Let There Be Light

For Frank Cepollina ’59, it was probably the most terrifying moment in his entire career. On April 9, 1984, astronauts from the space shuttle Challenger were attempting to capture and repair the failed solar research satellite Solar Max. Cepollina had championed the mission—and the untested notion of repairing spacecraft in orbit—despite the significant skepticism of his bosses. Many in NASA as well as in the private sector doubted that such repairs could be done effectively or were worth the cost. On that day, the skeptics appeared to be right.

Not only was Solar Max spinning too fast for the shuttle’s robot arm to grab it, the spin was preventing the spacecraft’s solar panels from catching sunlight. Solar Max’s batteries were steadily draining, and within a dozen hours the spacecraft would be dead. Worse, the first effort by Cepollina’s engineers to use Solar Max’s torquer bars, designed to create a small electrical field that could interact with Earth’s magnetic field and slowly ease the spacecraft’s spin, failed because of an error in the software. New software had to be uploaded, which would take hours. Then it would take hours more for the torquer bars to neutralize the spacecraft’s spin.

As far as Cepollina could tell, the future of in-space servicing—an idea that he had been promoting for years—was about to die. Along with it would go his career.

Engineers managed to upload the software, however. And the spacecraft drifted into daylight long enough—a mere 10 minutes—for its solar panels to draw sufficient power to re-energize its batteries.

The next day, Shuttle Commander Robert Crippen flew the shuttle in formation with the satellite so that crew member Terry Hart could grab the satellite with the robot arm. Two astronauts went out into the shuttle cargo bay and successfully repaired Solar Max, installing a new attitude control module as well as new electronics.

“They had to do two EVA days’ worth of stuff in one day, and they finished it all,” remembers Barbara Scott, the Hubble Flight Software manager at the Goddard Space Flight Center in Maryland. “Everything got done!”

Cepollina and his team at Goddard would go on to lead a number of spectacular satellite rescue missions. Perhaps Cepollina’s most enduring legacy is this: He organized and in many ways conceived the spectacular repair missions that made the Hubble Space Telescope the most successful scientific instrument ever launched into space.

He is known as “Cepi” by practically everyone in the aerospace industry. He has led five repair missions of Hubble. By the time of the last shuttle Hubble repair mission, in 2009, Cepollina’s ideas of repairing and reusing spacecraft were no longer considered radical or impossible. In fact, today those ideas are about to become routine.

“Cepi is a visionary. He could always see the future better than his peers,” says Elmer Travis, who has been an engineering branch chief at Goddard during most of Cepollina’s career there. However, “Sometimes he created trouble for himself. He would do something that would turn out better in the end, but his supervisors didn’t see it as soon as he did.”

When the 1906 San Francisco earthquake hit, the banker Giannini found himself in a ravaged city with about $2 million in cash that he had salvaged from the wreckage. Giannini arranged for Cepollina’s grandfather to bring his garbage wagons into the city. They secretly loaded the cash onto the wagons and hid it under the garbage, then brought it safely out. That favor made it possible for Giannini’s bank to reopen immediately—when other banks couldn’t. That favor also got Frank Cepollina’s father a job at what became Bank of America.

As a child growing up on the farm, Cepi was tasked with maintaining the tractors. “I used to have fun, taking things apart and seeing how they worked,” he says. That didn’t always turn out well. “The mechanics would sometimes look at me and just shake their heads.”

By the time Cepi was in high school, engineering seemed to be the ideal career for him. He continued to take things apart to see how they worked—and wanted to make a living at it. His grandparents encouraged that notion. “My grandfather always used to tell me, ‘You never want to work with your hands.’ And my grandmother added, ‘You want to go to college, learn a profession!’”

His mother and father were more doubtful. “My parents told me, ‘You will never be an engineer. You’re not smart enough. You won’t work hard enough.’”

Cepi arrived at Santa Clara University in 1955 to study mechanical engineering. “I had to work my butt off,” he says. “The first two years were really tough. I can remember a lot of times working four or five hours in the lab and coming back to the dorm to immediately write my report so I wouldn’t forget, working until 10 p.m. on a Friday night.”

