Looking for What Went Wrong

By Otto Friedrich

It may take weeks or even months for federal investigators to be certain of the answer, and perhaps they will never know for sure. At least ten separate teams of investigators are now probing every aspect of the mystery of why Challenger exploded; and Acting NASA Administrator William Graham vowed they would "provide the best national capability to study this, to analyze it, to find out how to correct it, and to ensure that it will never happen again."

Although attention at week's end was focusing on a possible burn-through of the casing on one of the shuttle's two solid-fuel booster rockets, Space Flight Director Jesse Moore warned against premature speculation, saying "it will take all the data and a careful review of those data before we can draw any conclusions."

The first step was to gather every scrap of evidence that could be found, including the remains of Challenger. A fleet of 13 vessels, four planes and nine helicopters began searching an area that eventually grew to 6,000 sq. mi. of Atlantic coastal waters, picking up thousands of pounds of wreckage, including a large section of the shuttle's fuselage and the nose of a booster rocket.

Then there were all those pictures that the whole nation had seen, over and over again, and that the experts now had to study, in slow motion and with computer enhancements, over and over again. NASA not only had 80 of its own cameras filming the Challenger launch, but it impounded all the film in 90 remote-control cameras that various news organizations (including TIME) had installed near the launch pad.

Finally, there were the billions of signals sent between the doomed shuttle and NASA computers at Cape Canaveral's Launch Control and in Houston's Mission Control before and during the 73 seconds of its flight. The shuttle contained an extraordinary array of monitoring devices (sensors to detect pressures, temperatures, fuel flow, and so on), which reported their findings thousands of times a second. This flow of information, or telemetry, was so constant and so enormous that a lot of it was not sent either to the shuttle cockpit or to the consoles at Launch and Mission controls. Instead, the data that were nonoperational--that neither controllers nor crew could have done anything about--were simply stored away in computers. Thus while controllers at Cape Canaveral and in Houston apparently noticed nothing abnormal on their consoles until telemetry from Challenger abruptly stopped, the stored data could contain the clue that the space agency is seeking. Though NASA experts repeatedly objected to all public guessing about what caused the explosion --all employees of the agency were ordered not to speculate--it was virtually impossible to prevent people from doing just that. But the space agency promptly took issue with some of the early theories.

One was that ice, which had formed on Launch Pad 39-B during Cape Canaveral's 27 degrees F weather the night before the lift-off, had somehow damaged the shuttle. In fact, engineers at Rockwell International, the prime contractor for the shuttle, saw the ice in televised shots of Pad 39-B and telephoned NASA to urge a delay in the launch. But Space Flight Director Moore said that an "ice team" had inspected the shuttle. "We checked just 20 minutes prior to launch, and the consensus of the reports was good," he said. "It was decided that very low risk would be involved."

Another theory was that the external fuel tank's insulation had been damaged during a minor accident Saturday, when a derrick arm supposedly scraped the tank. But Moore insisted that the derrick had not touched the tank, only part of the launch-pad equipment. "It was a small box," he said, "a heater box that had about a quarter of an inch of insulation, out of five inches, scraped. It was a very minor scrape, and it was repaired."

Since the videotapes played early in the week seemed to show a small ball of fire suddenly appearing between one of the solid rocket boosters and the large external tank, most of the speculation centered on the possibility of a failure in either the tank or one of the boosters. Either situation could have caused the uncontrolled ignition of hydrogen gas, which needs only a spark and the presence of oxygen to explode violently. When the flame first appeared, a bit more than a minute after blast-off, the shuttle had just experienced its maximum aerodynamic stress, which a few experts thought might have caused some flawed part to crack.

