The New York Times, by Staff

By Richard Witkin

A study by the space agency shows that a solid-fuel booster rocket of the shuttle Challenger experienced a sudden drop in power about 10 seconds before the spacecraft exploded, a source close to the inquiry said yesterday.

Officials believe that the most plausible explanation for the power drop would be flames burning through the side of the right-hand rocket, the source said. NASA officials were said to believe that flames shooting out of the booster burned through the skin of the external fuel tank, to which the booster rocket was strapped, and through the liquid oxygen line inside.

Began With Fire

Fire erupting in the external fuel tank was believed to have led to the explosion that took the lives of the seven astronauts on board, the source added.

At liftoff the external fuel tank holds 500,000 gallons of liquid hydrogen and liquid oxygen, which when mixed together are very explosive. Under control, combustion of the explosive mixture propels the rocket.

Additional data supporting the theory that the oxygen line was ruptured were sent back from the Challenger, the source said. The data showed, he said, that the three main shuttle engines, which burn liquid fuel from the external tank, shut down almost immediately after the drop in the booster rocket's power.

The data sent from the spacecraft and recorded on the ground were said to have showed specifically that the solid booster suffered a pressure drop of just under 30 pounds per square inch and a loss of about 100,000 pounds of its 2.5 million pounds of thrust.

Steering Correction

In addition to the three engines of the shuttle stopping immediately, the source said, the ground data show that the nozzles of those three engines and of both solid-rocket boosters swiveled to one side as the steering mechanisms sought to correct the course change produced by the uneven rocket thrust.

Jim Mizell, a NASA spokesman who worked on the shuttle program for many years as an engineer, said the space agency would not confirm or deny the report. ''I haven't heard that, but it is exactly the kind of thing the review board is looking at,'' he said. ''But we cannot confirm whether this particular piece of information is correct at this time.''

The information raised the question of why the problem was not detected by controllers in the National Aeronautics and Space Administration's control room at the Johnson Space Center in Houston before the explosion.

The source said the drop in power was subtle, representing only about 4 percent of the power of the solid rocket. The percentage loss was so small, the source suggested, that controllers might have thought indications of it were routine electronic fluctuations. Only when the stored data were analyzed after the fact by space-agency technicians could the precise readings be refined, he said.

Of the wealth of data sent back from the shuttle, only a tenth is monitored by ground controllers during the launching. The rest is saved on computer tapes, and it is this archival material that is one of the main focuses of the NASA review board.

No Escape Possible

In any case, the Challenger was at a point in its flight where, apparently, absolutely nothing could have been done to avert disaster, the source said.

It was emphasized that it would take more analysis to determine the exact sequence of events leading to the explosion of the Challenger. But evidence recorded on the ground pointed to the solid rocket as the place where the disaster got its start.

''The space agency assumed that solid rockets were 100 percent reliable,'' the source said.

The liquid oxygen that combines with liquid hydrogen to power the shuttle's three main engines is contained in the top third of the external fuel tank. The shuttle is joined to the fuel tank at two points, front and aft. The two solid-fuel boosters are attached to either side of the tank.

The television pictures of the Challenger explosion seen by the public show the right booster on the far side of the camera, behind the shuttle, which had rolled on its back. NASA is examining photographs taken from other angles, and it was reported that one view suggested a jet of flame shooting from the side of the right booster rocket.

Both the liquid oxygen and the liquid hydrogen are fed to the three main shuttle engines through pipes from the bottom of the tank through the aft attachment point. The liquid oxygen reaches that attachment area through a long pipe running from top to bottom under the skin of the main fuel tank.

The space agency's investigation team was reviewing more data before stating publicly the precise sequence of events that might have precipitated the explosion.

''They want to be ironclad sure'' the informant said. He said this was probably because of the extreme sensitivity of the agency, its contractors and the public on the issue of where the blame might lie for what was the worst catastrophe in the brief and spectacular history of manned spaceflight.

By David E. Sanger

The manufacturer of the space shuttle Challenger's solid-fuel booster rockets disclosed yesterday that they were not designed for use when the temperature of their propellant was below 40 degrees Fahrenheit.

At the time of the launching Jan. 28 the temperature at Cape Canaveral was at the freezing mark, after a night in which the Challenger had stood in sub-freezing weather awaiting the launching.

It is unclear whether the propellant had become as cold as the surrounding air. An engineer with the National Aeronautics and Space Administration said yesterday that the rocket's interior did not have a temperature gauge to inform ground controllers about the condition of the propellant.

Concern About Vibration

The news came amid reports that as long as two years ago some NASA engineers were concerned about dangers posed by vibrations and rapid pressure changes in the rockets. According to several reports that NASA officials have refused to confirm, an unusually sensitive pressure meter was placed in the boosters for several shuttle flights starting in 1983. The meter was later removed, and agency officials have not made public the results of that study.

The manufacturer's disclosure yesterday refocused attention on whether the launching site was too cold for the Challenger to take off safely.

It also appeared to give credibility to the theories of several rocket scientists that a combination of factors - including cold that may have cracked the rocket's solid fuel, and strong pressure changes that may have widened the crack - could have led to a ''burnthrough'' of the rocket's thin skin.

That flame, in turn, could have touched off a larger explosion of the adjacent liquid-fuel tank, consuming the Challenger and its crew.

''There are more and more people coming to the conclusion that it was awfully cold out there,'' said Edward Price, a professor of aerospace engineering at the Georgia Institute of Technology, a leading authority on solid-fuel rockets.

Scientists say they have known for some time that solid-fuel boosters, which were not used on manned space flights until the shuttle program, are particularly plagued by vibration. But some say the problems are especially acute in the shuttle's boosters, because of its size and construction.

''The oscillations are a serious issue,'' said Dr. L. King Isaacson, a professor of mechanical engineering at the University of Utah, not far from where Morton Thiokol Inc. builds the shuttle's solid-fuel rocket segments.

The connections between the booster's four main segments, he said, can create ''pressure barriers'' that result in a wave of pressure resonating through the rocket, much like the continuous waves in organ pipes. Questions on Temperature

Since the morning of the disastrous launching, numerous questions have been raised about the unusually low temperature at the site, one of the few factors that distinguished this flight from the 24 shuttle missions that preceded it. But most of those questions have focused on icicles that formed near the shuttle.

The day after the accident, however, Jesse Moore, NASA's associate administrator of space flight, said, ''We were confident that there were no problems relative to the solids,'' a reference to the solid-fuel booster rockets.

Since Saturday night, however, when NASA published photographs of a ''plume'' escaping from the skin of the shuttle's starboard rocket, near a seam, attention has shifted once again to the possibility that the cold weather could have weakened the bonding agent that holds the granules of propellant together.

A crack in the propellant could have disastrous consequences, experts say, because it could create a channel through which the hot gases rushing through the rocket could have reached its skin.

Operating Temperature Range

At Morton Thiokol, which has generally refused comment on the accident, a spokesman, Gilbert Moore, said yesterday that the storage temperature of the solid propellant ranges from 32 to 95 degrees Fahrenheit, and that the operating temperature ranges from 40 to 90 degrees.

Those temperatures are for the core of the rocket, not the outside air, he said. ''The atmospheric temperature can vary a great deal without having any significant impact on the temperature of the motor,'' which has ''insulative qualities,'' he said.

Among the many unanswered questions surrounding the condition of the solid boosters is why its internal temperature was not monitored by NASA ground controllers, whose computers survey thousands of critical sensors before and during a shuttle flight.

''I can't answer that,'' said Mr. Mizell, the NASA spokesman. ''Either there was a decision about that made somewhere, or there was a miscommunication with Morton Thiokol.''

Test Results Unknown

Also unclear yesterday were the results of the tests on the solid-fuel booster that were conducted in response to fears by several scientists, including Prof. Gary Flandro of Georgia Tech, about pressure oscillations in the rocket boosters.

