A spectrograph works like a prism, breaking light down into its component parts, allowing us to collect data about the object being observed, like its temperature and chemical composition. The public loves the incredible images we get from Hubble’s cameras, but for scientists, spectography is a vital part of the telescope’s utility. Which is why the single biggest, most worrisome task of the mission was repairing Hubble’s other spectrograph, the Space Telescope Imaging Spectrograph.
The STIS was installed on Servicing Mission 2 in 1997 and had stopped working in August 2004 due to a failed component in its low-voltage power supply board. It had been resting in “safe mode” ever since. It was a vital piece of equipment; at the time it failed, it accounted for 30 percent of the research being done with the telescope. The STIS allows us to study the relationship between black holes and their host galaxies. When it was working, the STIS allowed us to examine dying stars to understand what happens to them and why. Most important, it enabled us to study the atmospheres of distant planets with the hope of finding other places in the universe that are capable of sustaining life.
We needed to get the STIS back.
Typically, with Hubble’s scientific instruments, we never repaired them. We simply swapped the old for the new. That alone was challenging enough. But we had no replacement unit for the STIS and no budget to build one. We did have room in the budget to try to repair it. There was only one problem: The STIS was never meant to be repaired. The Hubble’s instruments were designed to survive the violence of a shuttle launch and the brutal conditions of space. They weren’t designed to be opened up. By anyone. Ever. They were sealed up as tightly as possible. Imagine all the great heist movies you’ve seen, like Ocean’s 11 or The Italian Job, where some ragtag crew of misfits gets together to break into an impregnable vault or crack the safe that can never be cracked. That’s what repairing the STIS was going to be like.
The power supply we needed to replace was housed behind a panel about 14 by 26 inches. Holding that panel down, in the top left corner, was a metal clamp held in place by two torque-set screws. A handrail that was used for installing and removing the instrument was also blocking the panel. It was held in place by four hex-head screws. The clamp and the handrail and the six screws holding them in place all had to come out.
The chief problem with doing intricate repairs inside the telescope is the risk of foreign object debris: my old nemesis, FOD. The same way you don’t want to FOD the jet when you fly a T-38, you cannot allow anything to get inside the telescope: a loose screw, a speck of dust, particles of gas. If particles of gas get inside and condense on the Hubble’s mirror, they might render the telescope useless. When you’re inside that machine, you can’t even rub your boots together, because it might give off static electricity. This is why astronauts always compare working on Hubble to performing brain surgery. Once you open the patient up, the tiniest mistake can be fatal.
The screws we needed to remove from the clamp and handrail weren’t magnetized, so they weren’t going to adhere to the drill bit when we pulled them out; there was a good chance they’d go flying off. Also, each of these six screws had a washer behind its head. When we pulled the screws out, those washers would float off, too. In addition to the washers, to make sure these bolts never, never, never came loose, the threads of each screw were covered with glue. They were literally glued into place. When we pulled these screws out, microscopic bits of dried glue would come flaking off and float away as well. And since astronauts are working with oven mitts in zero gravity and half the time in complete darkness, grappling with microscopic bits of floating debris is close to impossible.
And we haven’t gotten to the hard part yet.
Assuming the clamp and the handrail came off clean, the panel housing the power supply was held in place by 111 very tiny screws. The reason it had so many was to keep the STIS from overheating; each screw acted as a mini-radiator, allowing heat to escape and dissipate into space. Which was a rather brilliant engineering solution to the problem of regulating the instrument’s temperature, but it was a design you’d use only if you never planned to open up the instrument ever again. Every one of those 111 very tiny screws also had a washer that could go flying loose, and the threads of every one of those 111 very tiny screws were also glued in place. If that weren’t bad enough, 2 of those 111 very tiny screws were covered up by a metal plate. When we move things in space, it’s helpful to know where the center of gravity of the object is because that’s the point we want to rotate the object around. To help us, the engineers who designed the STIS put a label, a metal plate, marking where the center of gravity on the instrument is. Only now that metal plate wasn’t helping us. It was in the way and it had to be sheared off, also without creating any debris.
Then, assuming we could remove the clamp, the handrail, the metal plate, and the 111 very tiny screws, the four sides of the panel were sealed shut by a rubber gasket that had to be peeled loose. Then, once we peeled that loose and made sure no bits of dried rubber gasket were floating around, the panel was still connected to the instrument by a grounding wire. The grounding wire had to be cut without causing any electrical problems. Then, once that wire was cut, we would finally have access to the power supply itself. The power supply was a flat card, about 9 by 14 inches. It looked like any motherboard you’d find inside a computer, and it was held in place by channel locks, also known as launch locks, designed to protect the power supply from the violent shaking that goes along with launch. A launch lock consists of a long screw that, as it’s driven in, forces the metal plates holding the board to sandwich together and clamp down tight.
Then, assuming we successfully made it through the clamp, the handrail, the metal plate, 111 very tiny screws, the rubber gasket, the grounding wire, and the channel locks, we still had to remove the failed power supply board, which was seated with a 120-pin connector at the back. We had to slide that board out perfectly straight, making sure that none of these tiny metal pins broke off inside the instrument. Then we had to slide the new power supply in—again, perfectly straight—making sure every single one of the 120 tiny metal pins went in flush. If even one of them bent or broke, the whole repair would be blown. Everything up to that point would have been for nothing.