A rupture of a seam in the external fuel tank is another possibility. But the inner hydrogen tank would also have to break for the hydrogen to escape, and that would be hard to explain. Then, too, a leak would have lowered the tank pressure, a change that should have showed up in readings on the controllers' consoles. Charles Donlan, a former NASA official who is now an aerospace consultant, suspects that something may have gone wrong in the fuel lines connecting the external tank to the orbiter's engines. Former Astronaut Frank Curtis Michel, now a professor of astrophysics at Rice University, speculates that trouble may have started "in the rear of the orbiter, where all of the action is going on. There's a lot of plumbing there, all of which has to be leakproof." Even though the videotapes seem to show fire breaking out in the forward part of the craft, he notes, it could have quickly reached there by moving through pipes or even the payload bay.

Technicians at the Cape thought that the forward attachment, or strut, connecting the orbiter to the external tank might have broken and severed fuel lines. Or, they say, a failure might have occurred in the barrier between the liquid hydrogen and liquid oxygen containers within the big external tank.

At NASA's Jet Propulsion Laboratory in Pasadena, Calif., however, some rocket experts suggested at midweek that the explosion had been caused by a burnthrough in the 3/8-in.-thick steel casing of the solid-fuel boosters. After watching replays of network tape of the Challenger disaster at J.P.L., officials said it looked "very much like films we've seen of burnthroughs during tests at Thiokol." (Morton Thiokol Inc., which manufactures the booster rockets at its plant in Utah, has since the 1970s been conducting tests in which the boosters are deliberately overpressurized to see how much stress they can take.)

. The J.P.L. experts interpret the tape as showing a bright sphere of flame appearing well above one of the boosters' lower skirts. It is on the interior side, facing the external tank and pointing away from the orbiter. A fraction of a second later, the sphere of flame becomes a cone-shaped jet of fire. The pointed end of the cone emerges from the booster, and its rounded end seems to aim at the fuel tank, apparently burning a hole in its side. The next thing to be seen is the huge fireball, engulfing everything.

Indeed, a near burnthrough at a different site on a booster occurred on an earlier Challenger flight, during the summer of 1983. In that case, the insulating material on the interior of the nozzle's throat was scorched away to within half an inch of the nozzle's outer skin.

Still, the way the boosters continued flying after the explosion prompted some experts to reject the likelihood of a burnthrough in either one. Hurled away from the exploding external tank, both rockets appeared to be moving rather stably, producing the awesome Y-shaped pattern that millions of Americans will never forget. A burnthrough on the side of the casing, several rocket specialists say, would have sent the booster cartwheeling wildly through space. Bob Truax, a retired engineer who directed the Thor missile program in the 1950s, agrees. "After the explosion, they were continuing on a fairly normal trajectory," he says. "Even if you get a small leak, that hole would get bigger in a matter of seconds, and you'd have hell in a handbasket very quickly."

But at week's end the New York Times reported that NASA technicians had found evidence amid the reams of telemetry that seemed to support the burnthrough theory. According to the unnamed source, the data show that the right solid- fuel booster had a pressure drop of nearly 30 lbs. per sq. in. and a loss of 100,000 lbs., or about 4%, of normal thrust about 10 sec. before the explosion --the kind of decrease a burnthrough would have caused. Later the same day, NASA released new pictures and a videotape showing what it called "an unusual plume" of flame streaking from an apparently enlarging gap in the side of the right booster immediately before the explosion. That seemed to be strong evidence.

But NASA still refused to jump to any conclusions. The agency named an interim investigative panel to take charge of the search and called upon two aircraft "crash detectives" from the National Transportation Safety Board - for help. The NTSB experts, more experienced than the space agency in reconstructing accidents, will assist in building a "fault tree": a split- second, item-by-item analysis of the flight's progress, as portrayed by telemetry, voice recordings, eyewitnesses, photographs and videotape. With NASA and industry engineers, the NTSB investigators, like paleontologists trying to reconstruct a dinosaur, will piece together every available scrap of Challenger debris--the same procedure they follow in investigating an aircraft crash. Says one former NASA scientist: "I just hope there's enough evidence to find the smoking gun."

With reporting by Jay Branegan/Washington and Jer ry Hannifin/Cape Canaveral