Several Possible Causes Studied

Kathy Sawyer
Washington Post Staff Writer
Sunday, February 2, 2003

"Columbia, Houston, we see your tire pressure messages and we did not copy your last. . . . "

"Roger, buh . . . "

That was the final, broken exchange between flight controllers in Houston and Rick D. Husband, commander of the space shuttle Columbia, at the moment when, preliminary evidence indicates, the left wing came off.

The first indication of trouble arrived abruptly, shortly before 9 a.m. EST, when a series of sensor failures occurred in sequence from the trailing edge of the wing to the front. "It was as if someone just cut the wire," said an exhausted and distraught Ronald D. Dittemore, space shuttle program manager, in a briefing late yesterday.

The shuttle was passing through a period of peak heating at the time, with temperatures reaching 3,000 degrees on the leading edges of the wings.

Dittemore and other top NASA officials emphasized that no cause had been determined, and that the investigations -- three were announced yesterday -- are likely to take days or weeks. Possible causes include damage from foam insulation debris loosened during the launch, structural failure, catastrophic heating possibly after the loss of heat shield tiles, and a failure in the flight control system.

The U.S. Strategic Command in Omaha, which closely tracks space debris, said there were no signs that any foreign object had struck Columbia. But a spokesman stressed that the probe is only beginning.

A key to the cause of the disaster could come from the collection and reassembly of the shuttle debris, as in an airline crash, Dittemore said. Fragments of the spacecraft were spread over parts of at least two states -- Texas and Louisiana.

Considerable attention yesterday focused on a piece of foam insulation from the shuttle's large external fuel tank. The foam came loose and appeared to nick the protective tiles on the left wing during the launch. Columbia, the oldest in the shuttle fleet, lifted off from the Kennedy Space Center in Florida 15 days, 22 hours, 20 minutes and 22 seconds before the shuttle disappeared from NASA's telemetry.

Dittemore said earlier extensive analysis of the foam impact had indicated that it posed no threat to safety. He said there was no indication that hard metal, as opposed to foam, had detached from the tank.

It was the second time in the past three flights that a bit of errant insulation from the tank was detected in post-flight videos. In October, the shuttle Atlantis lost a piece of foam that struck the rear section of one of the two solid-fuel boosters. Analysis then indicated that no harm was done, and the flight proceeded.

Before yesterday's tragedy, flight managers had already made plans to take a closer look to determine whether these events were symptoms of a serious problem.

Dittemore said the investigation team, which mobilized as soon as the tragedy became known, will be looking intensely into that and several other failure scenarios.

"As we look at that now in hindsight . . . we can't discount that there might be a connection," Dittemore said. But he cautioned against a "rush to judgment . . . because there are a lot of things in this business that look like the smoking gun but turn out not even to be close."

There are about 20,000 tiles on the underbody and leading edges of each shuttle. If they come loose in a crucial spot or across a wide area, the craft can overheat disastrously. Once a shuttle reaches orbit, there is no way to repair any tile damage.

Columbia was not carrying its 50-foot robot arm on this mission, and therefore the astronauts were not able to use the arm-mounted cameras to inspect for damage. Flight managers had been eagerly awaiting the return of video footage taken by the astronauts during their ascent, which might have captured a detailed view of the area in question -- but that potential evidence was lost along with the shuttle.

Another possibility, Dittemore said, was a structural failure of some kind. The winged spacecraft was in a 57-degree bank at the time contact was lost -- an angle he called acceptable -- as it executed a series of "S" turns to bleed off energy and to slow down for the landing at the Kennedy Space Center expected about 16 minutes later.

"We had nothing to indicate the event that occurred. . . . We had no indication of any control problem prior to loss of data," Dittemore said. Neither was there any sign the spacecraft was off its flight path.

Officials said there was no indication of a terrorist act, and that the shuttle was too high to have been shot down by a surface-to-air missile.

They also discounted the shuttle's age as a factor. "The vehicles are kept in just pristine shape," they said.

But Seymour Himmel, a former NASA rocket engineer and former member of the agency's safety advisory panel, said: "The chief thing we were concerned about on the panel was the aging of the beast. . . . Things do wear out, and sometimes they are very subtle and you can't tell." He cited budget constraints that might have limited needed upgrades.

NASA officials clearly were determined to avoid the missteps that had compounded the fallout from the January 1986 Challenger tragedy, after which NASA seemed to react with a mixture of confusion, defensiveness and secretiveness. This time, Administrator Sean O'Keefe wasted no time in commissioning both inside and outside investigations.

NASA quickly convened a televised briefing, conducted at 3:30 p.m. EST by shuttle managers in Houston who had been on duty since about midnight. "Somewhere along the line we missed something," Dittemore said. "But I guarantee we are going to fix it."

As the day began, the mission control team at the Johnson Space Center in Houston was in an upbeat mood, looking forward to a triumphant landing after an unbelievably flawless mission, Dittemore said.

Then, shortly before 9 a.m. EST, readings from temperature sensors in the hydraulic systems on the left wing started to disappear. "They were followed seconds and minutes later by several other problems, including loss-of-tire-pressure indications on the left main gear," Dittemore said.

Milt Heflin, chief flight director, fought back tears as he elaborated on the chronology of the emerging horror.

The vehicle had been performing fine, he said, adding: "We had no indications of any problem." The sensor readings that disappeared are known to go on and off at odd times and ordinarily would be used for post-flight assessments of how the vehicle performed. A couple of minutes later, at about 8:58 a.m. EST, temperature sensors embedded in the structure of the vehicle went "off-scale low" -- or were lost.

By around 8:59 a.m. EST, about eight measurements were gone. One of them produced a message to the crew on their cockpit displays. In their last exchange, Heflin said, "[we] think they were acknowledging that measurement that they saw."

The mission control team tried repeatedly to reach the crew on at least two radio channels.

After some confusion, Dittemore ultimately concluded that there was no information in the data to indicate excessive heating. "All we have is the indication the sensors stopped working," he said.

After the accident, NASA officials impounded the data from Columbia and space shuttle hardware at plants around the country, and halted processing of the three remaining shuttles at the Kennedy Space Center in Florida, in order to preserve evidence, he said.

"We will be poring over that data 24 hours a day for the foreseeable future," Dittemore said.

The space shuttles, unlike commercial airliners, are not equipped with hardened "black boxes" that record onboard data and voice exchanges. Flight managers rely instead on the streams of flight data that flow constantly to mission control -- all of which has been preserved from Columbia's flight.

After conferring with Bush administration officials, O'Keefe, in Florida to celebrate the scheduled landing, announced that he was setting up both an internal probe and an outside, independent investigation to ensure thoroughness and objectivity. Bill Readdy, NASA's top space flight official, said a standing contingency plan went into effect immediately, including a search-and-rescue effort.

The investigators -- with the aid of the Federal Emergency Management Agency, the FBI and other agencies, as well as local law enforcement -- were scrambling to set up a debris collection center in northeast Texas, and to coordinate the daunting task of analyzing the fragments.

The independent investigation will include a variety of experts from the Air Force, the Navy, the National Transportation Safety Board, the Department of Transportation and other federal agencies.

A command post was established at Barksdale Air Force Base in Louisiana. The NTSB was planning to examine radar and voice communications data from air traffic control facilities in Texas. Although the shuttle breakup was too high to have been recorded on radar, radar might have picked up some of the showering debris.

Rep. Sherwood L. Boehlert (R-N.Y.), chairman of the House Science Committee, announced that his staff will also investigate.

The White House was considering whether to create its own independent panel, but officials said no decision had been made.

View all comments that have been posted about this article.

© 2003 The Washington Post Company