Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

I am interested in building a rocket with feedback control system

  1. Jun 28, 2007 #1
    I am interested in building a rocket with feedback control system....

    so I have been interested in model rocketry lately and I was wondering how feasible it would be to design and build a model rocket for my senior project complete with a control system to ensure it flies straight up. I was considering using two fins with adjustable flaps to control the rotation of the rocket about its axis and 4 fins to control the angle of the rocket. 2 fins would be used to control each of the two components. I would require 3 independent feedback systems. One feedback system for ensure the rocket does not rotate about its axis, 1 feedback system to control the 'x component' of the tilt of the rocket and another feedback system to control the 'y component' of the tilt. The rotation angle, x component, and y component will be controlled to a value of zero.

    This rocket will be designed to go straight up only.

    How feasible would this project be? I know I would need access to a wind tunnel that could simulate aerodynamic drag and cross wind effects. Also, I would need to beef up on my control systems. I would want to use digital control systems, but I am a little weak in this subject.
  2. jcsd
  3. Jun 29, 2007 #2
    Actually, you might not need to have roll, Four fins with control surface is enough to have the rocket point to any direction it wants, and four wings too. To avoid roll, you need to make sure the fins are perfectly parallel to tube and it doesn't flutter.

    For directional control, can't you just use a photo or IR sensor, wait for the sun to go 12noon, so the rocket will point to the sun and flies straight up. Or.. you could also fire a flare and the rocket will intercept it.

    It's much simpler than using gyroscopes, at least IMO.. I've used photo sensors in the past to build a light following vehicle that used only three electronic components, a photosensor, and two transistors :) It has lightning-quick response and has good dampening qualities.

    Ok, back to control, since rockets/missile and the like are extremely sensitive to flight controls due to their speed, you really need some 'dampening' qualities built into the design so it doesn't 'snake' out of control.. Unless you can design military-grade control system that would do everything for you..

    -One is putting your center of gravity a good distance in front of center of pressure or center of lift, and it will have good stability. But the controls won't become effective until it's flying fast enough.
    -Use conventional layout, avoid canard. Control surfaces in the rear and fixed wing at the front.
    -Avoid freeplaying in the control surfaces. Using Flaps instead of having the whole control surface move in one piece is also a great idea to improve control stability. If you use one piece 'stabilator' though, the pivot point should be slightly ahead of center of pressure to have stability.

    Finally, I would suggest you to start with unpowered models first to minimize danger of the controls going haywire. They can be launched by compressed air, spring or rubber strands. Also try to get books on 'control systems' you need to learn about stuff like dampening feedback. It's a very complex thing really, but sometimes, the solutions are very simple ;)

    Tell me if you like to go the IR sensor way and I may give you illustration that might help, including how to rig it in the rocket.
    Last edited: Jun 29, 2007
  4. Jun 29, 2007 #3
    It is good that I might be able to do without the roll control. However, any slight change in roll position will seriously affect the 4 other control fins and the axes they control so I need to be sure the roll position is at 0 degrees at all times.

    I will check into using an IR sensor and control beacon or some sort. However, in order to control this system I will need to be able to collected x and y axis coordinate data somehow and this would require multiple photosensors I would think.

    And yeah, I realize that the control system would need to be very very underdamped just for safety purposes.

    I plan on placing the center of gravity a couple of rocket diameters in front of the center of pressure.

    I considered having the entire control surface move about a pivot point, but upon further consideration I realized it would be better just to have fixed fins with pivoting control flaps as you suggest.

    Also, I have taken a control systems course and I am familiar with the fundamentals or lead/lag compensator design, PID controller design, RL plots, and Nyquist plots. I think I would be able to design an analog control system to yield a particular response given the rocket and fin parameters (the transfer functions for the various servos and fin configurations). However, I think I should go with digital control and not analog. Unfortunately, I am not too strong with digital control systems (z-transforms and whatnot), so I will have some studying to do.