So how do you crack a safe that can’t be cracked when you’re 350 miles up in space?
The easiest part, hands down, would be removing the handrail. It took up a single, short entry in our checklist. Line 28: “Disengage handrail fasteners (four).” We were nervous about the 111 very tiny screws, which were hard to work with. But the handrail had four big hex screws with large interfaces, easy to engage. Piece of cake. Nothing to worry about. Removing the clamp would be slightly more complicated but also straightforward. We had a clamp removal tool specially designed to help pry it off and capture the screws as they came loose.
To deal with the 111 very tiny screws we designed a fastener capture plate that would fit with an airtight seal on top of this panel. The plate was made of metal, about a quarter inch thick, with clear plastic compartments that had small holes in them that lined up squarely with each of the very tiny screws. Each hole was big enough to allow a small drill bit through to drive the screw out, but small enough that the screw and the washer and the flaking glue debris couldn’t escape. The plan was to go methodically, tiny screw by tiny screw. When we were done removing them the debris would be safely contained. For the two screws covered by the metal label, the fastener capture plate had a built-in blade that, with a turn of a knob, would shear the metal off, exposing the two screws underneath.
The trick would be attaching this capture plate to the panel itself. To do that we had to remove 4 of the 111 very tiny screws from the panel to give us holes where we could drive in guideposts, stanchions, to bolt the capture plate in place. Since those four screws would be coming off before the capture plate was attached, we needed a way to capture the screw and the washer and the dried glue. Each one had to be removed using a
capture bit, a special bit on the end of our power tool that grips the head of the screw with teeth and captures it. Once we pulled the trigger to remove the screw, it would remain attached to the capture bit and we would carefully stow it and move on to the next one.
That still left the issue of the washer, which could still go flying. The plan was to insert a washer retainer, a split ring, that could be pushed over the screwhead and snap into place around it. When the screw came out, the washer retainer would hold the washer in place. Then we would use a washer extraction tool to remove the washer and the washer retainer, clearing the way for me to insert the guide studs for the capture plate. Once the capture plate was attached, we’d drive the remaining 107 very tiny screws, remove the panel, peel off the rubber gasket, cut the grounding wire, swap the old power supply for the new, and finally install a new cover panel. The replacement cover was simple. It worked as a radiator without needing 111 very tiny screws. It simply had two levers to be pushed and locked into place. Thank goodness for small favors.
When the flight was assigned, it was assumed that Grunsfeld would be EV1, leading three space walks, and I would be EV3, leading the other two. Drew was partnered with Grunsfeld and Bueno with me. Per Grunsfeld’s standing pronouncement about Hubble, the telescope knew we were coming to fix it, so new things started to break. Three months after we were assigned, the Advanced Camera for Surveys suffered an electronics failure that rendered two of its three channels inoperable. Like the STIS, the ACS needed to be opened up and given a new power supply, something that also was never intended to be done. This repair was at least simpler; accessing the inside of the instrument required the removal of 32 screws instead of 117, using a similar capture plate mechanism. Now that repair had to be developed and planned and added to the EVA schedule as well. After everything shook out, Grunsfeld and Drew were assigned the WFC3 installation, the COS installation, and the ACS repair. Bueno and I were assigned the STIS repair and the replacement of the Rate Sensor Units that housed the six gyros. Each team was assigned to handle one of the battery module replacements and a few other miscellaneous repairs.
I was glad that and Bueno and I got the STIS repair job. I wanted it. I’m not supposed to say that. I’m only supposed to talk about the team and how I was happy to play my part and all that, but if I’m being honest I have to say that I wanted that job as much as I’ve ever wanted anything. My feeling through all of this, since Grunsfeld tasked me with the robot mission, was that this flight was a date with destiny. All the challenges and obstacles I’d faced in life had brought me here, and the STIS repair—if we pulled it off—would be the most intricate, delicate, and complex task ever undertaken by a spacewalker. It was an opportunity to do something that had never been done in space before.
The STIS repair was such a difficult undertaking that the Hubble engineers built us our own model STIS to practice on in Houston, separate from the replica at Goddard and the mock-up in the pool. We had a special room across the hall from our office that we called the Annex, and we kept the model STIS in there along with space suit gloves and a full set of tools. Every moment of free time we had, Bueno, Drew, and I would be in there practicing. It was my job to do the intricate work of performing the repair while Bueno supported me and Drew ran the checklist. I don’t know how many run-throughs we did. We must have done it hundreds of times: Remove the clamp. Drive the guide studs. Remove the handrail. Attach the fastener capture plate. Over and over and over again.
Bueno was my EVA partner, but Drew was the guy in my ear. He and I went through something of a mindmeld. He’d tell me what to do, and I’d do it. It was like we had one brain. We developed our own language. This repair was so intricate and so complicated that every item on our checklist had to be broken down into a half dozen micro-steps that all had to go perfectly. There was zero margin for error. Drew would give me verbal cues from our annotated checklist, and I would execute. Then we’d do it again and again and again. A true friendship developed. I came to realize that I’d been completely wrong about him when we met. His breezy attitude came from his confidence and natural ability, but he was dedicated and conscientious and I knew he’d have my back 100 percent if anything went wrong.
