The ultimate 4-wheel-drive: How ESA's keeping XMM-Newton alive after 20 years and beyond
You thought that yoghurt in the back of fridge was time-expired? Behold X-ray boffinry YEARS past its design-life
Space Extenders Sure – that telescope can be serviced by Space Shuttle astronauts. But how do you keep one running for years past expiration without a prodding by spacewalkers? Behold ESA's XMM-Newton.
ESA's X-ray Multi Mirror (XMM) telescope, named for English physicist and mathematician Sir Isaac Newton, was launched on an Ariane 5 back on 10 December 1999. Funded for an initial two-year mission, with a ten-year design-life, the biggest science satellite built in Europe (at the time) is now entering its third decade of service.
The XMM-Newton story, however, begins a little earlier with an astronomy mission proposed in 1982. A number of working groups were established in 1985, and by 1988 XMM was approved by ESA as a Cornerstone mission in the Horizon 2000 programme.
They do it with mirrors
The spacecraft itself, weighing in at 3.8 tonnes and measuring 10 metres in length, comprises three X-ray telescopes, each with 58 high-precision concentric mirrors arranged to capture the greatest amount of X-rays possible. In addition to the telescopes, XMM-Newton is equipped with three European Photon Imaging Cameras (EPIC), a Reflection Grating Spectrometer (RGS) and an Optical/UV Monitor (OM), which observes the same regions as the X-ray telescopes, but in UV and visible wavelengths.
With mirrors among the most sensitive ever developed, XMM-Newton's mission is to provide the data to enable scientists to solve cosmic conundrums; from what happens in and around black holes to the formation of galaxies.
The spacecraft has a limited amount of onboard automation and is controlled by the European Space Operations Centre (ESOC) via a live ground station connection.
XMM-Newton's first brush with a mission-ender came in October 2008, as it approached the end of its design-life. The spacecraft was approaching the perigee of its highly elliptical, 48-hour orbit of Earth when ESA controllers lost contact.
After procedures to recover the spacecraft failed, there were fears that some sort of catastrophic event, such as a collision, had occurred or a malfunctioning thruster had sent the observatory into a tumble. These were assuaged when ground-based astronomers spotted that the sunlit spacecraft was stable in its expected orbit.
ESA's 35-metre antenna in Australia – using a mode developed for deep space missions – eventually detected a weak signal from XMM-Newton and recovery could begin.
Despite the feeble strength, controllers theorised that the issue was a stuck Radio Frequency (RF) switch. Following simulations, the team used NASA's 34-metre Deep Space Network station in Goldstone, California to send a command to set the RF switch back to its last working position. Communications were re-established, and the switch has not moved since.
The spacecraft's deputy ops manager, Dietmar Heger, said at the time, in a bit of an understatement: "It's been a thrilling moment for our team."
More thrills would be in store as the mission continued to a doubling of how long anyone might have reasonably expected XMM-Newton to last.
Four-wheel drive in Spaaaaaace
XMM-Newton's current spacecraft operations manager, Marcus Kirsch, moved into the role from the science side of the spacecraft around 2008. He arrived in time for the next issue: fuel. The spacecraft was showing every sign of enduring years beyond its design-life, but fuel would soon present a problem.
XMM-Newton, like many spacecraft, uses reaction wheels to control its orientation. The telescopes peer at a given point in space for a given period of time and then move to another in a manoeuvre called a "slew". The thrusters are not used for these orientation changes since the tanks would soon be emptied, instead, explained Kirsch, "we have wheels on board and these wheels are sitting on a tetrahedron. So, they're getting three dimension and angular momentum. By just spinning them in different speeds, you can move the spacecraft."
However, the wheels "have to be unloaded every now and again, and for this you need fuel."
After launch, the fuel usage was in the order of 6 kg per year.
XMM-Newton was launched with plenty of fuel for the original mission, but by 2009 there was every danger of running dry before 2018. In addition, ESA had moved on from the three-wheel model in favour of four in subsequent spacecraft: "This gives them much more possibilities," explained Kirsch. "Imagine that you have a three-dimensional problem which you can solve in four spaces. So, it makes things much easier."
