Model rocket reaction wheels

[rpayne88]

I've had a Quest Zenith two stage model rocket sitting in my basement for the past three years. I have yet to launch it.

Problem: I don't have any first stage motors (the ones with zero delay before firing the ejection charge.)

Solution: Make a home made reaction wheel.

Question: How can I calculate how fast I need to spin the wheel given a certain mass.

What I'm looking for is an equation to calculate it given speed, mass, and diameter.

I know I could go out and buy a 1st stage motor, but that is much less of a challenge.

[Nuke]

use a gyro for feedback and then use a pid controller to control the wheel(s). you will probibly need a motor controller too. you will also need a microcontroller to run the show. it will get data from the gyro, feed it into your pid loop, and command the motors appropriately. you dont need a lot of io for your micro. most gyros are i2c, and thats 2 wires+2 resistors. your typical h bridge driver will use a pwm pin and 2 other pins per channel. so for a 2 axis reaction wheel system you will need 8 mcu pins or 11 for a 3 axis system. an attiny84 should be up to the job. an sn754410 is a good 2 channel motor controller you dont need that much current since the wheels will never approach a stall state. then just use any mems gyro. get one on a breakout board because they are really hard to solder (this is also convenient because a lot of them can be bolted to a hard point in the rocket and attached to the rest of the electronics with a few wires). for power just use a 3.3v regulator like a lm1117 and a small 2 cell lipo battery.

to do this the easy way just get one of the small form factor arduinos, like the nano, then get some strip board and jumper wire and solder your motor controllers to it and some female headers to connect it to the arduino. you still need the pullup resistors for the i2c bus, 1.5k usually is fine (you may be able to use the micro's internal pullups, idk), then make a header to match your gyro board. you could make it small enough to fit in a tp tube.

Edited March 10, 2014 by Nuke

[Camacha]

I am unsure how reaction wheels are going to replace a motor without delay. Could you explain that?

[rpayne88]

I am unsure how reaction wheels are going to replace a motor without delay. Could you explain that?

The rocket has a payload bay. The plan is to place the rocket on the pad, turn on the reaction wheel, replace the nose cone, and launch. It will be spooled up by launch and won't even be needed for a few seconds afterwards when the 1st stage burns out.

The first few seconds would be stabilized by the fins. The reaction wheel is only needed for the four seconds between 1st stage MECO and 2nd stage ignition.

The idea is to keep the rocket pointing straight up during that time. What I want to know is how to calculate the requirements (for lack of a better word) of the reaction wheel.

[K^2]

The rocket should be coasting at good speed at that time, allowing aerodynamic stabilization. If it's not, then you've messed something up in terms of engines you've selected for your rocket.

[Camacha]

The rocket should be coasting at good speed at that time, allowing aerodynamic stabilization. If it's not, then you've messed something up in terms of engines you've selected for your rocket.

That is what I would say. If the rocket slows down so much in those few seconds that it loses aerodynamic stability, something is probably wrong. Although I do read that it is advised to ignite the second stage as soon as possible to prevent arcing over.

My main worry would be the ignition of the second engine - I am not sure that a motor designed for parachute expulsion is capable of igniting a second stage. Or do you plan to use a seperate ignition system?

[Seret]

I know I could go out and buy a 1st stage motor, but that is much less of a challenge.

Indeed. What you're suggesting as a solution to your problem is actually a reasonably sophisticated robotics approach, akin to something like a ballbot. It would be awesome if you got it to work, but it's a lot of extra complexity and weight.

Personally I think I would just spin stabilise the rocket, or look at a way of igniting the first stage without the delay.

[Nuke]

shouldn't the wheels be stationary initially. the only time they would need to spin up is to cancel out some angular momentum.

[Whirligig Girl]

If you didn't live in some place like North Carolina you could just build your own motor...

[acrain116]

@Nuke, AHHH!! I just finished a class about controllers and now you're bringing me back in! But yeah, a controller/feedback loop sounds like exactly what you need since that is what real reaction wheels use. As for the equation you want, it would be a function equating those input lateral forces ([force of gravity and/or drag]*[cos(angle of deviation from vertical, theta)]) to the centrifugal force created by the reaction wheel, or (g+F(drag))*cos(theta)=F(centrifugal). Theta will change over time which is why you need a controller. Drag and centrifugal/centripetal forces depend on the rocket but equations can be found online. This should hopefully at least get you started. Or I could be a big idiot who didn't learn anything in my class

[rpayne88]

Never mind this post.