NASA’s MAVEN Mars orbiter has been in orbit around the planet since September 2014. For the majority of that time, and following science commissioning (Sept-November 2014), the spacecraft has been studying the Martian atmosphere, yielding valuable science. Except for the past three months, that is.
On February 22nd, 2022 – ironically the day Shannon Curry, appointed to take over the role of MAVEN’s Principal Investigator in August 2021, was making a three-hour presentation on the vehicle’s science findings at the conclusion of its latest 6-month mission extension – when Things Went Wrong.
We finally finished the presentation, I turn my ‘phone back on, and our project manager calls me immediately. I’m thinking, he’s calling me to be like, ‘Congratulations, you did it, you’re doing great!’ And he was, ‘Shannon, we’re in safe mode.’
– Shannon Curry
Regulars to Space Sunday will know that “safe mode” is when a spacecraft has encountered a condition that exceeds its programmed parameters / expectations, causing it to shut down most of its non-essential systems and services and ‘phone home with a call of “I’m in a spot of trouble, folks!”
Safe modes are rarely easy to diagnose and resolve remotely, with MAVEN (Mars Atmosphere and Volatile EvolutioN mission), the issue would prove to be almost catastrophic.
In order to both study Mars and communicate with Earth, MAVEN must periodically re-orient itself. Up until 2017, it did so by using one of two Inertial Measurement Units (IMUs) to calculate its position, attitude and rotation. However, from 2017 through until the end of 2021, MAVEN has been reliant on just one unit – IMU-2 – after IMU-1 experienced data issues.
By the start of 2022, IMU-2 was starting to show issues of its own, so a project was started to write new software to enable MAVEN to orient itself using the stars in what the mission team called “stellar mode”, a project that would take until late 2022 to complete. In the meantime, the vehicle was instructed to switch back to using IMU-1, with the power to the unit being periodically recycled to help with keeping it operating smoothly.
However, on February 22nd, 2022, with MAVEN oriented to communicate with Earth, a power recycle was started and IMU-1 crashed, and when IMU-2 automatically started, it had absolutely no idea of where it was, and MAVEN went into a loop of trying to restart IMU-1 after shutting down all science operations.
When it was clear IMU-2 was “lost”, and IMU-1 was not going to recover, risking MAVEN drifting out of communications alignment, the mission team took a desperate step: heartbeat termination.
That term is not just for dramatic effect: basically, it’s like ripping the cord out of the wall. We ordered the vehicle to shutdown and reboot its primary computer without switching to the back-up. When that failed, we had no choice but to then swap to the back-up and we’ve never been on that before.
– Shannon Curry.
Whilst the switch to the never-used back-up computer was a risk, it nevertheless allowed position data to be given to IMU-2 to ensure communications could be maintained with Earth. This allowed the mission team to accelerate the work on developing the “stellar mode” software.
On April 19th, the first version of the software was uplinked to MAVEN five months ahead of its due date. However, it could only be tested by shutting-down IMU-2. If the software failed, there was no guarantee either IMU would reboot, leaving MAVEN to drift out of its communications orientation within hours. Fortunately, the software demonstrated it could keep the vehicle correctly oriented, and the mission team were able to continue to refine the software and add the tasks required for MAVEN to use stellar mode for both communications and science operations.
In May, work had reached a point where the science instruments could each be brought out of safe mode and tested to ensure they had suffered no long-term damage. Then on May 28th, the order was given for MAVEN to fully transition all operations to use the stellar mode for navigation / orientation, allowing science operations to resume.
There will still be periods in MAVEN’s operations when it will have to rely on an IMU, but for now the mission team has brought the mission back from the brink of disaster, and are now focusing on ways in which the craft can better deal with possible data hiccups from the IMU systems.
Starship + Crew Dragon Update
The FAA report on the SpaceX starship facilities at Boca Chica, Texas, will now not be published until June 13th. In the meantime, it has been confirmed that the first orbital launch attempt will be undertaken by Ship 24 and Booster 7.
At the time of my last Starship update, Booster 7 had suffered a failure with a downcomer pipe, resulting in the booster being returned to the production facilities for examination, together with speculation that Booster 8 might replace it for the orbital launch attempt. However, repairs were made to Booster 7, enabling its return to the launch area.
At the end of Mays, Ship 24 was been rolled out to the test stands where cryogenic tests using liquid nitrogen commenced – only for a feed pipe connected to its LOX header tank to fail, throwing heat shield tiles off of the vehicle as the hull flexed. As a result, the pipe in question went through a rapid pipe redesign whilst on the test stand, with additional expansion joints being fitted to prevent any over-pressurisation.
With engines now being fitted to both ship and booster, and deliveries of liquid oxygen, liquid methane and liquid nitrogen being made to the tank farm, SpaceX appear confident the FAA report will give the green light for the orbital launch test – a test that will include a test deployment of Starlink satellites through the small payload slot.
Even if this first flight test is a success (which is unlikely), it is perhaps important to note it is not a prototype test flight per se, but is rather an initial proof of concept. This is because the starship vehicle is far from its final configuration (Musk has announced first possible changes to the design). Nor is Ship 24 reflective of an “operational” starship: it has no means to carry the volume of payload promised (100-150 tonnes), the mechanism(s) required to support such a mass during launch, or the means to deploy it payload bay doors and their mechanisms. As such, there is a long way to go before starship reaches an actual prototype flight, with a lot more to prove. Even then, the realities of its promise are still highly questionable – something I hope to be looking at in a future Space Sunday.