It is sadly familiar. Technical experts dig through mountains of data, examine shards of metal and graphite and epoxy, read log books, review weather reports, listen to tape recordings of radio transmissions and grill people about what they knew and when they knew it.
A formal report will point a finger in one direction or the other, everyone will breathe a sigh of relief and the space program will go on, as President Reagan has commanded.
Despite its exotic nature, NASA's space shuttle is, at its simplest, a transportation system to haul cargo from here to there. Thus it is no surprise that the investigation of the Jan. 28 disaster that claimed the shuttle Challenger and its crew of seven looks much like an inquiry the government might conduct into an airline disaster or a dramatic train wreck.
Such accidents are inevitably the result of dozens of factors in the human/instititutional system coming together the wrong way. Aircraft crash investigators call such a sequence the accident chain.
Public attention has focused on the explosion and the conjunction of physics, chemistry and engineering that caused it. A true accident chain involves more. What follows, based on facts and theories that are in the record and interviews with academic, government and industry sources, is just one of several possible chains. It includes, at least:
*Pressure on the National Aeronautics and Space Administration to produce launches on an increasingly reliable, rapid schedule.
*A result of that pressure being a subtle message of haste to the NASA world.
*Evidence of some sloppiness in NASA's performance.
*An apparent insufficiency of continuing concern for a solid-fuel rocket system deemed reliable.
*The removal of sensors to reduce weight and increase shuttle payload.
*The weather, an element man has never been able to control, and the decision to launch despite weather-related concerns.
None of these factors, taken individually, may be to blame for the explosion. The concept of an accident chain suggests that it takes many factors, some more important than others, all interrelated, to make an accident. The chances of an accident occurring are reduced as the factors are eliminated.
NASA is under great pressure, some of it self-imposed, to produce a string of shuttle launches on schedule. Money is one of the reasons.
The White House's 1982 National Space Policy says, "The first priority of the STS [Space Transportation System, the shuttle] is to make the system fully operational and cost-effective in providing routine access to space." By Oct. 1, 1988, the White House had decreed, "It is the U.S. government's intent to establish a full cost recovery policy for commercial and foreign STS flight operations."
In other words, the shuttle had to run with the reliability of a good airline and pay for itself through fees charged its customers, the very definition of a Reagan administration transportation system.
To pay its way, the shuttle needed customers. James M. Beggs, who is on leave as NASA administrator, told Congress that the fiscal efficiency of the program depended on the amount of business for the shuttle. "The larger number of times you fly, the more you can spread . . . a fixed cost," he said. That "fixed cost" includes the shuttle's standing army of 29,500 federal and contractor employes.
The cost-recovery goal was threatened for several reasons. The Pentagon, once committed to the shuttle for all its launches, wanted a backup in the event of a shuttle breakdown and was exploring the limited use of unmanned rockets, known in the space business as ELVs (Expendable Launch Vehicles). They provide launch on demand; the shuttle doesn't. "The loss of a significant part of the military market would hurt us," Beggs testified, "because this system is . . . uniquely sensitive to flight rate."
The Transportation Department has been designated to encourage U.S. entrepreneurs to launch ELVs, but as long as NASA is subsidizing prices, there is little room for private business.
Another threat came from Europe and an ELV program called Ariane. It too is subsidized, so NASA feels it cannot raise prices for the shuttle, at least not until its reliability is high.
This need to launch may have sent a message to the NASA world of government employes and contractors: We've got to make this go (but be careful).
The most ambitious launch schedule ever, calling for 15 missions, was laid out for 1986.
NASA's Aerospace Safety Advisory Panel, a group of space experts who have the role of official nag, was concerned as early as 1982 that heavy scheduling might put pressure on shuttle safety. The panel said in its 1985 report that "NASA management would be well advised to avoid advertising the shuttle as being 'operational' in the airline sense when it clearly isn't . . . . The continuing use of the term 'operational' simply compounds the unique management challenge of guiding the STS through this period of 'developmental evolution.' "
The Hartford Courant reported last week that, because of the need for a stepped-up launch schedule, NASA had recently eliminated shuttle inspections by separate inspectors and substituted inspections by the workers themselves, a common airline industry practice.
