I am beginning to understand Burt Rutan's approach. Basically, he lights the rocket and stands it on a 90° AoA (α), but it's not clear that SSO is actually hitting supersonic as it climbs out of the atmosphere. On the way back in, it is. That much is clear. However, it comes back down producing so much drag that it slows immediately to subsonic flight, the feather is "defeathered," and the plane resumes flight as any glider would. As the NASA orbiters have, in fact, for decades.
Airframe after airframe, I would get to the point with rocket motors (supersonic flight is a lot easier with turbofans, incidentally, because the progression is a lot slower and you can always just ease off. Rockets are not good with throttle.) where the craft went supersonic and all of a sudden my control surfaces were worthless. I had violent side-to-side shaking, up-and-down shaking, it seemed like it wanted to stall – and bear in mind this is a still accelerating rocket – and the only way I found to manage supersonic flight with control surfaces was to have at least two sets of horizontal surfaces (wings) as well as a set of rear stabilizers. I also borrowed the Su-37 concept of the ventral (cf. dorsal) vertical control surfaces, and threw a pair of them down there like the Russians have. When I let X-Plane do a lot of the automatic handling of the frantic roll and pitch (basically, I guess flaps and aielerons don't do so good at supersonic speeds, or don't work how I'd expect them to), and I developed a really light touch for pitch and rudder, the plane actually sorted itself out, and I survived a flight. But this is with front canards that almost counteract the main wing bodies, pitching the nose down because the rear, for some reason, wants to go down (I guess when you stomp on the gas maybe this is weight transfer or something similar?), while the flaps extend, and the lower (ventral) "vertical" control surfaces act as both elevons (which is weird, given their configuration and speed brakes.
And, in the end, you wind up with something that looks, not real surprisingly, a lot like something Burt Rutan would build. Lotsa wing. It would never really work in practice with a hybrid motor, of course. They're pretty all-or-nothing, so you can't say "Let's take off from DCA at 8% throttle, work our way up to 35,000 feet and then pour on the throttle until we're wide open at 55,000 feet." Having all that wing (those front canards have a wingspan of eight feet; the wings twenty-four) under full thrust down here where the atmosphere is treacherous would have ripped it to shreds.
As a matter of fact, I thought that's what X-Plane was trying to tell me when I finally got up to full throttle and the "transsonic" indicator started bleeping at me, crap, I've torn the bloody wings off the aircraft.
What really bugs me is that while I, today, sort of kind of learned how to handle supersonic control surfaces, I know I don't know it anywhere near enough to start teaching the Arduino about it, and that means that I can't put it in a rocket.
I'm thinking that the best design for a supersonic rocket (at this point it's looking like a nine-motor, three-stage, clustered E job in either CF or "blue tube" chassis) is to have very, very sensitive, and very clean canards up front – almost imperceptibly tiny ones even – and then very small slats on the ass end of the thing. Or just giving up alltogether until somebody like, say, an employer can buy software that will show me what supersonic airflows look like and I can just teach the software.
Until then, I'm terrorizing the district in phantasmagorical throttled hybrid rocket-powered many-winged Scaledesque craft. I had hoped to be further. I am beginning to wonder how hard it is to actually build a liquid fuel rocket, due to the vastly increased specific impulse and thus longer flight (and the ability to, you know, throttle).