Ordinarily the pressure in the 149-foot-long boosters is around 600 pounds per square inch. But even seemingly small changes of one to four pounds per square inch, experts say, can affect the rocket's thrust by up to 100,000 pounds. Each booster rocket has a thrust of about 3.2 million pounds at launching.

Such variations can be particularly harmful when the rocket is going through the period of maximum dynamic pressure, the stage of ascent that the Challenger was just leaving when the plume appeared at about 58 seconds into the flight.

''The oscillations can put enormous stress on the supporting structure,'' Professor Flandro said. ''I don't think that alone they caused the breach, but they could have made it worse.''

By Philip M. Boffey

The space agency was warned last year that seals on the space shuttle's solid-fuel booster rockets might break and cause a catastrophic accident, according to documents from the agency's files.

The documents show that engineers at the headquarters of the National Aeronautics and Space Administration and its Marshall Space Flight Center in Huntsville, Ala., were concerned that leaks might occur where segments of the booster rockets are mated.

Such leaks would allow hot gases and flames to escape through the side of the rocket instead of through the nozzle that channels the gases out the rear, possibly causing severe damage to the shuttle or an explosion, according to space experts. A Stark Warning

One NASA analyst warned in an internal memorandum last July that flight safety was ''being compromised by potential failure of the seals.'' He added: ''Failure during launch would certainly be catastrophic.''

A 1982 ''critical items list'' for the booster also warned that if the seals should fail the result could be ''loss of vehicle, mission, and crew due to metal erosion, burnthrough, and probable case burst resulting in fire and deflagration,'' or rapid, intense burning.

It is not clear what action if any, NASA might have taken in response to the warnings in the documents, but the issue was listed as a matter of concern in agency documents as recently as December.

Space agency officials declined today to respond to questions about the concerns raised in the internal documents.

Two Officials Are Informed

David W. Garrett, chief of the agency's news and information branch, was informed of the substance of those documents and said he had informed both L. Michael Weeks, deputy associate administrator of the office of space flight, and David L. Winterhalter, acting director of the agency's shuttle propulsion division. He said they had declined to comment.

Jesse W. Moore, associate administrator for space flight, the agency's top shuttle official, did not return a telephone call to his home, although a family member said he was there.

The internal documents describing problems with the seals were made available to The New York Times by a solid-fuel rocket analyst who has worked closely with propulsion engineers from the Kennedy Space Center in Florida, which assembles the booster rockets; the Marshall center in Alabama, which is responsible for their design, and NASA headquarters in Washington. The Leading Theory

Although no one knows exactly what caused the explosion that destroyed the Challenger Jan. 28, space agency officials have said that the leading theory, based on films of the flight, is that a plume of flame emerged from one side of a booster and set off an explosion of the shuttle's giant external fuel tank.

Space officials have said they cannot identify precisely where the plume emerged and thus do not know whether it burned through a seam or through the metal side of the rocket. ''It did appear to happen at least near a seam,'' Dr. William R. Graham, Acting Administrator of the space agency, said last Sunday. He said the plume appeared to start ''near one of the field joints'' but that measurements had not yet established whether the plume occurred ''at the seam or just near the seam.''

The safety of the seals also became an important issue Thursday at the first meeting of a Presidential commission that is investigating the causes of the accident. The space agency acknowledged that it had consulted with the rocket's manufacturer, Morton Thiokol Inc., about concerns that cold weather at the launching site might have weakened the seals. But an agency official told the commission the manufacturer had concurred that the launching should proceed.

The official, Judson A. Lovingood, deputy manager of shuttle projects at Marshall, also acknowledged that there had been concern after previous shuttle flights about erosion damage to some of the seals, but he indicated that this problem had been thoroughly investigated.

The seals are needed because the booster is not a single long structure but rather four large cylindrical segments that are bolted together, along with other components, at the Kennedy Space Center when the rocket is being prepared for launching. Although the side of the rocket may look leakproof to the naked eye, there is room for gases to escape at the seams. Thus rocket engineers have devised a series of seals and other barriers to keep the gases in.

The two most important seals are O rings, essentially large doughnut-shaped pieces of synthetic rubber that fill the tiny gap between two cylindrical segments that are bolted together. The O rings are themselves protected from heat and flame damage by an initial barrier of putty.

Tighter Under Pressure

If flames and hot gases are to escape through the joint between segments of the rocket, they generally must first pass through the putty, then through the primary O ring, and finally through the back-up ring. The rubbery O rings are designed to seal especially tight when they are hit by the high-pressure gases, much as a rubber washer on a faucet seals tight to prevent water from leaking.

At the Presidential commission's meeting Thursday, Mr. Lovingood, from the Marshall space center, was asked if experts had looked at the joints in the reusable boosters after previous shuttle flights to see if there was any evidence of leakage. ''We have seen some evidence of erosion of those seals, the primary seal,'' he said. ''We've never seen any erosion of a secondary seal. But we have seen evidence of soot in between the two seals.''

When asked if this was a cause for concern, he replied: ''Oh, yes. I mean that's an anomaly and that was thoroughly worked, and that's completely documented on all the investigative work we did on that.''

The possibility that cold weather might weaken the seals, by causing shrinking or stiffening or through some other effect, is not explicitly addressed in the internal documents. Instead, those memorandums focus on erosion and heat effects observed on the seals after previous flights. One memorandum does suggest, however, that ''environmental effects such as moisture'' could be an indirect factor in causing erosion.

A memorandum prepared within the comptroller's office at NASA headquarters last summer used dire terms to describe the potential problems of charring and erosion that might damage the effectiveness of the seals. The memorandum, dated July 23, 1985, was addressed to Michael B. Mann, head of the resources analysis branch for the shuttle program, from Richard C. Cook, a subordinate.

Ring Was 'Destroyed'

Mr. Cook warned that ''the charring of seals,'' which had been observed on recent shuttle flights, posed ''a potentially major problem affecting both flight safety and program costs.'' In the joint between the nozzle section of the rocket and the adjoining segment, the memorandum said, ''not only has the first O ring been destroyed, but the second has been partially eaten away.'' The memorandum did not say how often this had occurred.

The joint referred to in the memorandum is the one nearest the spot where the plume of flame was seen to emerge from the side of the rocket just before the explosion that destroyed the Challenger.

The memorandum said the cause of the erosion problem had not been determined. ''There is little question, however, that flight safety has been and is still being compromised by potential failure of the seals,'' it said, ''and it is acknowledged that failure during launch would certainly be catastrophic.''

The memorandum said the leadership of the space flight program ''is viewing the situation with the utmost seriousness.''

Another memorandum prepared at roughly the same time by Irving Davids, an engineer in the shuttle rocket booster program at NASA headquarters, described a visit he made to the Marshall Space Flight Center on July 11, 1985, to discuss ''seal erosion problems'' that had affected the O rings on several shuttle flights. A Dozen Instances

This memorandum said there had been ''12 instances during flight'' where there had been some erosion of the primary O ring at the seam where the nozzle segment of the rocket is bolted to the adjacent segment.

The memorandum said that in two cases soot actually blew by the primary seal, and in one case the backup seal showed erosion as well. Both this observation and Mr. Cook's memorandum appear to contradict Mr. Lovingood's assertion that no erosion of a secondary seal had been observed.

The document added that the prime suspect in causing the erosion was the type of putty used. It said Morton Thiokol, the manufacturer of the booster rocket, believed that the putty, made by another, unidentified manufacturer, could develop holes under certain conditions and that these holes would have a ''jetting effect,'' an indication, apparently, that the holes could focus hot gases on the seal. ''There doesn't appear to be a validated resolution as to the effect of the putty,'' Mr. Davids wrote.