The launch schedule this winter was seriously threatened by a record seven delays in the Shuttle flight just before the one that crashed. As a result, there were only 16 days between launches, a shorter interval than ever before.
In the days before the Jan. 28 launch, as delay followed delay, NASA officials told reporters they were aware of the potential for fatigue among members of the ground crew. Seven-day weeks and long days were already a source of concern, according to many familiar with operations at the Kennedy Space Center at Cape Canaveral.
Some sloppiness crept into the system at Cape Canaveral.
Last March 8, as the shuttle Discovery was being prepared for launch, a platform to lift loads into the cargo bay fell, breaking the leg of an employe and smacking a hole in the cargo bay door. NASA's internal report detailed a series of errors including inadequate training for the platform's operators, nonexistent safety switches, mistaught procedures, an absence of inspections and other carelessness.
NASA's investigators noted, "It is obvious that some of [the] findings have implications reaching beyond this particular equipment and these locations."
On Nov. 8, one segment of a solid rocket booster not used on the Challenger flight was damaged as workers failed to use the approved procedure while removing a huge handling ring. Under the correct procedure, a crane would be attached to the ring, 132 bolts loosened, 129 shipping pins removed and the crane activated to lift the ring.
On this and several previous occasions, investigators discovered, the crane was used to ease pressure on the bolts before they were loosened and while the shipping pins were still being removed. "This failure . . . coupled with an apparent malfunction of the crane load indicating device, was the major cause of the incident," NASA's report said.
Investigators also found universal mistrust of the cranes, but few reports of the problems.
It is not known how many other things have gone wrong in the past year. NASA is not answering questions and has required formal Freedom of Information Act requests to produce documents usually available on a press room table. Airlines and airplane manufacturers often behave similarly after an accident.
Because there have been 24 straight successful shuttle launches and one of the most reliable components has been the solid-fuel rocket boosters, it became a subjective assumption that the boosters will always work reliably.
Acting NASA Administrator William Graham conceded as much last Sunday on television, when he said the boosters "were considered primary structures and not susceptible to failure."
Gary Flandro, a solid-fuel rocket expert at the Georgia Institute of Technology and former consultant to the company that manufactured the boosters, said, "When we get 24 successful flights, we feel we have experience. That had to happen at NASA as well; they wanted it to happen."
The attitude is as predictable as the expectation your car will start. However, Flandro said, "reliable doesn't mean you are free of problems, whenever you have a complicated piece of machinery."
Just as the solid rocket boosters came to be "reliable," their manufacture at a Morton Thiokol plant and subsequent assembly at Cape Canaveral may have come to be regarded as routine.
Solid rocket fuel is mixed and poured into the booster segment casings, then cured to the consistency of hard rubber. X-rays are taken to assure there are no bubbles or voids in the fuel. Similar tests check the bond between the fuel and a plastic liner and between the liner and the sides of the casing. "They do a meticulous job with that sort of thing," Flandro said.
Completed segments are shipped to Cape Canaveral, where four segments are joined, end to end, to form one booster. Assembly is a "field job." A field job, is, by definition, one with less positive control, less assurance of integrity, than a job performed under more sterile factory conditions.
Films of the Challenger just before the explosion, showed a "plume," perhaps a flame, flaring from the side of the booster about 14 seconds before the final explosion; the plume appeared to be coming from the area of a seal. A faulty seal between the segments is a likely culprit in early speculation about what went wrong.
"Personally, I believe that the grain [the fuel itself] isn't going to come apart," said Raymond E. Weich, a solid-fuel expert and the president of a manufacturing company in San Diego. "I'd be far more concerned with problems relating to delamination [unraveling of the bonds], because the integrity of the plastic-to-metal seal is what the whole thing depends upon. If the tiniest pinhole occurs, then those gases are going to work their way down through, and once it starts, nothing will stop it."
Primarily to increase the shuttle's payload by reducing the weight of the vehicle itself, some sensors that monitor solid rocket boosters were removed.
In 1981 and 1982, during the first four test flights of the shuttle, sensors were carried on the boosters that might have warned of trouble such as a burn-through. The sensors were removed, according to NASA sources, to reduce the weight of the shuttle. If that was the motive, it is as old as powered flight. Wilbur and Orville Wright had to build their own engine for the first airplane because they could not buy one off the shelf that was light enough.