And, perhaps more importantly, "When you have four wheels instead of three, you need much, much less fuel for unloading."
XMM-Newton actually had a fourth wheel, but it was a spare, to be brought into service when one of the others failed.
It was over beers, as most of the best brainstorming sessions often are, that a contractor put the idea to Kirsch, who recalled being told "Ah! With four-wheel service we would be much better... and we should try this, but ESA never liked to do this."
After all, the observatory had a good few years of life left in it – why take the risk? It was a running spacecraft, working well and with a decade in orbit under its belt. If something is working, for goodness sake don't fiddle with it, especially not the reaction wheels.
"You would never change your wheel on the car while running. Right?" he laughed.
Back at the bar, Kirsch replied, "Let's do a study. And if you can convince me that we really would save fuel, then we can think about implementing some software changes to get this going."
The result was that "you can save 50 per cent of the fuel if you operate a fourth wheel on top," remembered Kirsch.
"Nobody thought about this before," he told us, "and we had a life horizon of maybe another eight, nine years."
The change could add a possible 10 more years on top, potentially moving the end of the mission to beyond 2030.
"So we did some studies, we involved industry and we did various things and it came out that we can change the onboard software to activate the fourth wheel and to operate in this so called four wheel drive."
It took some time for the team to persuade the bigwigs that the idea was a good one: "In the beginning," laughed Kirsch, "it took a while, you know, to overcome the inertia: 'we never did it, we will never do it, blah, blah, blah...' But then, on the other hand, they said, 'Well, 10 years more? We can't say no to 10 years'..."
Indeed, the mission had a lot to lose if things went wrong, but so much more to gain. The science community was keen, the team was keen and so the XMM-Newton gang went ahead.
As with all the ESA operations teams we've spoken to, Kirsch was full of praise for the spacecraft manufacturers, originally Dornier and now Airbus (after the usual round of acquisitions and mergers.) It was trickier, however, to put the band back together: "The software guy was already retired," Kirsch recalled, "and we got him back from retirement."
With the mission potentially lasting for almost another 20 years, the team decided to have a younger member do the work, with support from the older Brit who had been returned from retirement. "He provided the perfect support," said Kirsch. "He's one of these computer freaks, who say, 'Oh, yeah, yeah, this change, it must be in code line 4370...'"
While not a huge change to the code, the modifications were significant. "Most of the effort," recalled Kirsch, "went into testing."
The team had an emulator and simulation hardware available. Airbus also proved helpful once again: "We even reactivated some old simulators at the contractor site at Airbus in Stevenage in UK," he remembered.
A few short years after the idea was had, the team was ready to update the spacecraft.
Kirsch remembered saying to one of the team from industry, having invited some of those involved over to his home for paella: "You know, this is really dangerous. If we lose the spacecraft, we don't have an X-ray observatory anymore. I mean, you are convinced, but I'm very nervous.
"And he said, 'No, you should not be nervous, it will work fine. It will work'."
Kirsch was still, however, twitched about the whole thing. His team was a little more relaxed, having been closer to the actual code and testing.
"If you're more on the management side," he told us, "you have to trust the people. It's not that you can control everything yourself.
"You have to trust the people."
Despite all the documentation, validation plans and test campaigns, Kirsch told us: "I can tell you I was nervous until the first moment where the wheel was spinning up. And I was still nervous for a week after that."
He added: "It's not easy to make such a decision."
As it transpired, it was the right decision. With "four-wheel drive," Kirsch told us, "our extrapolations go until December 2030."
The change, implemented in 2013, saw the annual fuel consumption halved.
The propellant could be eked out even longer if safe mode operations, where the spacecraft uses its thrusters for orientation, are avoided. Each safe mode can cost up to half a year of operations, and so is factored into calculations: "We allow for one safe mode per year in terms of fuel budget. So, every time where we have no safe mode in one year, we gain a little bit."