He also found help when he needed it. “If I had a problem or a question, I could go talk to the professor, and he would always take the time to go through it and explain. That was a great thing about going to a small university.”

In gaining his degree in 1959, Cepollina learned one crucial lesson that he would apply for the rest of his life: “Never believe the word ‘no’!”

It was also the time of Sputnik and the beginning of the Space Age. When Cepi was a sophomore, one professor invited students to attend a science conference focused on aerospace engineering. Cepi was entranced. “We saw all these rockets blowing up on the launchpad,” he remembers. “‘Oh my God,’ I said, ‘that looks like fun!’”

Another way of looking at it is: “Your dreams take shape as you are going through college. I got caught up in that era, the great space adventure.”

Upon graduation Cepollina went to work for the Army Security Agency in Warrenton, Virginia. Working there put him in contact with people at Goddard, where the 1960s’ space race was going full tilt, with Goddard designing and building most of NASA’s unmanned science probes. “This struck me as being more exciting. I thought the people [at Goddard] weren’t afraid to try new things, weren’t afraid to push new technology.” So in 1963 Cepi went to NASA and joined the space race.

Initially, most of the projects he worked on were not successful. His first NASA effort was the Advanced Orbiting Solar Observatory, which got canceled before launch in 1965. Then he worked on the Orbiting Astronomical Observatory (OAO) program, a series of space telescopes that in the 1960s yielded fairly mixed results. The first OAO failed shortly after launch. The third never reached orbit when its rocket shroud—the cone that protected it during the climb into orbit—failed to jettison.

Nor was this disappointing track record unusual. During the 1960s, approximately 30 percent of NASA’s spacecraft failed within 10 days of launch. “Some would go in the drink, with others the boosters would blow up,” Cepollina remembers. “Some would go up, get turned on for a few hours, and then die.”

By the early 1970s, George Low, then deputy administrator of NASA, was pushing the agency to find a way to make spacecraft both more reliable and less expensive. Cepollina’s boss, Joseph Purcell, put together an ad hoc committee to look into the problem. They assessed that if spacecraft were standardized, NASA could save an enormous amount of time and money.

Cepollina, with Purcell’s enthusiastic support, took this idea and quickly expanded it, conceiving of putting various systems into modular units that could be easily replaced. Components like attitude control, power, data handling, and communications were required by all spacecraft. Built as standard modules with plug-and-play electrical connections, they could be designed and tested once and then be ready for installation into any satellite.

This concept eventually became the Multimission Modular Spacecraft program, headed by Cepollina, under which a number of satellites were built in the 1970s and 1980s, including Solar Max, Landsats 4 and 5, and the Extreme Ultraviolet Explorer.

Cepollina’s managers at NASA were not so enamored of the concept. Tom Young, the Goddard director, “thought it was an interesting idea, but he was somewhat skeptical of being able to carry it off from an agency political perspective,” as Cepollina puts it. “‘You’re going to convince the agency to do what again?’”

Young’s doubts were not entirely unfounded. A repair mission carried risk, something government managers like to avoid. It also carried cost, something that would come out of NASA’s science budget, which was already fully allocated to other projects.

With NASA brass, the position was clear. “NASA headquarters was totally against it,” says Joe Rothenberg, who at the time was working for Grumann but later became Cepi’s boss at Goddard. “They felt it was a high-risk, crazy idea.”

None of this mattered to Cepollina. To him, it made no sense to let Solar Max die. The cost of building a new satellite was far greater than getting Solar Max fixed while it was still in orbit.

“One of the things that’s driven me is this concept of stretching your capital assets for as long as you can to get every dollar of return you can possibly get from it,” he says. “The American taxpayers have paid for those assets. We should use them.”

Not only did Cepollina press his bosses to find the money to fly the mission, he made sure the press knew about it. “Cepi started to announce to the world that it would require a trivial effort to have the shuttle come up and repair Solar Max,” says Rothenberg. “It was in the papers before anybody in the NASA management chain even had a chance to approve—or more likely try to discourage—the idea.”

