Attitude control for a really big rocket

If this is meant for a model rocket it must be the biggest we’ve ever seen. [Scott] and [Trevor] took on the task of building a rocket attitude control system after reading about some research on the topic. But that researcher only tested the theories using simulations so they set out to build their own. The prototype above has a tank of compressed Nitrogen which can hold up to 3000 PSI. You can begin to understand why this needs to be used with a big rocket. The pressurized gas is connected through a regulator to four valves which feed nozzles around the circumference of the fuselage. An Arduino takes readings from a gyroscope and actuates the gas valves via a relay board.

You can check out the test rig in the video after the break. The prototype is suspended horizontally from a wire and its orientation held at one position by the system. There’s also a paper (PDF) if you’re interested in the equations that went into the stabilization control. This system would have been right at home on that huge sugar rocket we saw back in October.

39 thoughts on “Attitude control for a really big rocket

  1. This is neat and all, but I doubt it’ll be able to do more than straighten out very minor arcs in the flight. Having worked with high-power hobby rockets myself, it would seem that the other forces (air turbulence, fin flutter, and slight fin misalignment) would have an overwhelming effect on the flight of the rocket and all but drown out the small attitude corrections.

  2. It does seem to work well. I like it. I’m interested to see it in flight. However I thought there was a limit on the amount of time you can control a model rocket before your in handcuffs.

    All things considered it does seem a bit silly to use an arduino for such a project.

  3. I built the electrical controls for a payload on a Nike-Orion rocket. The payload frame was 18″ diameter and 48″ tall. The rocket produced 50-g’s at take-off.

    We had a larger budget than this project and so concrete forms, plywood and all-thread were not used in our construction.

  4. OP here–I guess I should clear up a few things. The system we built was intended to be a testing and prototyping platform, not for actual flight. It is also not intended to be a practical solution to lower-atmosphere attitude control. Mostly we just wanted something we could use to test attitude control algorithms, sensors, and attitude determination algorithms in a zero-g vacuum scenario…on a college student budget.

    @spiritplumber – Sometimes! Usually while a rocket is still in the atmosphere attitude control is accomplished using gimbaled motors or canards (rotating fins). Once it’s in space, though, canards are useless because there’s no air. The space shuttle and Apollo landers used thrusters in the same way we did, but theirs are/were actually little mini-liquid-fueled rockets (more power). I’m not sure how often cold-gas thrusters are used these days, and there are lots of other attitude control methods like magnetorquers, gravity booms, and spin stabilization. Attitude determination is an even messier problem, and even today these things aren’t easy problems.

    @ESylin – I guess it seems a little absurd to use an Arduino for a task as non-trivial as attitude control, but I think you’d be surprised at what you can accomplish with cheap technology these days. Using the 16MHz Arduino we were able to integrate the rate gyro readings, run the control algorithm, and send telemetry, all at ~40Hz. The valves take about 0.2 seconds to open and close, so at some point faster control doesn’t buy you a whole lot. People have been building guided rockets for a long time, and while it’s something of a toy now, I suspect that not too long ago an Arduino would’ve been a godsend in terms of performance and usability.

  5. This is a safety critical system, someone could potentially die if it goes wrong. And they are controlling it with an Arduino?!

    Really, that should be done with well designed and implemented code – most likely in assembler (C at a push), utilising at least a watchdog timer as a safety measure. It would not take too long to knock such a system up at all. And it’s much safer than relying on some third party boot loader who you have no real faith in.

  6. @Smoker_Dave – The system is really quite benign. Everything downstream of the regulator is between 0 and 100psi. People use an almost equivalent system to play paintball every day. Canned air uses higher pressures. Other than some catastrophic tank failure, what do you think might go wrong that would be prevented by a better flight computer? The Arduino has worked great for us, and at normal operating pressures (~30psi) the setup hardly seems safety-critical.

    Unrelated, is this you? –>

  7. The presence of an Arduino (or any microcontroller), does not, by itself, exclude a safety critical system from performing as intended. After all, we went to the moon (several times) using electronics. We fly entire airplanes using fly by wire technology. Most modern cars use fly by wire for acceleration AND braking.

    It is still entirely possible to build in heartbeat sensors, redundancy and other redundant mechanical systems and fail safer design methodologies to maintain safety.

  8. Hey, you think that if the Apollo people would not have used an Arduino, you are seriously mistaken. They would have jumped on it. They had to make every thing because no one else was making it.

    @ Scott Loved the use of the concrete form, that was a cool idea.

  9. Not only is it illegal in the US to do a DIY guided missile, but the US government pressured a foreign government into shutting down Bruce Simpson’s popular DIY cruise missile project in New Zealand:

    Comments in a previous HaD posting warn to stay away from guided rockets and stick to prop-driven UAVs.