The memorandum also described erosion of the O rings at the joints between other major segments of the rocket. It said there had been five occurrences during flight where the primary ring showed erosion and one case where the backup ring was affected by heat although not actually eroded. Problem With Back-Up Ring

One critical problem, it added, was that rotational forces generated as pressure builds up within the rocket caused a ''lifting off'' or ''unseating'' of the secondary ring, a problem which ''has been known for quite some time.'' One proposal for eliminating this problem, the memorandum said, was a ''capture feature,'' not otherwise described, which would prevent the seal from lifting off.

The memorandum from Mr. Davids was addressed to Mr. Moore, the associate administrator for space flight.

Through the rest of the year, the O rings continued to be a concern to some engineers and budget analysts.

On Aug. 21, 1985, a budget briefing prepared for top-level NASA officials listed charring of the rings as one of the top ''budget threats'' to the solid-fuel booster program, apparently a reference to the fact that fixing the problem could become costly.

Other Recent Reports

On Sept. 10, 1985, a status report and briefing prepared by NASA's propulsion division said that the most recently completed shuttle mission showed ''one minor erosion'' on the primary ring at the joint between the nozzle and the adjacent segment but no such damage at the other joints. It also listed charring of the rings at the top of a list of ''solid rocket booster issues.''

In December 1985, a monthly status report again listed ring charring as one of seven issues regarding the booster.

Concerns about the seals had been expressed in agency documents at least as far back as 1982. A ''critical items list'' for the solid-fuel booster rocket, dated Dec. 17, 1982, described the joints as in the most important category.

The document also said that ''joint rotation'' as the pressure rose in the rocket might knock out the back-up ring. This was the same problem that, according to Mr. Davids's memorandum, had still not been solved in July 1985, although it had been ''known for quite some time.''

By David E. Sanger

The top Morton Thiokol engineer present at the Kennedy Space Center before the Jan. 28 liftoff of the space shuttle Challenger said tonight that he had argued for hours with space agency officials not to launch the craft because of low temperatures.

He said that he persisted even after his own superiors had overruled him and given the agency a go-ahead.

The engineer, Allan J. McDonald, a 26-year veteran of Morton Thiokol Inc., which made the solid-fuel booster rockets for the shuttle, said that at a closed session last Friday before the Presidential panel investigating the explosion he recounted his ''somewhat heated'' exchanges with officials of the space agency.

Notes to Commission

He said in a telephone interview tonight that those exchanges centered on the rocket seals that have become a major suspect in the explosion that killed the shuttle's seven crew members.

Mr. McDonald, who is the director of Thiokol's solid-fuel rocket motor project, also said he turned over to the commission detailed notes made in the course of his dispute with the officials of the National Aeronautics and Space Administration.

He said he first warned NASA officials about potential dangers after calculations performed by Morton Thiokol engineers in Utah showed that the temperature of the O rings, which seal joints in the booster rockets, was about 30 degrees Fahrenheit. That is about 23 degrees lower, he said, than the temperature of the rings in a January 1985 shuttle launching that resulted in the largest amount of ring erosion ever seen by NASA officials.

Seals Suspected in Disaster

The O rings are the leading suspect in the apparent failure of the right-hand solid-fuel booster rocket. It is now believed that they may have set off events that led to the fireball that consumed the Challenger.

In testimony before a Senate subcommittee today, Jesse W. Moore, the space agency's top shuttle official, said he was not told about low temperature readings on one of the rockets the morning of the launching and would have asked more probing questions had the issue been brought to his attention. [ Page B9. ] Mark Weinberg, a spokesman for the Presidential commission, said, ''It would not be appropriate'' to say whether Mr. McDonald had appeared before the commission in its closed session at the Kennedy Space Center last week. Mr. McDonald said it was likely he would be called to appear at an open hearing of the commission next week. The commission has confirmed that it took testimony from several Morton Thiokol officials on Friday.

NASA officials said today that technical crews on the launching pad, using infrared sensors, had measured temperatures as low as 7 to 9 degrees Fahrenheit on the surface of the right-hand booster rocket, near the O ring seals that are suspected of failing.

A breach of the O rings would allow superhot gases to escape through the side casing of the 149-foot-long rocket, an event that NASA officials had previously considered unlikely, although some NASA engineers had warned more than a year ago that such a breach could be catastrophic.

Mr. McDonald said today that on the night of Jan. 27 he was arguing primarily with Lawrence B. Mulloy, who heads NASA's solid-fuel booster rocket project at the Marshall Space Flight Center in Huntsville, Ala.

''It was a very prolonged discussion, Mr. McDonald added. ''The engineers in Utah were largely in agreement with me.''

Nonetheless, he said, his superior, Joseph Kilminster, overruled his objections around 11:30 P.M. and transmitted to NASA a copy of a letter approving the launch. But Mr. McDonald said he believed the situation was so serious that he continued arguing his point even after the letter had arrived at Cape Canaveral.

''I argued before and I argued after,'' he said by telephone from his home in Ogden, Utah. ''The low temperatures make the O ring seals much harder, stiffer, and it caused them to shrink. It is hard to quantify, but qualitatively that is what happened.''

Mr. Mulloy did not return a message last night left on an answering machine at his home in Huntsville. There was no answer at Mr. Kilminster's home in Utah.

Mr. Mulloy has testified before the commission on the effects of the cold on the O rings. Mr. Kilminster has not, but his name has been mentioned before the commission in discussions about Morton Thiokol's agreement to go ahead with the launching.

Testimony Before Panel

In testimony before an open session of the Presidential commission last week, Mr. Mulloy, confronted with an impromptu test of the resiliency of the O rings by a panel member, Richard P. Feynman, said tests indicated that the rings lost resiliency as the temperature dropped toward 50 degrees Fahrenheit. But he added that specifications indicated they could operate safely down to 30 degrees below zero.

In his testimony Feb. 11, Mr. Mulloy said that data presented on the night before the launching by Morton Thiokol indicated that the resilience of the seal drops even further at temperatures of 20 or 25 degrees Fahrenheit. However, Mr. Mulloy said that ''under the conditions that we would see on launch day, given the configuration that we ran, that the seal would function at that temperature. That was the final judgment.''

He added, ''Ah, there were data presented as we have discussed, by some, by Thiokol engineering, that there was a suggestion that possibly the seal shouldn't be operated below any temperature that it had been operated on previous flights.''

Mr. Mulloy also said last week that Morton Thiokol had originally recommended against launching but reversed that position later the evening of Jan. 27.

At NASA headquarters tonight, Charles Redmond 3d, a spokesman, said, ''Clearly, the Morton Thiokol engineer has a point of view he is allowed to make, but beyond that I don't think we have any comment.''

At least some of Mr. McDonald's comments today appeared to contradict Mr. Mulloy's statements last week concerning the relationship between the cold and erosion of the O rings in previous launchings.

Last Wednesday, at a news conference, Mr. Mulloy was asked about the decision to launch despite the cold. ''What we went on,'' he said, ''was the basis of the data that we had that indicated that there was no direct correlation between the erosion and blowby on the primary ring and temperature, based on the previous flight data.''

''Blowby'' refers to a rush of gases past a seal.

But Mr. McDonald said today, ''That is not quite correct.''

'Not a Strong Correlation'

''There is not a strong correlation,'' he said. ''But there is evidence that things got worse with low temperature.'' Before the Challenger launching, he said, the O ring was particularly cold because it had sat, uninsulated, in below-freezing temperatures the night before the launching.

Mr. McDonald, who is 48 years old and holds a bachelor's degree in chemical engineering from Montana State University and a master's degree in engineering administration from the University of Utah, said tonight that he was uncertain why his recommendation against the launching had been overruled.

''They felt, I am sure, that the basis for the concern was not fully conclusive,'' Mr. McDonald said.

By Stuart Diamond

The space agency and its contractors have wasted billions of dollars on the shuttle and other space programs despite warning after warning by Government inspectors that such heavy losses were occurring through bad management, Federal audits show.