The publicity raised questions in Congress, where members started asking NASA management why they wanted money for new missions to replace Solar Max when they could get it fixed so much more easily. “In effect, Cepi applied external pressure on the agency,” Rothenberg says.

Or, as Cepollina puts it: “Keep your nose down, keep driving the frigging car, go as far as you can, as fast as you can, make sure you get it right and do a good job, and then your guardian angel will wake up and take you the rest of the way.”

After the mission’s success, many of the bureaucratic obstacles to Cepollina’s vision evaporated. When the Hubble Space Telescope was launched in 1990 and was found to have an out-of-focus primary mirror, Cepi’s team was already prepared to provide the equipment, tools, and training for astronauts to go fix it.

It helped that NASA had spent the six years following Solar Max doing more shuttle-repair missions, retrieving two satellites for refurbishment and relaunch, and repairing two more so that their engines could boost them to the correct orbits.

The first mission to repair Hubble, in 1993, topped all these previous rescues in complexity and difficulty. The mission included a marathon of five daylong spacewalks of alternating astronaut teams of two. The spacewalkers replaced the telescope’s main camera. They installed COSTAR, which provided corrective optics for Hubble’s other three instruments. Astronauts replaced both of the telescope’s solar panels, which, as originally designed, were too flimsy and were causing the telescope to shake. They replaced Hubble’s memory units and some insulation.

Mere weeks after this spectacular mission, astronomers were hailing a repaired and fully functional Hubble Space Telescope, able to see the universe in a way humanity had never seen it before.

Subsequent repair missions to Hubble in 1997, 1999, and 2002 were as stirring. On the 1997 mission, astronauts installed two new instruments, NICMOS and STIS, replacing two of the telescope’s original instruments with more-advanced designs. NICMOS, which stands for Near Infrared Camera and Multi-Object Spectrometer, gave Hubble its first ability to observe the heavens in the infrared. The astronauts also replaced two of Hubble’s gyros and one of its three fine-guidance sensors, and they installed a new solid-state data recorder.

The 1999 servicing mission had originally been scheduled for June 2000 and was to have included the installation of a new, even more sophisticated camera. However, when three of Hubble’s six gyroscopes failed in mid-1999, the mission was split in two so that new gyros could be installed sooner. This decision was fortuitous: A fourth gyro failed in November 1999, putting Hubble into safe mode and preventing scientific research.

Launched on Dec. 19, 1999, the emergency rescue mission had astronauts replacing all six of the telescope’s gyros plus a second fine-guidance sensor. Astronauts installed a new computer, a new voltage/temperature kit for the spacecraft’s batteries, a new transmitter, and a new solid-state recorder. They also improvised the replacement of thermal insulation blankets when they noticed damage on the telescope’s outer layers. Cepollina’s ideas about doing repairs in space had been so embraced by everyone at NASA that they were making repairs that even Cepollina had never considered.

The second half of the split servicing mission was finally launched on March 1, 2002. Once again, astronauts replaced a host of equipment, including two gyros, the telescope’s main power unit, and the solar panels that had been installed in 1993. They also installed a new permanent cooling unit on NICMOS, which brought the instrument back to life after its original unit had failed.

The new refrigerator for NICMOS was another example of Cepi refusing to take no for an answer. In order for NICMOS to detect the faint infrared heat from stars, the instrument had to be cooled to -321 degrees Fahrenheit. The first NICMOS cooling system used nitrogen ice in a dewar—essentially, a thermos in space. But the system had a limited lifespan because the ice eventually evaporated. And because of a design flaw, that system lost its coolant even faster than expected.

The solution Cepollina’s team came up with: Use a different coolant that would not be lost, thus making the refrigerator unit permanent. To make it work, however, they had to build a system that could pump the new coolant; and they had to find a way for astronauts to splice this new system into the old cooling lines already on Hubble.

“You’re crazy, it can’t be done,” Cepollina was told.