    Of course, propellors are not effective in the vacuum of space, where this research project is targetted. ;-)

  10. nevermind the fact the bootloader only runs for 2 seconds to see if there is new software to be um loaded then starts up the program in the leftover space and that it is a atmel STK500 setup and that program is compiled in avr GCC

    the haters are usually too wrapped up in imagining things to actually sit down and understand that its the exact same crap with a noobie wrapper and a crappy “ide”

  11. It doesn’t seem to be taking into account the inertia that each adjustment causes – the very last test showing it rapidly moving left and right because the inertia from each adjustment caused over-correction, which then required an opposite adjustment, which again lasted too long, etc, etc.

    I’m not sure how difficult it would be to take that into account, but it seems like shortening each burst would result in a slightly slower but more accurate attitude correction.

  12. @Scott (OP): As I recall, some geosynch satellites use cold-gas thrusters for fine-tuning orbits. I could swear that we used cold-gas vernier thrusters on a human-rated vehicle but I could be wrong. It wouldn’t make a lot of sense in anything heavy, as cold-gas vernier thrusters are pretty inefficient in a cold vacuum. Nitrogen thrusters will have a specific impulse of around 60-70 seconds, where a bipropellant rocket (like the hypergolic thrusters used in the Mercury, Gemini and Apollo programs) have a specific impulse somewhere around 450 seconds. But for a lightweight hobby rocket, the cold gas makes more sense.

    And as for the legal implications of a guided hobby rocket, I think that’s ridiculous. The Constellation program was canceled because “orbital spaceflight should be the realm of private companies,” and yet we’re being discouraged from developing these things on our own. Sure, it might look like weapons at first, but then again, the Atlas, Redstone, Titan and Saturn rockets were all developments of ICBMs, the Saturn being a development of Wehrner von Braun’s A-10 ICBM.

  13. @Malvineous – The control law does take into account the inertia of the system. However, to do that properly you need to know the inertia of the system and the torque caused by the thruster firings. The torque stays pretty constant, but the inertia drops significantly as we use up gas. So, we have to constantly estimate the angular acceleration caused by the thrusters as the test progresses. You can read about the control law in detail in the paper linked above. We believe the oscillating behavior you’re talking about on the 100psi tests is largely due to the swinging of the payload and the high angular rates achieved. Our rate gyro pegged out on those tests, causing our angular acceleration estimates to get thrown off. The 100psi test was really just to see how quickly we could potentially reorient, not a serious control test.

    @Rob Wenworth – Any reference to the laws regarding this topic? I can’t seem to find the specific statutes. It would seem pretty silly for the feds to get nervous over a couple college students’ summer project. We didn’t use any information that isn’t freely available online. However, if the guys in black suits show up to take me away feel free to say I told you saj03r2ja

  14. I dropped out of the local amateru rocket club when newly required Federal registration and certification was required. You had to pass a test and to agree to be bound by specific rules under “penalty of perjury”., which means that you are entering into a contract that has the potential of criminal prosecution if you break the contract. One of the rules was no guidance systems on the rockets, although that rule was already in effect before the the new rule enforcement, except after entering into a contract you can go to jail for violation of that rule.

    The Portland State Aerospace Society says “We’re trying to be one of the first amateur groups in the world to make an actively guided rocket”, so they either got a special exemption, or they have not gotten sufficient attention to get noticed by the feds yet.

  15. active stabilization != guidance
    There are a number of people who create gimballed motor mounts in the highpower rocketry hobby. Yes they have ‘licenses’ under governing bodies such as tripoli &c, but realistically those don’t mean much.

  16. I dropped out of the local fireworks club for the same reason. Interestingly, the new onerous regulations that clamped down on rockets and fireworks and the ability to obtain interesting chemicals did NOT start with 9/11. It all started with the War On Drugs and related expansionist policies.

  17. I don’t think that the feds have the manpower to enforce all these laws and regulations, but much like local loitering laws, if you draw too much negative attention, they will use those laws to shut you down.

    So, just like most things in life, if you want to color outside the lines, you must not draw attention to yourself. Unfortunately, that rules out publishing your results in a popular forum (like the New Zealand DIY cruise missile guy who was shut down by the US feds for getting too much attention).

    Active stabilization is fine as long as it cannot be interpreted as target acquisition. You must maintain clear intentions and not cross certain lines in such matters.

  18. Typically monopropellants (Hydrazine) are used with a catalyst for attitude control. Cold gas thrusters are not efficient. Its a cool project regardless.

    Regarding guided rockets: It’s not illegal to put a guidance system of any kind on your hobby rocket. Everyone loves to say it is, and its typically a rule within amateur rocket societies, but there is no law against it in the US. If you think you’ve got the stuff to compete with Raytheon no one is going to stop you. The biggest barrier is the technical difficulty involved in developing these systems. You can’t build a very accurate IMU with MEMs gyros…

  19. Okay, here are the rules according to PSAS:

    They think that their guided rocket plans are okay.