The problems that come to light now in the National Aeronautics and Space Administration reflect a far different picture from the one widely held in the years before the Challenger catastrophe: that NASA was essentially a smoothly run, trouble-free agency.

And the audits, many made available under the Freedom of Information Act, hold special significance not only in showing how NASA was operated in the years before the Challenger explosion of Jan. 28. The documents also take on added importance because the Government is facing key decisions about the future of the American space program, including who should direct it, and, some experts contend, about the future of NASA itself. Criticism of NASA Chief

Dr. James C. Fletcher, who headed the agency in the 1970's and whom President Reagan has nominated to lead it out of the post-accident era, is said by the Federal auditors to have misled Congress and the public in the 1970's about essential costs of the shuttle program. Dr. Fletcher's confirmation hearings open in Congress today.

In an interview, one of a series with past and present NASA officials, Dr. Fletcher conceded he was ''over-optimistic,'' but he denied he attempted in any way to mislead Congress.

The pattern of management problems, as well as broken promises on costs, schedules and performance, emerges from a review by The New York Times of more than 500 audits, other Government documents and economic reports by outside experts and from interviews with American space experts.

In the last 15 years, bad administration and spending abuses have been found in virtually every aspect of the NASA operations, from running the shuttle to developing planetary probes, from satellites to construction of buildings, from space experiments to employee overtime, from headquarters to field centers, according to the documents.

Experts inside and outside Government say such faulty administration procedures have severely hurt the American space program and are impossible to separate from the safety problems that culminated in the Challenger explosion and the death of its seven astronauts. At Least $3.5 Billion Lost

The waste amounts to at least $3.5 billion, according to amounts cited in the audits. But in many cases the audits did not give dollar figures for instances of significant waste, and Francis P. LaRocca, the attorney for the NASA Inspector General's office, said, ''We always assume we have only seen the tip of the iceberg - there is probably much more out there.''

NASA officials say that, in general, the audits have been fair and accurate. They concede that their agency was slow to correct financial and management problems, taking several years in some cases to do so. But they insist that their flaws have been strictly administrative, not technical, and they deny that the flaws led to problems with shuttle safety.

Among other key findings, most of them new to the general public, that these documents and interviews disclose are these:

* The NASA Inspector General's office reported only last month in an internal document that the Marshall Space Flight Center at Huntsville, Ala., did not properly monitor what happened to defective equipment. Marshall was in charge of the booster rocket whose seal problems are considered the probable cause of the shuttle explosion. And last November, just two months before the disaster, Congressional auditors repeated previous findings that Marshall was weak in keeping track of and handling shuttle equipment.

* Millions of dollars in NASA equipment has simply disappeared, either stolen or lost.

* Hundreds of millions of dollars has been lost through failure to enforce contracts in a wide variety of projects. Contract failures included paying NASA or contractor employees for vastly inflated expenses.

* NASA spent as much as $1 billion more than it needed to for the new Tracking and Data Relay Satellite System.

* Hundreds of relatively small contracts resulted in the waste, overall, of hundreds of millions of dollars. In one, for example, the Government agreed to pay $3.6 million in rent for a building that it had just sold for $300,000.

The audits, by NASA's Inspector General's office, Congress's General Accounting Office and the Pentagon's Defense Contract Audit Agency, allege thousands of NASA or contractor violations of Federal law or regulation. And they portray the staff of the agency, which had a serene, always-in-control reputation, as constantly arguing behind closed doors with Government inspectors, and often ignoring criticism. Patterns Of Management Typical of the problems the audits disclose is the case of the shuttle's external fuel tank, the huge vessel for hydrogen and oxygen that forms the backbone of the vehicle and, in the Challenger disaster, caused most of the destruction when it broke apart. NASA officials have said they believe the tank was pierced by the booster rocket whose seal is believed to have failed.

The NASA Inspector General's office found in 1982 that the costs of developing and making the tank had more than tripled, to $502 million, from its original figure. Auditors attributed the increase to the lack of competitive bidding, unrealistic schedules, design changes in the middle of construction, failure to set prices before work began and allowing parts to be built before their design and testing were finished.

The auditors documented 2,714 changes on the tank by 1980. NASA officials changed some key components as many as five times as parts were bought and work was done, stopped, ripped out and redone.

They said the ''current external tank management is not cost effective and, in our opinion, is a microcosm of the overall shuttle program management difficulties.'' The tank project was managed by the Marshall Space Flight Center, which also managed the production of the booster rocket faulted in the shuttle explosion. Auditors said many of the problems resulted more from Marshall's management than from budget constraints. Disagreement With Findings Marshall officials told auditors that they disagreed with many of the 1982 findings and declined to follow major recommendations, including one that Marshall do more equipment testing and better document its reasons for cost changes, as required by law.

The Federal inspectors replied that Marshall officials ''do not have the arbitrary authority to act in contravention to the NASA procurement regulation.''

One of NASA's biggest mistakes was cited by its Inspector General's office in reports from 1981 to 1984 dealing with the Tracking and Data Relay Satellite System, an orbital communications system to replace ground stations that track space vehicles. The second unit of the system, a $100 million satellite, was aboard the Challenger when it exploded.

The Inspector General concluded in 1981 that the agency had wasted up to $1 billion, or nearly half of the $2.1 billion cost of the system, through a ''questionable'' leasing arrangement. Because NASA did not have the money to buy the satellites, it guaranteed a Government loan from the Federal Finance Bank to a contractor, Spacecom, a Western Union unit, to buy the system and lease it back.

But the auditors said NASA then failed to adequately monitor the contractor, letting it withdraw millions of dollars in advances for work not yet done, for expenses expressly disallowed and for payments on other contracts.

There were many delays and ''excessive'' cost increases, but auditors said NASA managers at the Goddard Space Flight Center in Greenbelt, Md., had agreed to pay for all delays, regardless of who caused them. And NASA agreed, ''in direct contravention of law,'' to repay the Government bank all money that Spacecom withdrew, even if the contract were to be terminated for any failure of Spacecom to furnish required services, auditors said. They called these agreements a kind of predetermined ''confession of judgment against NASA.''

The result, in part, has been a program that is still not operational, is more than six years late and is costing taxpayers $200 million a year instead of the $50 million anticipated, the auditors said.

As such, it has failed to fulfill the promise of being a cost-effective replacement for the $70 million-a-year ground-tracking stations. Auditors have urged NASA management to consider liquidating the contract and ''take timely and effective management actions'' to put the system ''on a prudent business basis.'' The agency did not liquidate the contract; the project's total cost has since risen to $2.8 billion.

James Elliot, a Goddard spokesman, declined to comment on the problems.

More broadly, in May 1984 the General Accounting Office said NASA's systems to track and correct financial, equipment, property and management problems were ''generally ineffective.'' It said 10 of the 37 management problems that existed in 1981 persisted until May 1984.

Only last November, the G.A.O. auditors repeated those conclusions and said the failures to detect and correct problems were still widespread. Auditors wrote of ''attitude'' and ''general leadership'' problems that prevented correction of persistent abuses. The report singled out the Marshall center, which just two months later gave the order to launch the Challenger without informing higher management of problems with the booster rocket's joints that had existed for at least six months.

The auditors made no specific mention of the booster rocket but repeated a 1984 criticism that there were ''problems with the entire logistics spares program'' for space shuttle engines at Marshall, including control over equipment both in purchasing and after it was bought.

The report of last November went on to say that NASA was telling the President and Congress that it had no significant internal weakness at the same time that its own auditors were telling agency management of significant system weaknesses.

On March 27, two months after the explosion, officials at Marshall were cited by the agency's auditors for violating various NASA rules in the handling of hardware that did not meet specifications.

The audit said Marshall personnel ''cannot determine the final disposition of nonconforming items.'' That is, if a part was broken or had some other manufacturing defect, Marshall people could not tell whether it was fixed or not before use.