For a while, it couldn’t. “Twelve times we failed,” he says. “I used to get calls at 2 and 3 in the morning after three days of testing, telling me, ‘It failed again.’”

The 13th test was the charm. With the technology proven, astronauts were able to install the first permanent cooling system on an infrared instrument in space.

After 2002, NASA planned one more shuttle-servicing mission to Hubble. When the shuttle Columbia was destroyed during its re-entry in 2003, however, NASA administrator Sean O’Keefe decided to cancel that mission. The risk of sending astronauts to Hubble didn’t seem justified. Astronomers had already made it clear that they wanted to shift funding from Hubble to the James Webb Space Telescope, already under construction.

This decision did not sit well with many people, however, especially Cepollina. He had already begun design studies for robot servicing, since many scientific spacecraft were increasingly being placed at distant locations that humans couldn’t access, even with the shuttle. Webb, for example, was going to be placed in solar orbit, about a million miles from Earth.

There was no reason to abandon Hubble, Cepi argued. If humans couldn’t fix it, robots could! And as he had done with the Solar Max repair mission, he began a campaign to convince NASA, Congress, and the world to let him fly a robot mission to do exactly that.

“Our center director said, ‘You’re nuts, Cepollina!’” he says. “I know I’m onto something when they tell me I’m nuts.” By the 2000s, it had become, as Cepi puts it, “almost a game. They say, ‘This isn’t going to work, you’re never going to be able to do it.’ I say, ‘Thanks, now I know you want it.’”

Once again, rather than take no for an answer, Cepollina approached the NASA administrator directly with his ideas and convinced him to give the OK. Cepollina also approached the press, as he had in the 1980s with Solar Max. Story after story appeared describing how NASA could service Hubble with robots. Among the results of the crusade: “[His boss at headquarters] was getting calls from the administrator to put Cepi back in a cage,” remembers Rothenberg.

Unlike Cepollina’s previous campaigns, however, this one did not succeed. And it failed for a very ironic reason. The manned shuttle-servicing concepts Cepi had helped create and prove were now what everyone favored. While robot servicing was not rejected outright, the sense was that there wasn’t enough time or money to get the mission launched. Better to fly astronauts to Hubble.

Sean O’Keefe resigned in 2006. The new administrator, Mike Griffin, quickly reinstated the manned shuttle-repair mission. For Cepollina, the loss of the robot mission was hardly a failure. He and his team were put in charge of assembling that last manned shuttle mission to Hubble. Rather than do a limited repair, which was what the robot mission would have done, they could now organize what became the most ambitious shuttle-repair mission ever attempted, fixing everything on the telescope as well as installing the latest state-of-the-art instruments.

The work was astonishing, including one repair job that required the removal of 111 screws to get at a failed circuit board. When the astronauts finished, Hubble was more capable than ever—with every single instrument that had been launched on the telescope in 1990 replaced with something newer. The telescope’s initial lifespan of 15 years had been extended to 25, with the possibility (as now demonstrated) of many years beyond. And all this because Frank Cepollina wanted to maximize the government’s capital assets for as long as possible.

Meanwhile, the repair and reusability concepts that Cepollina has been championing for decades are finally finding their way into the private sector. In November and December 2015, two different American companies, SpaceX and Blue Origin, launched rockets and—rather than allowing the first stage to fall into the ocean as garbage—successfully landed the first stages vertically, making them available for reuse. Blue Origin then flew its used rocket again, in January 2016. Those triumphs bode well for the future of spaceflight. With costs lowered, more can be achieved with less.

Neither success nor age seem to have dimmed Frank Cepollina’s vision. It’s fair to say that his tough, ethical approach to his work has enabled him to teach NASA—and the entire space industry—that there are always better ways to do things. So what’s next? He talks of humans colonizing the solar system. “I am still a fervent believer in humans in space,” he says. “Planets and suns only have a finite lifespan, even if numbered in millions or billions of years. Humans have to be prepared to move elsewhere.”

Robert Zimmerman is a science journalist and space historian who has been covering space since the mid-1990s. He is the author of Genesis: The Story of Apollo 8.