June 5th, 2024, the CST-100 Boeing Starliner with Suni Williams and Butch Wilmore launched for a three hour tour eight day mission to the International Space Station (ISS).
This was the third flight of the Starliner that was contracted with Boeing in 2010 and was supposed to launch in 2017.
History
- First unmanned orbital test flight: December 20, 2019, launched via an Atlas V rocket. This was a partial failure likely due to insufficient integration testing and the mission clock being wrong putting the capsule into the wrong orbit and preventing docking with the ISS. The capsule landed two days later in White Sands, New Mexico.
This failure was widely regarded as a software failure (some 61 corrective actions were needed), although little was said about any thruster or hardware failures. Changes were made to filter out radio interference and the software was re-certified on January 18, 2021.
A series of valve failures in thirteen system values contributed to an additional year of delays. - Second unmanned orbital test flight: May 19, 2022, launched via an Atlas V rocket. Two thrusters (out of twelve) failed during orbit insertion and the software failed over to backup thrusters. Starliner was able to dock with the ISS and brought approximately 500 lbs of supplies. This mission was deemed a success, albeit with some minor docking issues. Capsule landed six days later in White Sands, New Mexico.
- Third orbital test flight, manned: June 5th, 2024, launched via an Atlas V rocket. Preflight anomalies in the capsule (aside from the oxygen leak problem in the Atlas V) included a helium leak — helium which is used to pressurize the thrusters. NASA elected to not delay the launch for that. After launch, five of the 28 maneuvering thrusters failed. In addition, docking with the ISS required manual overrides. Capsule was certified for up to 45 days of duration in space.
Present
The current situation is that of the five thrusters that failed, four appear to be nominal after cooling off, and on-the-ground testing of the same thrusters suggest a bubbling of teflon within the thrusters prevents fuel flow. (One of the five has been permanently disabled for the duration of the flight, but there is redundancy for that thruster.)
In addition, the thrusters were designed for satellites, not space capsules, and generally short bursts rather than long sustained usage.
Teflon: According to NIST (1956), Teflon/PTFE plastic goes through state changes (particularly at 20°C and 30°C) as a function of temperature as well as having a modest expansion coefficient.
The crew and capsule have now been in space for more than NASA’s 45-day certified duration, and there are several issues to be considered (in no particular order)
- Current version of the software is incapable of doing its own reentry burn without occupants
- Failure of thrusters upon heating still happens
- To mitigate that temperature-related failure, new software with an intermittent pulse series of thrusting maneuvers must be developed, debugged, and deployed
- Danger of reentry thrusters failing or partially failing remains
- SpaceX Crew Dragon does not have additional seats (although we could theoretically cram the additional crew into the Dragon)
- Crew has Boeing pressure suits that are incompatible with SpaceX hardware (although new, not fitted pressure suits could be flown up)
- There is a possibility of launching the next Crew Dragon with two empty seats
- This is costing Boeing a lot of money, and if the crew does not return on Starliner, it will cost Boeing even MORE money as they will have to re-fly this test. (At this point they almost certainly will have to re-fly to become certified.)
Analysis of Risk
I simply have to look at the risk side of the equation: (1) the capsule is already past its rated duration. Yes, it was designed to spend a half-year on orbit but they didn’t certify it for that for this flight, and the risk only continues to increase as a function of time. (2) Currently they have not stated _WHY_ the teflon is bubbling up, although since it does reproduce when tested on the ground, I’m skeptical. Any number of repeat tests can be performed, and forensic analysis of the components undoubtedly has already been performed. This implies that the thrusters should not be trusted. (3) I expect a new software program that uses intermittent thrust pulses to manage temperature / teflon expansion to be implemented, but testing that for the first time with people would be a last-resort measure. It also reduces redundancy since more thrusters have to be part of the program, leaving fewer available for fail-over. (4) Management will choose to wait and not commit if they can find any way to wait. (And management has a history of understating risk.)
I find the odds, barring an emergency evacuation, of the Starliner crew returning to Earth on Starliner to be increasingly small. The financial challenge (cost of seats on SpaceX, mission rescheduling), and the engineering challenge (deorbiting Starliner by itself and/or new fault-tolerant distributed thruster maneuvering software) cannot be ignored, but it will still be much safer to switch the crew over to a SpaceX return.
The longer it takes before their return, the higher the odds that our castaways will not be returning on the Minnow Starliner.
Further, I speculate that this will kill Starliner if, for no other reason, to stop the hemorrhage of money.
TL;DR: They’ve not coming home either soon or on Starliner. Starliner’s dead, Jim.
