Grieving NASA scientists began painstaking detective work yesterday to pinpoint the cause of the catastrophe that tore apart the space shuttle Columbia as the aging spacecraft plummeted into the atmosphere at more than 12,000 mph.
They have a critical clue in the first sign of trouble during re-entry: the successive failure of temperature sensors embedded in the left wing, the left tire well and the hydraulic system that controls the left wing flaps.
"It's as if someone just cut the wire," said NASA shuttle program manager Ron Dittemore.
The fact that the sensors cut off, sending a warning message to the astronauts' display screen that they appeared to acknowledge in their last radio transmission, might mean that ferocious heat caused by air friction was burning through the protective tiles that wrap the craft.
The failing sensors could also mean trouble with the hydraulic controls, interfering with the precise, computer-controlled choreography of the flaps that steer the shuttle.
The speed of re-entry at 39 miles above Earth would mean that even the tiniest unintended flap movement would send the 90-ton, 122-foot vehicle into a lethal tumble, instantly breaking it into pieces.
NASA officials were also considering the possibility that the left wing was damaged shortly after takeoff Jan. 16, when a chunk of foam insulation peeled off the external fuel tank and struck the leading edge of the wing.
A final verdict on what caused the first fatal accident during re-entry in 42 years of space missions will not be possible for days or weeks at the earliest. "There are a lot of things in this business that look like the smoking gun that turn out to not even be close," Dittemore said.
But NASA officials and spaceflight experts said yesterday that even as the strapped-in astronauts enter the last minutes of their return home, the physical conditions outside the craft during re-entry create a dangerously unforgiving situation.
Critical flight moments
The three most hazardous moments in any shuttle flight are takeoff, the point one to two minutes later when forces on the shuttle reach their maximum, and re-entry, said Mark Lewis, an aerospace engineer at the University of Maryland who is leading an effort to design the next generation of space shuttle.
The Challenger disaster, in which the spacecraft blew apart 73 seconds after takeoff, led to intensive research on the first two critical points, but the re-entry phase has gotten less study, he said.
"I've told my students for years that the next time we lose a shuttle, it will be on re-entry because we're paying so much attention to launch," said Lewis, director of the university's Reusable Launch Vehicle Institute.
As with all shuttles landing at the Kennedy Space Center in Florida, the descent of Columbia began about 200 miles over the Pacific Ocean. The shuttle, traveling 17,000 mph, rotated tail-first and at 8:18 a.m. fired its main engines for about two minutes, slowing the vehicle.
NASA considers this "burn" the most critical moment of re-entry: If it is too short, the shuttle could hit the atmosphere at too shallow an angle and skip off into space. If the burn is too long, the shuttle could come in too steeply and be incinerated.
Once the shuttle wings bit into the thin upper atmosphere, the spacecraft began a series of S-shaped banking maneuvers to slow its descent. At that point, the shuttle was coasting like a glider - though engineers say it more closely resembles a falling brick. Had all gone according to plan, Columbia would have touched down at Kennedy at 9:16 a.m.
At Mission Control, the first signs of trouble came at 8:53 a.m., when hydraulic sensors on the left wing stopped functioning, chief flight director Milton Heflin said. At 8:56, the temperature gauge in the left main landing-gear well stopped signalling.
Those irregularities, however, did not raise undue alarm. It's not unheard-of for sensors to wink out. "The vehicle was performing fine," said Heflin.
Then, at 8:58, three more temperature sensors embedded in the left wing cut off. A minute later, the left inboard and outboard tire pressure sensors began to malfunction.
NASA flight controllers radioed:
"Columbia, Houston, we see your tire pressure messages and we did not copy your last. ..."
Columbia responded: "Roger, uh . ... "
"Then we lost all vehicle data," Heflin said.
At last contact, Columbia was at an altitude of 207,135 feet, traveling at Mach 18.3, or about 12,500 mph.
Scraps of data
The scraps of sensor data radioed back to the ground in the last minutes will be the core of the investigation. "We're getting some hints of where we need to look," Dittemore said.
Officials said it doesn't appear that Columbia's flight computers or hydraulic controls malfunctioned because the spacecraft hewed perfectly to its flight path until the moment it was lost. Nor did the age of the oldest U.S. shuttle - which first flew in 1981 and was completely overhauled two years ago - seem a likely factor in the accident, they said.
