Reaction wheel failures leave Swift rescue mission spinning in orbit

TL;DR

A catastrophic failure of two out of three reaction wheels aboard the Link cargo spacecraft has left a critical rescue mission to the International Space Station in jeopardy, with ground teams assessing whether the vehicle can safely dock or must be scuttled. The anomaly, confirmed during preliminary investigation on July 28, 2026, threatens a fast-approaching crew-support timeline aboard the orbiting laboratory.

What Happened

At 14:32 UTC on July 28, 2026, flight controllers at NASA’s Johnson Space Center received a telemetry alarm they had dreaded: two of three reaction wheels on the Link cargo spacecraft had inexplicably ceased rotation, stripping the vehicle of its primary means of precision attitude control. The Link vehicle, launched just 36 hours earlier on an emergency resupply and crew-relief mission code-named Swift-9, is now drifting in a stable but uncontrolled orbit 408 kilometers above the Earth, with a docking window to the International Space Station that closes in less than 72 hours.

Key Facts

• The Link spacecraft is an upgraded Cargo Dragon 2 variant, built by SpaceX, flying its 23rd mission overall but only its second configured for crew-capable rescue operations.
• The Swift-9 mission was declared a priority Level 1 emergency after a micrometeoroid strike on the ISS’s Nauka module on July 22 caused a slow cabin-pressure leak, forcing the station to reduce its crew complement.
• Reaction wheel assemblies on Link, supplied by Honeywell Aerospace, each produce up to 100 Newton-meters of torque and are designed for a 15-year orbital lifespan; the failed units had accumulated only 18 months of in-service time.
• Both failed wheels stopped at different orbits on different axes — Wheel 1 at perigee over the Pacific, Wheel 2 at apogee over the Indian Ocean — ruling out a single common-cause event like a power surge.
• Preliminary telemetry shows no thermal exceedances, no radiation spikes, and no commanded shutdown signals in the moments before each failure, suggesting an internal mechanical or bearing seizure.
• Without at least two working reaction wheels, the Link vehicle cannot maintain the pointing accuracy required for autonomous docking with the ISS’s Pirs docking port, and manual backup systems can only hold orientation for about 90 minutes at a time.
• The last comparable failure in a human-rated orbital vehicle occurred during Space Shuttle mission STS-93 (1999), when a reaction wheel failure forced a manual docking with the ISS, but that vehicle carried four wheels total as redundancy.

Breaking It Down

The reaction wheel is the unsung workhorse of spacecraft attitude control. Unlike thrusters, which consume finite propellant, a reaction wheel spins a heavy rotor inside the spacecraft; to rotate the vehicle, the wheel is sped up or slowed down, and the spacecraft reacts by turning the opposite direction. Three wheels, mounted orthogonal to one another, give the vehicle full three-axis control. Lose one wheel, and the spacecraft can still manage with the remaining two — but the control becomes slow, the pointing stability degrades, and any failure that forces the remaining two wheels to work harder accelerates their own wear.

With two of three wheels dead, Link has lost 67 percent of its attitude control capability — a failure that in any previous NASA cargo program would have triggered an immediate stand-down and return-to-ground checklist.

The engineering analysis here is brutally simple. A single reaction wheel can only provide torque along its single axis. To even hold a steady orientation, a spacecraft needs torque authority about all three axes simultaneously. With only one functioning wheel, Link can control rotation on one axis but will freely tumble on the other two. The spacecraft’s star trackers, which rely on fixed pointing to identify constellations for navigation, will lose lock within minutes. Its main S-band antenna will lose line-of-sight to the Tracking and Data Relay Satellite System, severing high-bandwidth telemetry within roughly one full orbit cycle (90 minutes). Ground teams confirmed at 15:47 UTC that Link had already entered a slow tumble of 0.8 degrees per second — well within safe structural limits but far beyond the 0.03 degrees per second tolerance for docking.

SpaceX’s propulsive thrusters on the trunk section can provide attitude control, but they are designed for brief orbital tweaks, not sustained station-keeping. Using them to replace the lost wheels would consume propellant at a rate of roughly 12 kilograms per hour — exhausting Link’s entire deorbit fuel reserve in under 24 hours. That calculus leaves engineers with an excruciating choice: try to rush a manual or semi-autonomous docking using thrusters alone, or abort, preserve the propellant, and bring Link home with its cargo undelivered and two ISS crewmembers stranded an extra month.

What Comes Next

The SpaceX and NASA joint anomaly review board will converge at Kennedy Space Center tomorrow morning at 08:00 EDT (12:00 UTC) to decide the vehicle’s fate. The meeting agenda, obtained by Ars, includes three critical time-dependent variables.

1. August 1, 2026 — The absolute drop-dead docking window. After this date, the ISS orbital alignment shifts such that Link would require a 36-hour phasing maneuver — an impossible fuel cost given the attitude-thruster burn scenario. A decision to abort must come by July 30 to allow for a safe southern-Pacific deorbit burn.

2. Telemetry from a third-party optical tracking pass. The U.S. Space Force’s Space Surveillance Network will perform a high-resolution radar pass over Link at 03:18 UTC on July 29 to look for debris shedding or anomalous spin — signs of a cascading mechanical breakup.

3. A hardware-in-the-loop test at SpaceX’s Hawthorne facility. Engineers will mount a flight-replica reaction wheel on a test stand and attempt to replicate the failure signature by injecting simulated bearing faults. Results are expected by 20:00 UTC July 29.

4. The ISS crew’s immediate health status. NASA flight surgeon Dr. Laura Chen reported at the 17:00 UTC briefing that station oxygen supplies are now at 72 percent of nominal, and the leak rate has not changed — meaning the crew can survive without resupply until September 3, buying some margin if the vehicle is forced to abort and relaunch.

The Bigger Picture

This failure lands squarely at the intersection of two accelerating trends in human spaceflight: the Commercial Cargo Resupply Model and the Crew Safety Margin Debate. NASA’s decision to certify SpaceX’s Cargo Dragon 2 — and by extension, the Link variant — with a three-reaction-wheel architecture was a calculated risk. The Space Shuttle flew with four wheels; the Russian Soyuz uses a combined gyrodyne-thruster system; Orion flies with six. The reasoning was simple: statistically, the probability of losing two out of three wheels over a 30-day mission was calculated at **0.