NOTE: The first half of this page was written about “Jujucee”, the previous incarnation of my “Jujubee” rocket. I’ll be rewriting this to apply to my new system, but for now I’ve left the information for historical purposes.

This page is about the Jujubee project, my cannon-launched rocket. The cannon makes it possible to add a theoretically unbounded amount of energy to the rocket without adding an ounce of weight to it. This is, notionally, one way to decrease the costs of sending things into space. A more futuristic version might use a long maglev accelerator instead of a pneumatic tube, but the idea is the same — get it moving as fast as you can, and then let the rocket take you the rest of the way. With my current setup the launcher alone can theoretically add the thrust of an E or F engine without adding an ounce of weight or costing a penny (since it’s just compressed air, after all).

“Jujucee” is the 4th rocket (and 5th launcher) I’ve built as part of this project. It uses a minimum-diameter 1.1″ phenolic airframe around a 29mm Dr Rocket motor. It can be flown on anything from a D to a small H motor. The launcher itself can be used reliably with anything from about 25 psi to about 100 psi. At full pressure and with a moderate rocket weight, the cannon in theoretically equivalent to an F motor, with an average thrust comparable to many Ms out there (over 1000 Newtons), albeit for less than a tenth of a second.

Launcher
Jujucee and launcher prior to painting
The launcher is constructed out of PVC. I used GGDT (Dave Hall’s Gas Gun Design Toolkit) to model the internal ballistics. That allowed me to use RockSim’s EngEdit program to create a model of a rocket engine that corresponds to the cannon (given specific parameters, such as the pressure of the cannon and the mass of the rocket). This model then allowed me to use RockSim to model the cannon-to-rocket launch as a two-stage rocket launch, giving estimates of speeds and altitudes.

The current launcher stands about 8′ tall, and at full pressure can fire a tennis ball well over 500 feet per second (simulated). In the picture, the black thing at the top is a homemade chronograph I built. I haven’t yet had time to really mess with it though.

Rocket
The rocket is a minimum-diameter phenolic airframe around a 29mm rocket engine. It can take motor reloads from a D up to an H (the motor casing I have only goes up to 180Ns though, so they’re small Hs). On the H238 and with the cannon at a decent pressure it should scrape Mach. One day I’ll have to test it that high, but right now the weather is too cold to pressurize the cannon too high.

Realizing that the rocket would have to take some very intense accelerations (up to about 300 gees) as well as very high velocities (possibly Mach), I tried to make the rocket as sturdy as I possibly could. To this end I decided not to use any couplers or allow any significant breaks in the airframe, since those are the sorts of things that would break first. So how do I get the parachute out? Well a small black powder charge is fired at apogee (using a G-Wiz MC2 altimeter), which pushes out the parachute and rocket motor. I static tested it at Phil’s place using about 1/4 of a gram of powder and it was perfect, the rocket motor ejected out the back and all three 15″ thin-mill parachutes unfurled nicely. I use 3 small parachutes because, due to the very narrow airframe, a single larger parachute simply wouldn’t fold up small enough. A thin Kevlar microbraided shockcord connects the rocket to the motor and parachutes ensuring that everything stays in one piece.

Construction

Launcher

Altimeter Bay
The altimeter is a PerfectFlite miniAlt/WD. This got me my Level 2. It is a barometric-only altimeter. Why not use an accelerometer-based altimeter, you ask? Two reasons: 1) the one I tried crapped out on me, and 2) the accelerations encountered inside the cannon are too much for any hobby-level altimeter I’ve seen out there, so the acceleration readings would be garbage anyway.

Open. Bulkhead/couplers are machined from solid PVC (dark gray). The top bulkhead/coupler (seen here at right) has a slot machined to snugly fit a 9v battery. The light gray pieces machined from PVC serve as anchor points for the thin Kevlar microbraided shock cord (seen on the right in yellow), and the lower piece also doubles as a shelf to support the ejection charge (to prevent it from tearing out under the high acceleration of the cannon). Wing nuts connect the electric matches to the altimeter while also holding everything together solidly. The 4-40 allthread is actually made in 4 pieces, and pairs of them are joined by the white threaded couplers. Finally, the 1/8″ thick ring of PVC that is exposed between the airframe and the altimeter bay is tapped with 10-24 threads: when the screw is removed, it releases a lever which in turn provides power to the altimeter, arming it.

Closed. Connects the top and bottom airframe tubes.

Launch Detection
This part is in progress. The plan is to have an AVR microcontroller connected via RF link to a ground-based control box. The ground station sends a confirmation code to arm. Once armed it waits until it detects a large acceleration roughly matching the cannon’s profile. Then a short timer is started, and when it runs out, the rocket engine is ignited. The optimal time delay for speed and altitude is actually zero, but I really want to see the smoke trail start mid-air.

Airframe
The airframe is 29mm phenolic. I’ve been using my little Taig CNC mill along with some small 1/16″ roughers to route the fin slots and cut the tube sections square.

Fins
Fin material is 1/16″ double-sided copper-clad board, for making printed circuitboards. It’s made of G10 fiberglass (a common fin material), but it’s laminated with copper on both sides, making them incredibly stiff but also tough. I chose a clipped delta design because they’re streamlined but very strong, with no sharp points to break during launch or recovery. The design itself was arrived at through trial and error in RockSim — basically I started with the smallest fins I could, and tweaked the shape until I found something tiny that still pulled the CP back a lot for its size. Note this was using the RockSim algorithms rather than Barrowman. Barrowman predicts the rocket to be hopelessly unstable. I guess we’ll soon find out!

Note, since the design of the airframe is so simple and since making extra fins is simply a matter of throwing a piece of material under the mill and hitting a few buttons, I’ve decided I want to try and make multiple fin sections. This way I can change the length to accomodate different engines, and I can also add different numbers of fins in order to optimize stability. According to the simulations this makes a very big difference.

The simple g-code I used to cut each fin:

g90

#100 = [0.0394/2] (tool radius -- set this yourself)
#200 = 3 (default feed)
#210 = 1 (plunge feed)
#300 = 0.05 (safe Z)
#310 = -0.1 (cut depth)

g00 #300 (go to safe z)
g00 x-#100 y-#100 (move to starting point)
g01 f #210 z#310 (plunge)

g01 f #200 x[0.1-#100]
y[-0.12 - #100]
x[1.1 + #100]
y-#100
x[1.25 + #100]
x[1 + #100] y0.565
y[0.565 + #100]
x[0.5 - #100]
x-#100 y-#100

g00 z#300

m30

Recovery
Currently the plan is to come down drogueless at apogee, and pop three 12″ chutes at low altitude. The arming box will hopefully also be able to help locate the rocket using the radio with a directional antenna.

Arming Box
I may not get time to finish this part, but I’m gonna give it a shot. This is to include character LCD display, a radio transceiver, a keypad, and some switches, LEDs, and other indicators. Ideally a “launch code” will be entered into the keypad and relayed wirelessly to the rocket. If the rocket’s launch detection circuit determines that the code is correct, the rocket is armed. The arming box will also have override capabilities, including the ability to manually ignite the rocket engine.

RockSim
Here is a link to the RockSim data file I’ve been working on. This only simulates the rocket part (no gas gun), and you’ll have to manually optimize the airframe lengths and fin count in order to achieve good balance for whichever motor you use.