    I would definitely be interested in using a IR sensor of some sort but it is not clear to me yet how one would use this method.

    oh, and thanks a lot for the advice.
    Last edited: Jun 29, 2007
  5. Jun 29, 2007 #4
    Trust me, in a four-fin 'missile' even with radical changes to roll or bank position, the flight path is not compromised noticeably. Having roll control therefore is redundant given the requirements.

    I've tested unpowered missile gliders, a scaled down version of AIM-120 missile. Only 2 inch long, it achieved good glide performance, with a measured glide ratio of 10. I've observed that it doesn't matter what the bank(roll) angle is, the missile always flew straight. This proves that aerodynamically, a four fin missile can glide or even turn left or right no matter what the roll angle is with the correct flap adjustments.

    Now that you may have no roll control at all. A 2D control system should be able to correct any deviations from intended path to target no matter what the roll position is or if even if it's rotating slowly..

    Now on the control system. If you use photo sensors, the idea behind the logic is that less or more light enters the sensor whenever the target moves to the left or right, same with up and down.

    So you need two sensors for target x-axis tracking and two sensors for Y-axis tracking. I would recommend you to research on IR heat seeking missile sensor design. They've figured out a way on how to do it.

    The logic would be for example, sensor (a) on x-axis would move servo motor 1 CW and sensor (b) on x-axis would move servo motor 1 CCW. For sensor (c) on y-axis would move servo motor 2 CW and sensor (d) on y-axis would move servo motor 2 CCW.

    I would advice that for not-so complicated approach is to limit your electronic components to transistors, resistors, sensors and diodes only, to minimize or eliminate feedback lag, avoid digital electronics and capacitors(I believe it's not needed and may cause feedback lag). 'LIghtning fast' response is certainly beneficial.. To save you the trouble of oscillations from the start

    If the electronic package is not enough to dampen oscillations or 'snaking'. I would leave the dampening to tweaking aerodynamics design and servo motor settings. Enlarging all the fins(front wings and rear control fins) to reduce wing loading, to make the missile respond quickly to any control inputs without stalling, and hopefully not overdo it(by reducing servo motor torque or push/pull force).

    So it's best to construct the missile as unpowered glider first, have it fly horizontally and intercept a 'hot' target. Gradually increasing launch speed to observe any fluttering or oscillation. If it succeeds in those tests, you might have better chance it will also succeed in powered flight with much greater speeds:biggrin:

    But I must remind you take my advice 'with a grain of salt'. First, radiation-seeking feedback control systems may not be the best solution to your requirement. Two, I never conducted any powered missile flights. And lastly, I've never used the control system in an aircraft yet, but wheels vs fins, it's nearly the same thing ;)

    Attached Files:

    Last edited: Jun 29, 2007
  6. Sep 26, 2008 #5
    Re: I am interested in building a rocket with feedback control system....

    hey this project sounds very interesting. On this topic, do you know of a good source for books on rocket control systems? Does anyone have any good titles? Any help would be greatly appreciated.
  7. Sep 30, 2008 #6
    Re: I am interested in building a rocket with feedback control system....

    First might want to check the laws in your country. This becomes a guided missile, and the ATF take a special interest in things like that

    slowly rolling during flight might be an unachievable dream. you have to be extremely careful when attaching your fins, and slight misalignment will give a parabolically increasing roll rate. (don't i know thats the truth)

    as far as building up speed for control effectiveness.....soon as you clear the rail, you'll have enough airflow to have effective controls, otherwise the rocket would immediately backflip off the rail

    definitely telling the rocket to "aim for the sun" with photosensors is probably the best (easiest) way of doing this, just make sure your control system can keep up with the speeds the rocket tends to change at.

    full flying tails would definately be easier to make, but would probably provide way too much control authority. flaps would be the way to go, but just thinking off the cuff about it, it would be a nightmare, and i would still probably try a full flying tail.

    i hope you have a high powered rocketry certification, casue thats what it'll take to get this beast off the ground. a level 1 cert should work fine i would think, 38mm motor with larger body tube (50mm range)

    you can do whatever works best for you. i wanted to do this for a senior project as well. i didnt for a multitude of reasons. needless to say, i've put some thought into this specific project.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook