought to utilise the flaps and spoilers, but, for some reason activating either seemed to create a huge amount of instability. With spoilers, they would just suddenly pitch my nose up by over 20 degrees, often sending me into a stall, and for flaps, they would just push my nose down by over 20 degrees, sending me into an unrecoverable dive. Aren't flaps meant to increase lift? So why are they pushing my nose down? Lift = nose up? I think I have a misconception of some kind.

On another note, is it possible to use slats in FAR? I think I best try to understand how flaps and spoilers work before attempting that, though. I think I know how the two of them work but when actually trying to use them and put theory into practice, I end up confused.

Edited July 15, 2016 by SyzygyΣE

[Van Disaster]

Look at the angle of the flaps, and where they are. Flaps in KSP/FAR add lift by changing AoA - now read my previous post ( or both ) again. Spoilers also change the AoA of a surface, but this time the aim is to increase drag rather than change the amount of lift - but you're still just changing the AoA of a surface.

You can't do slats. You can do drooped leading edges, they're just controls facing forwards - and with current wing code if you can stop them from stalling you will stop the main wing behind from stalling ( the wing can still stall from the back forwards, so you don't stop it *ever* stalling ). There's no fancy blown flaps or Fowler flaps either. Note that Ferram's been working on completely new wing code for a while.