Understanding Reaction Wheel
How a reaction wheel produces torque
Inside the unit, an electric motor spins a balanced metal flywheel. By Newton's third law, the torque used to accelerate that flywheel pushes back equally on the spacecraft, rotating the body the other way; speeding the wheel up turns the satellite one direction, slowing it turns it back. Because the motor's speed varies smoothly and continuously, the control is extremely fine — reaction wheels typically produce small torques (from a small fraction of a millinewton-metre on a CubeSat up to roughly a newton-metre on large satellites) while spinning at up to several thousand rpm. A set of wheels held near zero average speed forms a "zero-momentum" system, distinct from a single high-speed momentum wheel, which instead supplies gyroscopic stiffness about one axis. This quiet, propellant-free control is why almost every three-axis-stabilised satellite carries a wheel set.
Momentum saturation and "dumping"
Reaction wheels have one hard limit: they can store only so much angular momentum before spinning at their maximum rate. Small but relentless external torques — solar radiation pressure, the gravity gradient, aerodynamic drag in low orbits and residual magnetic effects — are continuously absorbed by the wheels, so a wheel spins ever faster until it saturates and can no longer respond. Operators pre-empt this with momentum dumping, or desaturation: applying an external torque to bleed momentum away while the wheel spins back down. In low Earth orbit, magnetorquers can push against Earth's magnetic field to do this for free; higher up, or when more authority is needed, small thrusters are fired instead — the one moment a nominally fuel-free system still consumes propellant.
Reaction wheels vs other attitude actuators
Reaction wheels are one of several ways to point a spacecraft, and most satellites blend a few. Wheels excel at fine, continuous, fuel-free control but hold limited momentum; magnetorquers and thrusters back them up, while high-torque control moment gyroscopes are reserved for large or highly agile platforms such as the International Space Station.
| Actuator | How it works | Best at | Main limitation |
|---|---|---|---|
| Reaction wheel | Varies flywheel speed for reaction torque | Fine, fuel-free 3-axis pointing | Saturates; needs desaturation |
| Momentum wheel | Spins fast at a fixed speed | Passive one-axis stability | Controls a single axis |
| Control moment gyro | Tilts a spinning rotor | Very high torque, agile slews | Complex; geometric singularities |
| Magnetorquer | Coil torques against Earth's field | No moving parts, no fuel | Weak; needs a magnetic field |
| Thruster | Expels propellant | Strong torque anywhere | Consumes fuel |
When wheels fail: the Kepler lesson
Because losing a single wheel can leave a satellite unable to point, reaction wheels are a classic single-point-of-failure concern — which is why a fourth, skewed wheel is usually flown so any three can maintain control. NASA's Kepler exoplanet telescope is the cautionary tale: launched with four wheels, it lost one in July 2012 and a second in May 2013, leaving too few for the ultra-stable pointing its planet hunt demanded. Rather than abandon the spacecraft, engineers salvaged it as the "K2" mission, balancing the craft against solar radiation pressure to act as a virtual third wheel, with brief thruster firings roughly every six hours to hold its roll axis steady. (A common myth to avoid: the Hubble Space Telescope's repeated pointing scares have come from its rate gyroscopes, not its reaction wheels.)