It has not been disclosed whether or not the booster rocket seal faulted in the Challenger explosion was defective or installed improperly, but it is now generally acknowledged that the seal design itself was faulty.

Dr. William R. Lucas has been the Marshall director since 1974 and was its deputy director from 1971 to 1974. He joined the facility in 1952 when it was the Redstone rocket research unit. He testified before the Presidential commission on the Challenger explosion that he was aware the night before the launching of contractor engineers' concerns about the booster rocket seals but did not inform officials at Kennedy. He said he had been aware for some time of the problems but ''never considered the seals as a safety-of-flight issue.''

A spokesman for Marshall, told of the various audit findings on Marshall management during Dr. Lucas's tenure, said Dr. Lucas was not available for comment.

NASA officials contend that most of the problems have been corrected. But the audits show that many persisted for years, causing serious cost increases, schedule delays and ultimately damage to agency programs. And the same types of problems continue to crop up while agency-wide weaknesses persist, the audits show.

NASA's general response is made by C. Robert Nysmith, associate administrator for management. He said in an interview that the Government audits had been ''relatively fair and relatively accurate'' over time. Keeping track of equipment and keeping information flowing properly is ''not a high priority'' with NASA, he said, but one that needs ''more attention.''

''Getting the agency to identify weaknesses is not a natural culture,'' Mr. Nysmith added, but he insisted that ''the system is improving every year'' and he denied that the administrative failures had led to systemic failures in shuttle safety.

But a debate has been raging since the Challenger disaster on whether money and other pressures compromised safety, and the faulty practices disclosed in the audits seem sure to refuel the debate. Implications Emerge From Audits One of the immediate consequences of the newly revealed audits may well be the additional spotlight they place on Dr. Fletcher as Mr. Reagan's nominee to head NASA.

Dr. Fletcher was the agency's Administrator from 1971 to 1977, when many of its major cost and management procedures and abuses originated, according to the audits. And it was essentially Dr. Fletcher who sold Congress on the shuttle by saying it would provide inexpensive and routine access to space through frequent commercial missions.

Government records now make clear that he and his aides were wrong by wide margins in virtually all their economic projections, projections on which Congressional approval of the shuttle was justified and voted.

In 1972, Dr. Fletcher predicted, in news releases and Congressional testimony, a cost of $10.45 million for each shuttle launching. The General Accounting Office said then that the figure was ''misleading'' and ''overly optimistic,'' since it accounted only for operating costs, not construction, and did not reflect considerable technical uncertainties.

Today the costs are $279 million for each launching, including construction and operations, and $151 million each time just for operations, according to Congressional Budget Office figures. Even after allowing for inflation, today's operations alone for each launching cost five times more than the original projections. The NASA projections said the shuttle would be uneconomical if launching costs rose more than 75 percent.

Dr. Fletcher predicted, too, that the cost of lifting shuttle cargo into orbit would be $100 a pound. The cost now is $5,264 a pound calculated on total costs and $2,849 for operations alone. Allowing for inflation the corresponding rise is ninefold to nineteenfold, the Congressional Budget Office figures showed.

In letters to Congressional auditors, Dr. Fletcher and his deputies consistently refused to correct many of the problems cited by the inspectors, including management errors, spending abuses, delays in hardware testing, lax monitoring of contractors and optimistic estimates of costs and technical problems.

On April 6, 1977, Dr. Fletcher wrote to the General Accounting Office: ''Cost effectiveness was a central feature in the initial approval for the shuttle, and remains so. On a payload cost basis, the shuttle is the most economical space transportation system.'' The previous year, he cited the ''past excellent NASA record for on-time within cost program accomplishment on major manned missions.''

Dr. Fletcher, now a professor of engineering at the University of Pittsburgh, said in a recent interview: ''In retrospect, maybe we were optimistic. The cost per flight rose way beyond what we anticipated. We're all disappointed it didn't come out the way we calculated.''

But he said that ''the running battle with G.A.O. was real,'' adding that in the mid-1970's there were 10 to 15 auditors scrutinizing the shuttle program full time at the NASA centers of Marshall in Alabama, Kennedy in Florida and Johnson in Texas. He said that what needed correction was a matter of judgment, his against the auditors. He also contended that the shuttle has still provided valuable scientific and military capabilities to the United States.

Dr. Fletcher said it was not possible to comment more specifically on events so many years ago. But he added that NASA was then under severe budget constraints and asserted that most of the problems with the shuttle occurred after he left the agency.

He agreed, however, that the agency now needed critical review, saying, ''The structure of NASA should be examined.''

NASA now has no Administrator. James M. Beggs, who was appointed in 1982 and recently resigned, is under indictment on fraud charges relating to his previous employment at the General Dynamics Corporation. Mr. Beggs said in an interview that he could not comment on the auditors' criticisms because ''my lawyers have put strict restrictions on me.''

Even before the Challenger disaster and the re-examination it has brought, it was clear that key decisions for the 21st century must be made now: Should the United States begin building its next generation of spaceships now or simply remodel the present craft? Will it proceed with the construction of a space station? Will it begin interplanetary exploration with a manned trip to Mars in the 1990's, as many in the space community are urging?

The audits show a wide gap between past promise and future reality in space-related construction as well as shuttle operation.

NASA has proposed building a shuttle to replace the Challenger. The estimated cost of $2.2 billion is more than three times the 1972 projection, and 75 percent more than the 1982 figure, even taking inflation into account. In nearly all major NASA projects since 1970, costs have significantly exceeded projections, so there is no guarantee the $2.2 billion estimate would be close to the final cost, some shuttle experts say.

Some space economic experts, noting that expendable rockets have been taking an increasing role in the plans of both NASA and private companies, say it may not make sense to build another shuttle. They also point out that the shuttle is now 10-year-old technology and that new reusable space vehicles are on the drawing boards.

NASA originally predicted that it could complete 48 flights a year with three shuttles, the number it now has left. The agency now projects a maximum of only 24 flights a year, with four orbiters. Some senior astronauts are now saying publicly what they warned privately, that anything more than nine is dangerous because of the time needed between flights for maintenance and preparation.

In fact, according to the audits, NASA takes almost eight times longer to prepare the manned vehicles for launching than it promised Congress when the shuttle program was proposed.

The management problems also raise the immediate question in some minds as to whether NASA should continue to exist in its present form. Should a separate organization manage the shuttle? Should NASA's decentralized space empire, with a Washington headquarters and 11 centers around the country, be consolidated? Almost all outside observers urge that the shuttle program not resume before systematic management improvement occurs.

''The management has simply been poor,'' one of these observers, Alex Roland, says. Mr. Roland, who was a historian and contract monitor for NASA from 1973 to 1981, is now a specialist in the history of technology at Duke University. ''A handful of new people in key positions could turn the agency around,'' he says. ''But the organization is going to have to start holding people more accountable for their actions.'' 'The Most Difficult Agency' Most of the management problems that are essentially new to the general public and vast in scope show up in audits done by NASA's Office of Inspector General and a related agency that often works with it, the Defense Contract Audit Agency.

The NASA Inspector General's office, which has a staff of about 100, is chartered by Congress to examine the agency's management and its spending practices. It is a separate entity not beholden to the Administrator, an arrangement similar to the one between the Pentagon and its Contract Audit Agency. Their recommendations to agency heads are not binding. Nor are they normally released to the public.

Figures are not available on NASA's cost management compared with that of other agencies. But some audits indicate that practices at NASA, so long regarded by outsiders as well-run, were on occasion worse than at the Pentagon, considered by many the worst for cost overruns.

