This is the second of a three-part series that looks back at the 30-year history of the U.S. space shuttle program.

Any summary of the shuttle program cannot go on without mentioning 14 lost astronauts and two doomed vehicles—Challenger on launch in 1986 and Columbia on reentry in 17 years later. Both events punctuated periods of exuberance, where the space agency pushed for the original goal of making flights to and from orbit routine and numerous; early on, NASA estimated that the system would pay for itself in five or six years if it could fly 30 times per annum. They would fall far short of their projection: In 1985 there were nine flights, the most of any year in the program. Tellingly, Challenger's loss came after Columbia had just landed from a mission 10 days earlier.

After each disaster, NASA wrung its hands, ground the fleet, set up independent commissions, analyzed its mistakes, rethought its policies, agonized over its "safety culture," and applied more stringent engineering and safety standards. Launches after Challenger were scaled down for a few years, then climbed to eight in 1992, averaging six launches a year until 2003, when Columbia was lost. After that, the average from 2005 until present (and including the upcoming flight) went down to three per year. (See NASA mission archive here.)

At the Pentagon, where there were already growing doubts about the shuttle's utility for their purposes, officials decided to pull the plug on its covert shuttle program after Challenger's loss; the military would design its hardware for expendable rockets.

The problem with the O-ring on the solid booster that sealed the fate of Challenger was addressed with a new design and fortified with launch restrictions disallowing flights if pre-launch temperatures dropped too low. But despite the new guidelines, for each liftoff of a shuttle, which lacked a robust escape system (except for a bail-out pole that supposes the failure leaves the orbiter intact and happens below 30,000 feet at subsonic speeds) there was concern. The doubts about launch were coupled with those about landing after Columbia's fiery demise. During its liftoff, a piece of the foam insulation broke off the fuel tank (a frequent occurrence during previous launches) and punctured the wing. On reentry 1,650–degree Celsius temperatures breached the ship and it broke up in the upper atmosphere over Texas—with no hope of escape for the crew. After flights resumed in 2005 with only one launch—that of Discovery to test inspection procedures and tile patching tools and techniques that had been developed in response to Columbia—an even more cautious NASA placed greater constraints on shuttle flights. Future missions would only fly to the ISS, and a complete tile inspection would be done by ISS astronauts as the shuttle rolled in front of the station before docking.

TILE SERVER: Discovery launched on July 26, 2005, on the first return to flight mission following the Columbia accident. Its main objective was to test and evaluate new space shuttle safety techniques, processes and equipment, including the Orbiter Boom Sensor System (pictured) Credit: NASA

Fear of Flying

Even flights to upgrade and repair the beloved Hubble space telescope, which had become increasingly hobbled by component failures, were problematic after Columbia. A final mission was canceled by then-NASA Administrator Sean O'Keefe, who felt it was too risky. Public pressure caused his successor, Michael Griffin, to reconsider and restore the mission. NASA feared that if reentry tiles were critically damaged, astronauts would not have an option to stay in the ISS until another shuttle could rescue them. (The agency worked around this by making an unprecedented provision: a second shuttle, Endeavour, stood on the launch pad ready to ride to the rescue.)

The growing lack of confidence in the shuttle's reliability contributed to the decision to end the shuttle era in 2010 and opt for a safer way to orbit—even at the cost of a four or five year interruption in U.S. access to space. (Launch delays in 2010 and the addition of an extra mission—STS 135—pushed the end of the program to this month.) With the culmination of the shuttle program, funds would be diverted to the development of a new space transport system capable of functioning both in orbit and beyond. First it was Constellation, which would put the U.S. back in the business of human spaceflight by 2014 and on the moon by 2020. Although canceled, it was recently resuscitated under a different name and with squishier target dates and vague mission goals.

Tomorrow, part 3the shuttle's role in solar system and deep space science