Reaction Wheel

Spin a flywheel one way and the entire spacecraft turns the other way — Newton's third law as a precision pointing tool.

Controls a spacecraft's orientation in space by accelerating or decelerating an internal flywheel. Because angular momentum is conserved in a closed system, speeding up the wheel in one direction causes the spacecraft body to rotate in the opposite direction. Three wheels mounted along perpendicular axes give full three-axis attitude control; a fourth is often added as a redundant spare. The wheels run entirely on electrical power, so no propellant is expended during normal pointing operations.

Each reaction wheel consists of a precision-balanced metal rotor — typically aluminum or steel — driven by a brushless DC motor. An onboard attitude control computer compares the desired pointing direction with sensor readings (from star trackers, gyroscopes, and sun sensors), then commands each wheel's motor to apply torque. Increasing wheel speed builds angular momentum in the wheel; by conservation of momentum, the spacecraft acquires equal and opposite momentum. Wheels operate bidirectionally, typically between −6,000 and +6,000 rpm, so they can exert torque in either rotational direction. Over time, external torques from solar radiation pressure, atmospheric drag, or magnetic fields steadily push momentum into the wheels until they approach their speed limit — a condition called saturation. A brief thruster firing or magnetorquer pulse then "unloads" the excess momentum while the wheels slow back to a useful operating range.

In plain terms

Sit on a spinning office chair holding a bicycle wheel by its axle. Now spin the bicycle wheel one way — the chair (you!) will start rotating the opposite direction. That's exactly what a reaction wheel does to a satellite.

Also called: momentum wheel, reaction wheel assembly, RWA.

Why it matters

No propellant consumed for normal pointing — a reaction-wheel-equipped satellite can repoint itself thousands of times over a decade using only solar power.
Pointing accuracy of arc-second class or better is achievable, enabling telescope operations, precision Earth imaging, and laser communications.
Scalable across a wide size range: CubeSat wheels store 15 mN·m·s; large observatory wheels store 40–100 N·m·s.
Kepler's planet-finding mission depended entirely on reaction wheels for 0.6 milli-arcsecond pointing stability; when two wheels failed in 2012–2013, engineers redesigned the entire mission concept around solar pressure balancing.

Where it lives

Hubble Space Telescope — four reaction wheel assemblies (4 N·m·s class, 3,000 rpm), used for slewing and guiding
NASA Kepler / K2 — four Ball Aerospace wheels at up to 4,000 rpm for ultra-precise stellar photometry
Nancy Grace Roman Space Telescope — six reaction wheels (18-inch rotors, ~45 lb each, up to 4,000 rpm)
Blue Canyon Technologies RWp015 through RW16 series — standard in CubeSats and small LEO satellites
Bradford Space W18–W45 series — medium and large satellites including GEO telecom platforms

Key numbers

Momentum range (small sat): 0.015–0.5 N·m·s (BCT RWp015 to RWp500 series)
Momentum range (large sat): 1–100 N·m·s (Bradford W18–W45; Collins RSI 68 series)
Typical speed range: ±1,000–6,000 rpm (Operational; hardware limits typically ±6,000–9,000 rpm)
HST wheel torque: 0.8 N·m (At up to 3,000 rpm; rotor inertia 0.84 kg·m²)
Roman wheel mass: ~20 kg each (18-inch diameter, six wheels total)
Saturation: Momentum unloaded via thrusters or magnetorquers (Required periodically as external torques build up angular momentum)

Surprising facts

Hubble's reaction wheels rotate the 11-tonne telescope at about the same angular speed as a minute hand on a clock — roughly 90° in 15 minutes — yet the same wheels keep it locked on a target to within 7 milli-arcseconds during observations.
A reaction wheel physically cannot change where a spacecraft is in space — only where it points. All translational maneuvers still require thrusters.
Kepler's two failed reaction wheels triggered a mission redesign in which engineers balanced the entire spacecraft against solar radiation pressure, effectively using the Sun as a third virtual wheel to maintain pointing stability.
The smallest commercial reaction wheels for CubeSats store only 15 mN·m·s of angular momentum — roughly the rotational energy of a golf ball rolling at walking pace — yet that is sufficient to repoint a 1 kg nanosatellite.

A short history

Momentum storage for spacecraft stabilization was demonstrated conceptually in the early 1960s. The Orbiting Astronomical Observatory program and Nimbus meteorological satellites used momentum wheels for attitude stability from the mid-1960s onward. Reaction wheels — operated bidirectionally around zero speed for three-axis pointing — became standard equipment on three-axis stabilized satellites through the 1970s and 1980s. Hubble Space Telescope (1990) brought widespread public awareness of reaction wheel failures, which recurred in 2003 and required Servicing Mission 4 in 2009 to replace all six gyroscopes and attitude control hardware. The most dramatic recent case was the Kepler space telescope, whose science mission was fundamentally reshaped after two of its four wheels failed in 2012–2013, illustrating both the centrality of reaction wheels to precision pointing missions and the ingenuity required when they fail.

Sources

Blue Canyon Technologies Reaction Wheel Specifications (https://www.bluecanyontech.com/components/reaction-wheels)
NewSpace Systems LIBRA Reaction Wheel Datasheet (https://www.newspacesystems.com/wp-content/uploads/2024/06/Reaction-Wheel-Datasheet-A4-5-1.pdf)
Kampmeier et al. — Reaction Wheel Performance Characterization Using Kepler Spacecraft, AIAA (2018) (https://arc.aiaa.org/doi/pdf/10.2514/6.2018-2563)
NASA Goddard — Moving Roman: Reaction Wheels (https://svs.gsfc.nasa.gov/vis/a010000/a014500/a014525/14525_MovingRoman_ReactionWheels_HTML_Transcript.html)
NASA Science — Operating Hubble with Only One Gyroscope (https://science.nasa.gov/mission/hubble/observatory/design/hubble-one-gyro-mode/)