The General Accounting Office, an investigative arm of Congress, has generally found the same kinds of management problems at the space agency as the NASA and Pentagon auditors have. In 1977, the G.A.O said NASA, unlike the Pentagon, had not determined for Congress what cost increases resulted from inflation, what came from changes in the size of a project or what came from technical problems. And while the Air Force was able to reduce contractors' prices through negotiation an average of 15 percent, the Marshall center's average reduction was 3.4 percent, NASA auditors said in 1984.

The G.A.O. inspectors say they have prepared more than 10,000 audits of Government agencies since 1970, including 335 audits of NASA.

J. Klein Spencer, a G.A.O. official who has audited the space agency for 10 years, said NASA was slow to provide material needed for audits and sometimes never provided all the information required by law.

He said ''NASA has been the most difficult agency'' in terms of obtaining information. ''We have problems with other agencies,'' he said, but adding that NASA consistently posed problems ''on every job.''

''It's just their management philosophy over there,'' he said. ''They want to provide a minimum of information, and it takes a great deal of effort to get the records we need.''

Mr. Nysmith of NASA says the agency's current policy is ''to provide them with any information they need - we have a turnaround time in 30 days.''

But Mr. Spencer said, ''There has been no change.'' Monitoring Contractors Among the themes that run through the audits reviewed by The Times is NASA's inability to manage or monitor the work of thousands of contractors paid billions of dollars for everything from manufacturing and maintaining the shuttles to painting buildings. More than 85 percent of NASA's $7.5 billion annual budget is spent through contractors.

Often, NASA has allowed contractors to start and sometimes finish projects before costs were determined. This has left the contractor virtually free to determine how much a job would cost, a 1981 audit said. Yet NASA's own procurement regulations specify that pricing must be done before work starts.

On the shuttle, the Rockwell International Corporation, the main contractor, spent $20 million on a propulsion system in the two years between receiving a work order and agreeing on a price settlement. NASA had expected a $3.2 million price but wound up paying Rockwell $19.2 million.

At times, more than $750 million in work was being done without price agreements, the auditors found, adding, ''The budget and planning cost estimates prepared by Rockwell International Corporation are often inaccurate and unreliable.'' Some of Rockwell's payment claims were submitted five years after work was started.

George W. Jeffs, who has been president of Rockwell's space division since 1976, says it is difficult to submit the required paperwork because of the tight schedule.

''When you have a change in the middle of a development program,'' he said, ''often the teams don't have time to prepare the proposal that goes to the Government. Oftentimes they get the job done first.'' The dollars go up as time goes on, he said, because ''changes are being made to the changes.''

It seems clear that one of the problems may have been the close associations that have developed between NASA personnel and its contractors' employees over the years. Mr. Roland, the Duke historian, says the Rockwell shuttle history illustrates ''the cozy relationship between agency officials and contractors.''

''They live in the same area, their kids go to the same schools; they even sound the same,'' he said. ''The only way you can tell the difference is by their badges.''

Another instance of the mismanagement of NASA contractors, one of the few such cases that was well-publicized, occurred in 1979. It involved what the Defense Contract Audit Agency later called ''unnecessary and unreasonable travel costs'' of $1.8 million. The billings included drives in rental or private cars by hundreds of workers from Rockwell's California headquarters to Florida to make major thermal tile repairs on a shuttle orbiter, as well as salaries during the cross-country trip and rental of beach quarters in Florida. The cross-country drives cost far more than airfare. Rockwell had also billed the Government for a trip abroad by four people to support publicity for the James Bond film ''Moonraker,'' which featured the shuttle.

Ultimately, NASA managers disallowed $252,000 of the charges.

Mr. Jeffs asserted that ''this was an emergency situation'' and the company's actions were ''extremely reasonable'' under the circumstances. The Perpetual Costly Circle What has happened has in a sense been circular and self-perpetuating. NASA's inability to monitor contractors properly was attributed in part to cuts in staff, made necessary because of shortages in funds caused in many cases by excessive contract spending.

In the Apollo program to put man on the moon, NASA had 28 contract monitors at the Johnson Space Center; in 1979 and 1980, it had two. There was one person to monitor the 44,700 purchase orders Rockwell issued in 1981, the auditors found.

At the Marshall center in Alabama, officials authorized Rockwell to purchase $1.6 million in spare parts that were not needed, a 1984 NASA audit said. At $730 million, procurement for more than 120,000 spare parts was the largest single Marshall contract, but the auditors noted that the cost proposals are reviewed by one full-time analyst with some part-time help. The auditors said 78 percent of the reviews were ''cursory,'' violating Marshall's own procurement requirements. Marshall approved fully 99 percent of the requests for spares submitted by Rockwell, the auditors said.

Whether because of inadequate staff monitoring or for other reasons, audits show that for years, instead of buying parts directly, NASA paid Rockwell or its subsidiaries for parts that had been bought from vendors, contrary to NASA policy. Markups sometimes exceeded 1,000 percent. At one point, Rockwell sold an electronic cooling fan to NASA for $159,000. It had cost the Rockwell subcontractor $5,215.

As late as September 1984, the NASA Inspector General's office said Marshall had ''not adequately considered or evaluated the monetary advantages which could be realized'' by buying spares directly from vendors. The auditors put the potential cost of this practice over five years at $10 million.

The shuttle contract for the extravehicular mobility unit, which enables astronauts to manuever around outside the shuttle, grew from $24.1 million to $71.8 million in part because of NASA's ''lack of timely review and consideration of significant changes,'' according to internal inspection reports. Instances Of Fraud Not all the extra cost to the shuttle and other space programs occurred simply because of sloppy procedures or staffing problems.

In many causes outright fraud occurred.

The space agency, as one audit put it, ''lacks essential internal controls to prevent fraud.''

In February 1985, Dr. Herman E. Thomason, deputy director of science and engineering at Marshall, pleaded guilty in Federal court to charges that he ''directed, recommended and approved'' the NASA purchase of millions of dollars in computer equipment from a company in which he had a substantial personal financial interest.

Dr. Thomason, who had been honored in 1979 for ''exemplary leadership,'' left the agency, paid a $50,000 fine and spent 60 days in jail for violating NASA regulations and ''knowingly making false and fraudulent statements'' about his outside financial interest, court papers said. Dr. Thomason's department evaluates space data, including that from the shuttle.

Many other instances of fraud also involved NASA employees. One space scientist at agency headquarters in Washington got a $3,000 finder's fee for helping a company to gain an agency contract. The director of NASA's Lewis Research Center in Cleveland was convicted of fraud for submitting more than $6,800 in phony travel vouchers. In 1981, 21 NASA employees at Langley Research Center were found by auditors to have improperly overcharged on their time and attendance, and the auditors said ''management had failed to correct the abuses.''

Other instances of fraud involved private contractors. Two officials of the Mayfair Construction Company of Cocoa Beach, Fla., were sentenced to three years in prison each for mischarges and false statements pertaining to construction contracts worth more than $20 million at Kennedy Space Center. But millions of dollars in wasted funds and lost time went unrecovered, auditors said.

Other cases of fraud, generally detected not by NASA's administrators but by the agency's Inspector General and Defense Department auditors, included false claims for overtime submitted by contractors and their employees as well as contract overcharges.

One contractor, for example, charged $12,000 each for parts he had agreed to supply for $5,000 and which were available on the open market for $2,000. In 1980 the Boeing Computer Services Company in Seattle paid $2.4 million to settle a $10 million overcharge it had made against the Government, the auditors said.

The audits cite numerous cases in which cost overruns on fixed-price contracts with such agencies as the Army Corps of Engineers, the Air Force and the Agriculture Department were charged to contracts with NASA. In 1985, for example, General Electric agreed to pay a $1.1 million settlement for improperly charging NASA and other agencies an estimated $3.1 million in cost overruns that the contractor could not collect on a fixed-price Air Force contract, auditors said.

