So it's the SYLVIE experiment?
Like Servo-controlled, Year-Long Validated, Interesting Experiment or something?
Paris 'cos Sylvie isn't available as an icon.
We invite fans of our Low Orbit Helium Assisted Navigator (LOHAN) project to raise a glass or two today to UAV specialist 3D Robotics, which has very kindly agreed to offer our Vulture 2 spaceplane a brain transplant in the form of the mighty Pixhawk autopilot. Two views of the 3D robotics Pixhawk As LOHAN regulars know, …
You read story after story of some Big Bad Corp or Corp v Corp sueball tournament and then suddenly as a bolt from the blue there is this bunch of enthusiasts and real world companies willing to put reputations (and money) on the line for pure escapist fanatasy projects.
Gawd bless 'em one and all
It looks like the autopilot will need 2 stages ~
Stage one, the truss swings and launches in a vertical - ish direction, the autopilot needs to either be nutral or adjust to a vertical trajectory for the duration of the burn plus a little get max altitude before ~
Stage 2, Normal horizontal flight.
Its actually a little more complicated, no doubt someone will comment about control surface throws for high speed thin air flight relative to normal thick air flight, relative to slowing down for a soft landing.
The way we're currently looking at it is:
1) Autopilot in "neutral" until after rocket motor burn
2) High-speed dive at high altitude (aka "plummet")
3) Low-speed glide to landing
Part 1) is actually the difficult bit. We need to work out how the autopilot is going to sense launch.
Ok, this might sound a bit stupid, but why doesn't the autopilot *initiate* the rocket burn?
This would cut down on the number of electrical connections from the launcher to the plane and make "detection of launch" easier.
If you need to communicate between the launcher box and the plane, for example to initiate launch early, you could do it via infrared (IR-transistor on the recieving end, transistor+resistor+IR diode on the sender; easy as PI).
If you really just want to passively detect the rocket ignition, looking for a sudden, 3 second forward acceleration followed by a rather steep temperature increase of the rocket casing and a rather steep fall in altitude[*] should be a good indication.
[*] If you detect a rather steep fall *followed* by a sudden deceleration it means to waited a bit to long.