Water Rocket Attitude Control for Satellites

ISEF Category: Engineering Technology: Statics and Dynamics

Subcategory: Aerospace and Aeronautical Engineering  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

Spacecraft do not steer like cars. They have to rotate without a road, brakes, or gravity helping them. That means a tiny burst from a thruster can matter a lot. You can model that problem on a tabletop and test whether your control system keeps a small satellite pointed where it should go.

What Is It?

This project studies how a small satellite can change and hold its orientation using cold-gas thrusters. Orientation means which way the object points in space. Think of a phone in your hand. You can spin it, tilt it, or roll it. A spacecraft does the same thing, but it has to track that motion with math and sensors.

The control system has three main parts. First, a sensor estimates attitude, which is the spacecraft’s orientation in 3D space. A quaternion is a compact way to store that orientation without the angle glitches that can happen with simpler methods. An extended Kalman filter, or EKF, blends noisy sensor data into a smoother estimate. Second, a controller, such as LQR, decides how hard and how long to fire each thruster. Third, the thrusters create torque, which is a twisting force that rotates the body.

You do not need real space hardware to study the core idea. A low-friction air-bearing tabletop lets a small platform rotate with little resistance, so you can test control logic in a realistic way. The water-rocket style frame and printed nozzles let you explore how nozzle geometry, thrust direction, and pulse timing affect pointing control.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real engineering system with clear numbers. You can measure pointing error, overshoot, settling time, and fuel use, then compare different nozzle designs or control settings. The project connects to spacecraft docking, CubeSat stabilization, and small-satellite attitude control. You will learn mechanics, sensors, feedback control, and data analysis in one project.

Research Questions

• How does thruster nozzle angle affect steady-state pointing error on a low-friction tabletop platform?
• How does pulse timing affect overshoot and settling time in quaternion-based attitude control?
• What is the effect of different EKF tuning values on orientation estimate noise and control stability?
• To what extent does an LQR controller reduce pointing error compared with open-loop firing?
• Which printed nozzle geometry produces the highest torque per unit gas use?
• How does adding sensor noise change the accuracy of the attitude estimate and the final pointing direction?

Basic Materials

• 1U-style lightweight frame or 3D-printed cube body.
• Paintball CO2 cartridges or other small cold-gas source, handled only with adult supervision.
• Assorted printed thruster nozzles with different exit angles and diameters.
• Microcontroller such as Arduino or Raspberry Pi Pico.
• IMU sensor module with accelerometer and gyroscope.
• Battery pack with proper voltage regulation.
• Breadboard, jumper wires, and connectors.
• Low-friction rotating platform, such as a Lazy Susan modified for air support.
• Shop-vac or other air source for the tabletop air bearing.
• Smartphone or overhead camera for motion recording.
• Tape measure or printed angle reference grid.
• Digital scale for mass checks.
• Safety glasses and gloves.

Advanced Materials

• 3D printer or access to one for custom nozzle and mount fabrication.
• Force sensor or load cell setup for bench thrust testing.
• High-rate IMU or motion-capture-grade sensor.
• Servo or solenoid valves for repeatable thruster pulsing.
• Pressure regulator and fittings rated for the chosen gas source.
• Data acquisition board for synchronized sensor logging.
• MATLAB, Python, or similar environment for EKF and LQR implementation.
• Air-bearing puck or custom air table hardware.
• Calibration fixture for nozzle alignment and torque measurements.
• High-speed camera for rotation tracking.
• Safety enclosure for gas testing.

Software & Tools

• Python: Processes sensor logs, fits control models, and compares pointing error across trials.
• ImageJ: Measures rotation angle from video frames and checks whether the platform follows the commanded motion.
• Tracker: Tracks marker motion in video and helps estimate angular response over time.
• GNU Octave: Runs matrix-based control design and state-space calculations without paid software.
• MIT OpenCourseWare: Offers free control systems and dynamics lectures for background on state estimation and feedback.

Experiment Steps

1. Define the attitude problem you want to solve, such as pointing to one angle, rotating by a set amount, or holding a fixed heading.
2. Choose the single variable you will change first, such as nozzle angle, nozzle diameter, or controller gain.
3. Build a measurement plan that lets you compare commanded motion, estimated motion, and real motion from the same trial.
4. Design controls that separate actuator effects from sensor effects, so you can tell whether failures come from thrust, noise, or tuning.
5. Create a baseline model of the platform before you test closed-loop control, so you know what the system should do in theory.
6. Plan how you will compare trials with statistics, not just by eye, so your conclusions rest on numbers.

Common Pitfalls

• Trying to tune the controller before the platform has a clean mass and inertia estimate, which makes the math disagree with the hardware.
• Letting the nozzles sit even slightly off-axis, which adds unwanted torque and hides the real effect of nozzle geometry.
• Trusting raw IMU output without checking drift, which can make the quaternion estimate look stable when it is not.
• Comparing video from different lighting setups, which changes marker detection and corrupts angle measurements.
• Using thruster pulses that vary from run to run, which makes it hard to tell whether the controller or the hardware caused the motion.

What Makes This Competitive

A stronger project goes beyond making the platform move. You would compare several nozzle shapes, several controller settings, and at least one baseline method, then quantify which design gives the best pointing accuracy for the least gas use. Strong entries also separate modeling error from sensor error and report uncertainty, not just average performance. If you can explain why your system works, and when it fails, your project starts to look like real aerospace engineering.

Project Variations

• Test whether different nozzle exit angles change torque efficiency more than nozzle diameter does.
• Swap the sensor setup and compare a quaternion EKF against a simpler complementary filter for attitude estimation.
• Change the platform mass distribution and measure how inertia affects LQR tuning and settling time.

Learn More

• NASA Technical Reports Server: Search for free papers on CubeSat attitude control, cold-gas thrusters, and spacecraft pointing.
• NOAA National Centers for Environmental Information: Use rotation and motion data concepts from geoscience records as practice for handling real-world measurement noise.
• MIT OpenCourseWare: Search for free lectures on feedback control, state-space systems, and dynamics.
• PubMed: Search for review articles on sensor fusion, Kalman filtering, and inertial measurement systems.
• Spacecraft Attitude Determination and Control: A standard textbook to look for at a library or preview through university catalogs.