For years, Robert Beasley had been devising ways to spin pure, thin strands of glass into heat-resistant shells to protect aircraft radar.
Then, in 1962, the chemist reached a eureka moment. Working with endless formulas, he created a lightweight material that he believed could be used to perform a duty far more dangerous and demanding: shielding spacecraft from the furnace of reentry.
Today, the tiles that came out of his spare-time research have become an integral part of the world's most sophisticated space-shuttle system - and one of the most problematic. They lie at the center of the current investigation into the Columbia crash and may, ultimately, be key to how quickly the shuttles resume their journeys into space.
Almost from the start, the tiles have bedeviled NASA engineers. As early as 1979, the Columbia lost 5,000 of its 28,000 tiles merely riding piggyback on a 747 from California to Cape Kennedy - an incident that set the orbiter's maiden launch back two years. More recently, studies in 1990, 1994, and 1997 warned of the tiles' vulnerability to debris damage during takeoff or in orbit - a prime focus of the current probe. While NASA did implement many of the changes that have been recommended over the years, the silica shield that protects the spacecraft is once again the source of engineers' attention and concern.
As speculation centers on whether a piece of insulating foam, possibly mixed with ice, damaged tiles on the Columbia's left wing during takeoff, basic questions are emerging: Is there a way to shield the tiles from maurauding debris of all kinds? Do the tiles themselves need to be changed in any way? And what about the next generation of protective material for spacecraft - some of which has been going on for years and is still being shaped by Beasley's early work?
The shuttle uses several tile systems on the orbiter, each tailored to the level of heat the surface is expected to receive. But, as the current investigation is highlighting, many of the tiles have clear shortcomings. They are easily punctured or dented, which can reduce their thickness and allow more heat to find its way to the aluminum-alloy skin.
Moreover, they are bonded to felt pads that in turn are bonded to the orbiter's aluminum-alloy structure. The pads are designed to isolate the brittle tiles from the flexing and bending wings can experience on launch and reentry. Yet those bonds can be weakened if foam or debris strikes or rubs against the tiles.
Damaged or missing tiles can also upset the smooth flow of air around the wings, undercutting the shuttle's ability to maintain the proper approach angle to minimize heating during reentry. That turbulence, combined with heat-weakened bonds on nearby tiles, can lead to a zipper effect that peels tiles away from the wing.
NASA officials acknowledge the design has no margin for error. Indeed, the thermal protection system may well be the only mission-critical system that doesn't have a backup, engineers say. Once a space shuttle reaches space, tiles can't be repaired.
"That was considered early in the program," says William Kauffman, professor of aerospace engineering at the University of Michigan, but it proved impractical. Each tile is custom-shaped for its position on the wing. Even if astronauts had a safe means of performing repairs and an adhesive that could be applied in the frigid vacuum of space, a shuttle would have to carry a full set of thousands of tiles on each flight. So NASA's only strategy for keeping the tiles whole is "debris management" at the launch site, according to Ron Dittemore, shuttle program manager at Houston's Johnson Space Center.
Over the years, NASA has improved tiles and installed them on the newer orbiters, particularly on vulnerable locations such as the landing-gear doors, says Doug Perovic, who heads the department of materials science and engineering at the University of Toronto. But "Columbia had the oldest design of tiles they ever used."
If it chose, NASA could replace the tiles beneath the wing with the hardier tiles it uses on the nose and on the leading edge of the wings, Dr. Perovic says. But the modification would come at the expense of added weight on an orbiter already highly criticized for being too pricey for every pound of payload it lofts. "How much do you sacrifice weight for safety?" he asks. "That's a constant engineering dilemma."
Several new thermal protection systems are also being researched. During NASA's X-33 program, the reusable space plane was built from a special nickel alloy much more heat-resistant than aluminum - chosen to cut down on the amount of ceramic heat shields needed.
Perovic says many groups are working on materials that self-heal if gouged. Others are looking at materials that can be applied as a coating - a heat-shielding blanket that can be sprayed onto a craft.
Dr. Kauffman notes that the Russian shuttle Buran, which flew once unmanned, has a thermal-tile system that is more resilient than the Columbia's and perhaps could be adapted for use in the US.
But waiting for a new generation of thermal-protection materials is not an option if the current fleet of three orbiters is to see space again, Kauffman says. The answer may lie in changing or getting rid of the external fuel tank's insulation, or replacing current tiles with Buran-like ones.
"It almost seems that flying one of these things without modification is a death sentence," he concludes.