Other devices, designed to keep track of pressure fluctuations within the boosters because of concern by Flandro and others about dangerous oscillations and vibrations, were tried on several flights and also removed, NASA officials have said. They have not said why.
Nonetheless, whether any warning from sensors could have saved the Challenger and its crew after the launch began is debatable.
The launch schedule encountered an unusual occurrence, subfreezing weather in Florida.
NASA continues to insist that weather was not an issue and continues to withhold routine weather data, such as the temperature at launch. However, it is known that temperatures in the area around Cape Canaveral crossed the freezing line of 32 degrees Fahrenheit several times on days before the launch and that the mission was delayed so ice could be knocked off equipment surrounding the Challenger.
Freezing temperatures make metal, plastic and solid fuel more brittle and cause them to contract. "There is water everywhere," said Weich. "Even the best plastics are generally not 100 percent immune to permeability of water, so when you go through temperature changes, the slightest amount of water in a system will behave just like anyone's water pipe [when it freezes]."
Morton Thiokol specifications say the fuel should not be used when its median temperature has dipped below 40 degrees Fahrenheit for a prolonged period.
Flandro said that solid fuel behaves differently at temperatures under 40 degrees. "I was involved in programs using solid rockets from the start," he said. "We discovered very quickly that ranges of temperatures had a major effect on performance and reliability."
An official from shuttle subcontractor North American Rockwell Corp. recommended against launch because of his concern that the subfreezing temperatures and ice on the equipment presented an unusual situation. NASA officials were concerned enough to meet with Morton Thiokol the night before the launch. A NASA spokesman said that, despite this, the agency was satisfied before deciding to launch that the weather was not a problem.
The elements were thus theoretically in place for what the engineers would call a worst-case scenario: an unsensored rocket, perhaps flawed in assembly, poised on the edge of controlled technology, undergoing an unusual freeze-thaw cycle in an institutional system pushed by financial and political problems.
What happened after the Challenger launch is both obvious and speculative. There was an enormous explosion following a breach in the wall of the right-hand booster. The flames from the breach somehow triggered the explosion. Maybe they cut like a blowtorch through the external tank and ignited the hydrogen. Maybe it was something more subtle, such as a pressure variation that overstressed the integrity of the external tank at a point when it was already under maximum strain from thrust and vibration.
The people interviewed for this article, both in and out of government, all had high praise for NASA and its commitment to safety, while agreeing that it was subject to real pressures and human failure, just like any other institution.
Lessons will be learned, as they are after aviation accidents. After the American Airlines DC10 crashed in Chicago in 1979, the nation's worst single-plane disaster, the world was told that an engine had fallen off the wing at takeoff. Everyone blamed McDonnell Douglas for designing its airplane poorly.
It turned out after several weeks of probing that the accident happened more because of a maintenance shortcut at American. The airline was trying to save time and money, and in the process damaged the engine mount. The damage grew until it produced a catastrophic failure weeks later, the deaths of 273 people.
The same shortcut had been tried earlier by Continental Airlines and the same damage had occurred. Continental discovered the damage and fixed it before returning the plane to service, but only a small part of the aviation world learned of the danger of the shortcut until after the American accident.
After the crash, changes were made to the DC10. Sensors were added to the wings so the pilot would have had a better chance of knowing what exactly had happened to his plane if the same thing happened again, because it was theoretically possible to fly after the engine fell off. A modification was ordered to the controls so they would remain in synchronization in the event of a similar failure.
Weich is philosophical about the whole investigative process, confident NASA will discover what happened and that the shuttle program will continue, despite the risks. "Anyone who climbs aboard one of those things is a real hero," he said. "They're riding a bomb and they know it. They're willing to expose themselves to these things because they feel the risk is worth it."
Teacher Christa McAuliffe and crew members Francis R. (Dick) Scobee, Michael J. Smith, Judith A. Resnik, Ellison S. Onizuka, Ronald E. McNair and Gregory B. Jarvis "never knew what hit them," Weich said. "The thing I have nightmares about is a group that gets stuck in orbit and can't get down."