But the sensor failures did raise new suspicions about possible damage when the falling insulation hit the left wing about one minute after launch.
After reviewing videotape, NASA officials had initially concluded that the foam hadn't caused any damage. In hindsight, said Dittemore, "we can't discount that there might be a connection."
It was the second time in four months that a hunk of fuel-tank foam has fallen off during a shuttle liftoff. In October, Atlantis lost a piece of foam that tumbled into the aft skirt of one of its solid-fuel booster rockets, causing superficial damage, officials said.
Scientists were studying the possibility that the foam might have damaged some of the 22,000 thermal tiles that protect the astronauts and the delicate machinery from the intense heat of re-entry. The nose and leading edges of the wings can reach 3,000 degrees Fahrenheit, hot enough to melt iron.
The tiles, each custom-made for a specific spot and labeled with a code number, are made of carbon composite or silica-glass materials selected for their light weight and insulating qualities. But they are brittle and vulnerable to physical damage, experts said.
While some shuttle orbiters have lost a few tiles during re-entry, the loss of certain key tiles could be very serious, Lewis said: "If you lose a tile - a single tile - in a critical location, it could be catastrophic."
Dittemore noted that the moment flight controllers lost contact with Columbia was precisely when the spacecraft would have been experiencing its highest temperatures. "If we did have a structural problem or a thermal problem, you would expect to get it at the peak heat," he said.
But a solution to the mystery is unlikely to be so simple. NASA officials pointed out that while the insulation appeared to strike the leading edge of the left wing, the first sensors to fail were on the trailing edge.
Only careful analysis of the debris scattered across southeastern Texas is likely to resolve the mystery, Dittemore said.
"We really don't have any information to say there was excessive heating. All we have is evidence that says the sensors quit working," said Dittemore.
During re-entry, astronauts are dressed in protective fireproof suits and helmets. They have parachutes, but those are useless at the altitude and speed of the Columbia when it failed.
Slim chance to bail out
In an emergency at subsonic speeds and below 40,000 feet, the astronauts can blow out a hatch on the left side of mid-deck and hitch their parachutes to a 10-foot stainless steel pole that extends out the hatch. The pole guides them to a point where they are safe from striking the leading edge of the shuttle's wing, and they jump from there, dropping to a safe distance where their parachutes open automatically.
"If you have a stable vehicle below Mach 1 [the speed of sound], and there's time to put the ship into a glide, you can jump out," said former astronaut Tom Jones. "But that's a very limited capability. Most of the time when you need to get out, things are not working properly."
Because spaceflight has come to seem routine, the public loses sight of the danger inherent in every shuttle mission and the intricacy of the craft's inner workings. The Columbia contained five computers, three hydrogen-powered fuel cells to produce electricity, and 230 miles of wiring linking thousands of electrical components.
"The shuttles are very complicated machines," said Andrew S. Douglas, a professor of mechanical engineering at the Johns Hopkins University. "By some measures, they're the most complicated things ever made by man."
While the public might be blase, the astronauts themselves are well aware of the risks they run, and NASA quietly assists them with estate planning and other contingencies in anticipation of the chance that any mission could end in disaster, Douglas said.
"This is not tourism," he said.
A General Accounting Office report on the space shuttle in 2000 raised concerns that personnel cuts and an aging workforce at NASA might make it difficult for the agency to handle all the shuttle flights planned.
But Lewis, who with his colleagues is trying to design a cheaper and safer replacement for the shuttle, said the relatively small number of flights in recent years - five in 1998, three in 1999, five in 2000, six in 2001, five last year - might, paradoxically, have hurt safety.
"The fewer flights you have, the less experience people get fixing the shuttles," he said.
Still, Lewis said, yesterday's tragedy does not necessarily mean the shuttle program was flawed. Spaceflight is inherently a high-risk activity, he said.
When a friend from Boston called yesterday morning and told him the news, Lewis said: "I wasn't all that surprised. We all knew it would happen eventually. It's one of the realities of flight, especially test flight, which is what the shuttle is."
Sun staff writer Frank D. Roylance contributed to this article.