Rockwell was discovered to have overcharged the agency for ''massive numbers of labor hours'' on a fixed-price contract with the Air Force for a satellite. The company agreed to pay $1.5 million in 1983. But two years later, on Oct. 31, 1985, auditors reported Rockwell had tried to get back $541,000 from NASA in legal fees from the case through a general claim for legal expenses. The claim was disallowed. Repeated Code Violations Cited Time and again the auditors cited violations of Federal cost and management regulations.

A significant category of these allegations, perhaps one of the most frustrating to the cost-conscious investigators, has been NASA's practice of buying large quantities of goods already owned by the Government.

In January 1984 Marshall said there were no ''off the shelf,'' or standard, items for the shuttle's main engines. NASA auditors arranged an independent search, which found 1,587 matches from Government sources.

From 1975 to 1981 NASA, in violation of its own contract and Federal codes, allowed Rockwell to buy parts on the open market when millions of dollars of identical parts were available much cheaper from Federal sources, the auditors said. In a case reviewed in newspaper accounts shortly after the Challenger explosion, NASA was found to have paid Rockwell's Rocketdyne division $120 each for bolt assemblies available from Government sources for $3.28 and $315 each for 3-cent metal loops available in Government supplies. A $1 washer cost $80, a $2 clamp cost $86 and a $78 bolt cost $1,621, the audit found.

In another contract violation, a NASA official hired his son and other novices for a painting job at the Goddard center, according to a 1979 audit. The $106,000 bill was said to be more than twice the standard amount because the NASA official estimated 3.8 million square feet to paint when the actual figure was 472,500.

At Kennedy Space Center, executives of EG&G Florida, the base operations contractor, were allowed to take $51,000 in Government money, buy oak office furniture and retain ownership, in violation of Government codes, according to a March 1985 NASA audit. Lost Time In Operations NASA's poor control over contractors caused the agency to pay for large amounts of idle worker time, records show. Many of the losses are said to have occurred for years before the audits and in some cases continued afterward. Often they were documented by the Defense Contract Audit Agency.

Repeat audits of Rockwell in 1979 and 1980 found the company had a 34 percent rate of idle time, costing NASA $3.7 million annually. The Johnson Space Center had been asked for more than a year to ''improve supervisory monitoring of employee work.'' In 1981 auditors found that $650,000 was being wasted because Rockwell employees at Kennedy Space Center were taking extended breaks, leaving work early and putting in excess overtime.

In 1983 auditors found NASA wasting $28 million a year because Rockwell engineers were idle more than 30 percent of the time. Auditors found NASA wasting $6.7 million a year because of a 20 percent rate of ''excessive nonproductivity'' at General Dynamics in San Diego.

NASA auditors found that $1.7 million a year was being wasted at the Langley Research Center in Hampton, Va., because the center's management ''had not sufficiently emphasized cost control'' over a maintenance contractor. Workers took nearly two hours to do an hour's work, engaged in ''cruising'' around the center in vehicles ''without an apparent reason'' and were not properly equipped with tools and materials, the auditors said.

The audits also show that employees of both NASA and contractors were using narcotics, marijuana, cocaine and other illegal drugs, and marijuana was actually found growing at one NASA facility.

In April 1984, auditors of the Johnson Space Center, whose Mission Control directs space flights after launchings from Cape Canaveral, said management must improve its controls over the use of narcotics at the center. Theft and Other Property Troubles The audits show that NASA had problems controlling its basic resource, more than $12 billion in property and equipment.

An audit by the General Accounting Office concluded that officials at the Johnson center consistently accepted the loss, presumably through theft, of tens of thousands of dollars of equipment each year in numerous small contracts. The officials told auditors that there was ''a general attitude that small dollar-value contracts are not worthy of attention.''

The audits document many specific instances of theft: Copper cable, aluminum, tools, $26,000 of gold that had been salvaged from rocket engines, rare metal used in the shuttle's solid-fuel booster rockets were stolen. At least $100,000 of silver engine-ring seals were taken, as was $450,000 of gold used in making shuttle thermal tiles.

As early as 1970, the Congressional auditors noted that the Kennedy Space Center had not accounted for more than $320 million of equipment for as long as four years. Marshall officials had been asked to account for their equipment in 1970, but at least $250 million was not tabulated by 1974. At Goddard, Johnson and Marshall, 3,779 items valued at $3.1 million were missing by 1974. Auditors said it was often unclear how much was stolen and how much simply lost because of poor accounting.

At the Ames Research Center in Mountain View, Calif., nearly $1 million in property was lost from 1977 to 1982 because required inventories were not done, losses were not investigated and center officials declined to penalize responsible officials, NASA auditors said in September 1982.

Ames promised to correct its control procedures but the next year, the auditors found that another $2.4 million, in computer equipment, had disappeared.

At the Langley Research Center, there was said to be ''no control'' over gold, silver, platinum and other precious metals used in laboratory experiments in 1981. Platinum was found stored in a brown grocery bag, and some $71,000 in silver ran down drains each year for three years because a reclamation filter had not been changed. Two years later auditors found some procedures still unchanged.

The Marshall center allowed Morton Thiokol Inc., the prime contractor on the booster rockets, to rent a building for $3.6 million over 13 years and charge the Government for the rent, although the Government itself had recently owned the building and sold it for $300,000, NASA auditors said.

Thiokol also rented three other buildings for rocket storage, assembly or manufacturing, although eight nearby buildings owned by the Government could have been used, auditors said.

In 1976, an audit by the General Accounting Office found that the Marshall center had fully 14 warehouses of equipment that had been ordered by NASA employees who had since left the agency or ''who had since died.'' Most of it was gathering dust.

And last year, two audits at Kennedy determined that more than $25 million was being wasted because two contractors were each warehousing and managing 25,800 identical pieces of equipment. The Lockheed Space Operations Company, which prepares the shuttle for launching, was allowed to spend $5.9 million for new warehouse space to store another 44,600 other items that were not needed, the audits showed.

John D. Williams, a spokesman for Lockheed, which assumed its duties on Oct. 29, 1983, said, ''We will continue to work on the problem areas cited in the audit.''

On the other hand, contractors at Kennedy were not stocking enough spares and had failed to order $55 million in requested equipment, a 1985 audit found. Auditors said lack of spares had delayed a number of shuttle launchings.

One result was that parts needed to replace bulky items on an orbiter waiting to take off were taken from other orbiters. ''The cannibalizing just horrified me,'' said Albert J. Kelley, a vice president of Arthur D. Little Inc., a former NASA project manager who later was a management consultant in technical fields. ''Every time you open a clamp, break a seal, damage a fitting to take out a part, you are looking for trouble.''

By Stuart Diamond

The space agency cut or delayed half a billion dollars in spending on safety testing, design and development from the time the shuttle program began to when the Challenger exploded in January, Federal audits show.

The range of work included testing the shuttle and its main engines for vibration, developing the booster rocket that probably caused the explosion and conducting communications and thermal tests.

For years Government auditors and aerospace experts implored the National Aeronautics and Space Administration not to cut testing and other development work. They said eliminating the spending or significantly delaying it - some of the money has still not been spent - would mean equipment would be assembled before components were fully tested. It would then be very difficult and costly to make changes. Ultimately, this would compromise reliability and make the machine less cost effective and less safe, they said.

But the agency, faced with ever-tightening budget restrictions, refused, saying the eliminated procedures were ''not cost effective,'' said the General Accounting Office, a Congressional investigative agency.

The cutbacks came in the same general period in which Government inspectors repeatedly warned NASA that it was wasting large sums of money through faulty management. This money, more than $3.5 billion in all, in turn helped create the budget problems. Review of 500 Audits

The reductions in safety testing, design and development and the patterns of excessive spending and mismanagement at NASA emerged in a review by The New York Times of more than 500 audits of the space agency by its own Office of Inspector General, the General Accounting Office and the Pentagon's Defense Contract Audit Agency. Interviews with space experts inside and outside the Government generally confirm the audits, most of which have not been made public. Many were obtained through the Freedom of Information Act.