We spent a lot of time talking about what might go wrong. That’s true of every flight, but it was especially true of this one. We trained and trained on the new techniques for inspecting and repairing the thermal protection system. We spent hours and hours sitting around a conference room table, doing sims for different contingencies, discussing survival scenarios, how we’d live on nuts and protein bars while waiting for rescue. There were certain failures, like some propellant leaks, that we didn’t even bother to sim; there was no way to recover from them if they happened. It was morbid and not at all pleasant, but it served to bond us closer as a crew. We knew we were doing something dangerous, but we were in it together and we’d either make it through as a team or we wouldn’t make it through at all.
We also had four rookies on the flight. On 109 we had a great crew, but we had five veterans, some of whom had flown three or four times. Flying is always more fun with rookies because everything is exciting and new. Because of the accident, they’d been waiting for years to get a flight. They knew Hubble was the flight that everyone wanted, but they also knew that, with the shuttle winding down, this might be their first and only trip to space. On my rookie flight I’d assumed I’d be back again and again. This time there was a definite feeling of Let’s appreciate this. STS-109 was a great flight, but it had been business as usual, astronauts doing what we had been hired to do. We didn’t know the shuttle program was going away. We didn’t know that this chapter was coming to a close and we needed to savor every second of it. This time, nobody took anything for granted.
Everyone at NASA was behind us, too. Mike Griffin told us if we needed anything—anything—we were to call him directly. The support we had from the Hubble team was unbelievable. Whatever we needed, we got. We needed helmet cameras that worked in the pool. We got them. They developed them for us. We needed a lighter power tool. They built us a brand-new one just to shave a pound off the weight. We needed a new mini–power tool to handle those tiny screws. They built us that. All told, we had over a hundred new tools designed and built specifically for the STIS repair alone. We were back in the Apollo days: Spare no expense, make it happen.
Every shuttle crew becomes a family, but that was never more true than with the crew from STS-125. Five of us, Megan and the four spacewalkers, had been together since the start of the development runs. Once we added Scooter and Ray J, the team was complete. We were together day in and day out for over two years, sharing an office, training in the pool, training in the shuttle simulator. We’d drive out to Ellington and get suited up and climb into our T-38s and fly in formation down to the Cape, up to Goddard, out to Ames. NASA sent us to Alaska for a kayaking trip, expedition training, bonding time. We came home from that closer than ever before. We ate dinner together every week. We were a tight group of people who clicked really well.
We knew we’d been given a chance to be a part of something special: the last great flight of the shuttle era. It was a risky, life-and-death mission, but that didn’t cast a pall over what we were doing. If anything, it kept us determined to have fun and enjoy every second of every day. There was a feeling that we were on this caper, a great adventure. Some days it felt like we were living in a movie. We were Ocean’s Eleven, the Magnificent Seven, the Dirty Dozen—a crack team brought together for one last job, one big score.
The media and the public picked up on the excitement surrounding the flight: We were in a race against time to save the most important scientific achievement of the modern space age. ABC News, the Discovery Channel, producers from Nova on PBS, they all asked to shadow us during our training to document the mission. NASA had a relationship with the people at IMAX films going back to the early 1980s and had made several films about the shuttle program, including The Dream Is Alive, Blue Planet
, and Space Station 3D, all with footage shot by astronauts on board the shuttle. Once STS-125 was on the books, Grunsfeld started talking with IMAX about following us to make a documentary about the final Hubble rescue mission. They loved the idea. Their camera people started filming us during our training and trained us to use their special cameras in space so that we could document the flight itself.
The media request that turned out to be the biggest deal of all was one that I didn’t even understand at the time. This thing called Twitter was just taking off. People were sending out 140-character status updates about what they had for dinner and such. The president had tweeted during his inauguration and apparently it was becoming the thing to do. About a month before we launched, NASA’s public affairs office approached me about being the first astronaut on Twitter. They wanted me to send out updates about our training and then send the first tweet from space.
I was happy to give it a try, but I had no idea what it would turn into. I also had no idea what one is supposed to tweet about, so I started telling people what I was doing. On April 3 we were at Kennedy for our Terminal Countdown Test, and I sent my first tweet:
In Florida, checking out our spaceship “Space Shuttle Atlantis.”
That was it. When we got back to Houston I started sending out updates a couple times a day:
In a simulator practicing for the first spacewalk on my mission
In a space shuttle simulation with my crew practicing our rendezvous with the hubble space telescope.
Practicing closing the big doors on the hubble space telescope with spacewalk instructors Tomas and Christy.
I sent out tweets from the NBL, from the shuttle simulator, from Daniel’s Little League games. I started following other people, and they started following me. They had tons of questions, and I tried to answer as many of them as I could. People asked about quarantine, about spacewalking, about our training. Mostly they asked how much I’d be tweeting from orbit: They wanted to share that experience and follow along.
Spaceman Page 25