    Their research looks much more thorough than mine, which was only what our local rocket club supplied us with at the time and seems inaccurate after reading the PSAS conclusions.

    So, I guess we can follow Grace Hoppers advice about how it is easier to apologize than to get permission. ;-)

  20. Cold Gas rockets have simplicity advantages over High Test Peroxide(HTP) however I think the energy density to complexity ratio favors Hydrogen Peroxide as a long term solution.

    It makes sense to me to proceed with initial research(computer science/IMU/PID tuning) based around a cold gas rocket as a proof of concept, but nitrous oxide hybrids and HTP show the most promise if you desire to achieve aerospace performance without messing with cryogenics or dicking around with BATFE/CPSC red tape.

    I wonder where the world would be if Samuel Colt had been thrown in the clink for his Nitrous Oxide business? Would we have standardized parts today?

    The very technology we need in order for our species to explore new frontiers is locked behind red tape and bureaucracy. What better way to inspire innovation than to criminalize the innovative process as “too dangerous”.

    The “common sense” measures drive innovators underground and discourage children from an interest in science.

  21. Here is the same thing:

    for this one :-)

    That is Hacking :-D

    New liquid engine with 7 tionnes thrust on it’s way.

    As you may know, the launch this summer was a no go, apparently because of a frozen O2 valve in a heavily forced launch campaign, but follow (and possibly donate to) this cool non-profit project here:

    There will be a new calmer launch this summer (Let’s hope it does not get to look like the Trident cousin mentioned above)


  22. Well, it’s not technically a model rocket, but it’s still an amateur rocket, and it’s way bigger and ambitious than this one:

    Their work is still in progress, but I would really like to see this DIY project completed! The same guys have already built a working submarine, so we’re talking serious hacking here:

  23. Complaints about using arduinos for rocket stabilisation are ludicrous.

    Do these complainers not realise that Scout rockets used to be able to moderately reliably fly themselves into orbit using a paper tape guidance and control computer? On the avr, the arduino is just a bootloader. It just listens to the serial port after reset and otherwise never gets touched – you just run programs on the bare metal atmega as usual.

    An arduino is a staggeringly capable device by the standards or orbital rocket computing power. If you can’t stabilise a rocket with it, that’s a reflection on your abilities as an engineer, not the arduino.

  24. Nifty…

    I’m no expert on it’s legallity but, even if it is technically illegal I’m sure test flights would be fine. Model rocketeers have been “guiding” rockets for a long time. Look up SunSeeker.
    As far as rocketry guidelines are concerned I understand there is a provision about putting an explosive payload on a rocket and aiming it at something. This doesn’t mean guidance and it is specifically speaking about explosive payloads.
    And for those of you touchy about following the letter of the law I’d be more wary of the writing in the Patriot Act. Under that mess using the word “hack” could be considered “terrorist activity”. Hell all of us at this web site could be taken in for participation in a terrorist cell.

    The bottom line is as long as the right intent is there and reasonable caution has been taken most johnny lawmen will just overlook technical violations.

    My official legal advice to someone avoiding legal trouble, never leave the house and work for the government.

  25. A second note. The reason gas is used on the shuttle is for actual spaceflight. Where there is no air fins won’t work no matter how fast you travel.

    In the atmosphere guidance is achieved mainly with slight adjustments in the motor nozzles. For model rocketry it’s far more efficient to just use fins to control attitude when there is an atmosphere to use as leverage.

  26. An electric arc assisted thruster could improve the response time of the thruster below the 200 msec needed to open the valve. Turn the nozzles into something resembling an offspring of a carnal relation between a cold gas thruster and a plasma cutter. Once the gas flows, its velocity can be modulated very quickly by igniting and quenching the arc.

    For atmospheric flight, you may like to consider using plasma for maneuvering; a plasma sheath selectively created over part of the vehicle could work quite like a control surface, by altering the drag over that part. As there are no moving parts involved, the reaction speed and reliability could be drastically improved.

  27. There should be a new regulation here: If you’re going to claim that something is illegal, cite relevant laws and/or legal precedent.

    As for the project:
    You’ll need roll stabilization, which is straightforward to do with rollerons on the wings (rear fins). You could also use the gas to drive fin positioning (you can even work out proportional control with this method)

    Bang-Bang guidance? it works, and is the minimum time solution. You’re going to use more energy with this control method rather than some sort of proportional method.

    This is probably ITAR, meaning it’s export controlled. Anything to do with guided rockets (including telemetry) is ITAR in the States. I’m not sure how the law applies to the self-funded amateur enthusiast, I’ve always been on the government funded side (research lab, intern @ NASA JPL). Don’t publish any specifics (e.g. control code) and you should be fine.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s