The agency issued a statement yesterday saying some of the problems cited in the audits ''are related to activities several years ago and have been corrected or are being corrected.''

''Furthermore,'' the statement said, ''during the existence of NASA it has used its own internal audits to uncover and correct deficiencies and to strengthen internal controls.''

NASA officials said in interviews that the audits had been fair and generally accurate. They also contended that the mismanagement was administrative, not technical, and insisted that it did not result in safety compromises.

Among the chief findings of the audits were the following:

* On the Challenger, faulty welds concealed through falsified X-rays by a subcontractor to avoid the cost of repair. They went undetected and uncorrected until NASA auditors received tips from former employees of the subcontractor.

* After a reduction in its inspectors' ranks, the Johnson Space Center, in direct charge of the shuttle flights that were to begin a year later, failed to detect equipment flaws so critical that they could cause loss of life or destruction of the spaceship.

* NASA misled Congress about costs and schedules for the shuttle and other programs, withheld critical documents and violated Federal codes in thousands of instances.

* NASA violated Federal policy by spending billions of dollars on shuttle equipment that had not been thoroughly tested.

* By 1976 the shuttle had become ''a higher risk program'' in terms of economics and performance than conceived, the General Accounting Office said. The Transition From the Apollo Era Experts familiar with the history of NASA say problems at its headquarters in Washington and 16 other space centers began shortly after the agency reached its founding goal of landing men on the moon in 1969 and after the Administrator who led NASA there, James E. Webb, resigned after eight years in office.

''There was a distinct change after Mr. Webb left,'' said Elmer B. Staats, who from 1966 to 1981 was the United States Comptroller General, the top position at the General Accounting Office. ''Mr. Webb was a good manager. He put together a good team. But afterwards, there was less openness, less willingness to listen to the kind of challenges we were making. It was an institutional thing; it was the whole system.''

One of the major changes was the way the agency viewed the design, testing and development of new space systems.

On the Apollo moon program and others of that era, each component was designed, built as a prototype and then tested, space experts say. Thus when the final system was assembled for a test, scientists had a high degree of confidence that it would work.

But NASA decided that to save money it would cut back on the kind of testing done in earlier programs and instead build major components and systems before they were fully tested, according to auditors, space experts and NASA officials themselves, including C. Thomas Newman, the NASA comptroller, who has been with the agency since 1963.

''The shuttle set out with some different objectives,'' Mr. Newman said. ''To produce a system of moderate costs, the program was not as thoroughly endowed with test hardware.'' Cutting of Costs Begins in Earnest Most of NASA's money problems -overspending for equipment and personnel and shortages of funds for testing - coalesced on the shuttle program, the auditors conclude. The program started as a relatively modest project and eventually took up 60 percent of the agency's budget.

The shuttle was NASA's choice to succeed the Apollo program as the agency's major endeavor after NASA was rejected in its attempt to undertake a manned flight to Mars and build a space station. The Nixon Administration backed the shuttle to aid scientific and military goals and help shore up the aerospace industry, according to experts familiar with decisions then.

But space experts said the program was sold to Congress, essentially by Dr. James C. Fletcher, NASA's Administrator at the time, based on the idea that it would pay for itself through commercial business, deploying and repairing private satellites and selling other space services. Dr. Fletcher has been nominated by President Reagan to head the agency again, and his confirmation hearings got under way in Congress yesterday.

The shuttle's economics, however, depended on relatively low operating costs to offset the high development costs.

To meet budget restrictions, NASA proposed a shuttle that would be reusable only in part. That kept down the development cost but made operating costs much higher, as the huge external fuel tank and other components had to be replaced each time.

To further cut costs, NASA projected 60 flights a year. ''They were assuming a ridiculous number of flights,'' said John E. Pike, space policy expert at the Federation of American Scientists. Since the Challenger explosion, some astronauts have said that any more than nine flights a year can put dangerous pressure on those preparing each flight.

C. Robert Nysmith, the agency's Associate Administrator for Management, said NASA officials consistently believed that, despite the cuts, enough protections remained to keep the space programs safe even if some testing was delayed.

The risk of the reduced or delayed testing, however, was that if major problems developed, costs would rise greatly as they were fixed and, in reality, safety might well be compromised, said auditors and such space experts as Dr. Albert J. Kelley, a vice president of Arthur D. Little Inc., the management consulting concern, who was once a NASA project manager.

''It's not the optimum way,'' Dr. Kelley said, ''and you take a risk. When you have pressures, you look for ways to cut corners, and you manage differently. What usually goes is design planning and hardware testing.'' The Issue Of Safety Despite NASA's views to the contrary, the Government auditors argued that there was no clear distinction between technical and administrative management, particularly because it often involved the same people.

And if there is no real distinction, they argued, safety issues cannot be considered separately from administrative problems.

Dr. Kelley, who holds a Ph.D. in aeronautics and astronautics from the Massachusetts Institute of Technology, said: ''Outside the laboratory, there are very few decisions that are only technical. The whole technical, management, procurement, schedule and cost issues are interrelated.''

The cuts and delays in early shuttle spending included $147 million of testing for the main engines, the most advanced ever developed for use in space, audits by the General Accounting Office said. About $118 million for testing and developing the orbiter was cut, $10 million for developing the external tank for liquid fuel, $68 million for developing the solid-fuel booster rockets and $70 million for launching and landing development.

Vibration tests for the orbiter and for the shuttle assemblage were deleted. Tests of tires and landing gear were cut back. Testing and development were eliminated or postponed for years on a wide variety of radar, communications links and other equipment, including that needed to operate shuttles in orbit.

The audits give numbers for a total of about $500 million of cuts, but not all of the deletions have price tags attached.

Some of the development work was moved to the production phase of the program and some to the operations phase, but a substantial portion was deleted: at least $110 million by 1976.

The cuts worried experts both inside and outside NASA. They warned that because the performance of much equipment could not be verified early in the program, the agency might wind up with unreliable and unsafe equipment. It also would not find out about major problems until an advanced stage of production, meaning that any changes could be very costly.

Some of those concerns were raised in 1974 in a study by 35 aeronautical and space experts led by Willis Hawkins of the Lockheed Aircraft Corporation. The team said it was worried that testing was being ''highly compressed.'' Orbiter testing, for example, was cut from 16 months to three. Main engine testing was slashed, the report said. It said there was little backup equipment for use in tests, so failures would delay the schedule.

The experts said that the likelihood of unforeseen technical problems was high because of the ''difficult technical goals'' but that there was little margin for error. Those concerns were echoed by three studies at NASA's Goddard Space Flight Center. And they were repeated by Congressional auditors in 1975 and 1976.

In 1977 the auditors said deletion of tests of tires, brakes and landing gear ''may result in loss of vehicle and crew.'' These systems have consistently caused problems on shuttle landings.

But NASA consistently disagreed with the auditors and experts. The agency rejected a plea by the Hawkins team to stop compressing the test schedule ''as not being cost effective.'' Major Troubles Begin to Develop NASA officials said they did not need all the tests because of experience gained from the Apollo program, though many experts argued that the shuttle was a new venture with its own potential problems. The General Accounting Office noted that some of the deleted tests had turned up many problems that required changes in design and other areas. Yet they were deleted, the auditors said, because NASA thought them ''too costly.''

On Feb. 27, 1975, Dr. Fletcher wrote to the General Accounting Office that NASA had a ''sound management process'' for the shuttle, adding, ''I have no reason to anticipate cost overruns'' beyond the original commitments. Two years later Dr. Fletcher said, ''The shuttle has encountered normal technical problems of a complex program.''

But major problems were beginning to develop by the late 1970's with the shuttle's thermal protection system, about 30,000 lightweight tiles that cover 65 percent of the vehicle and protect it from the heat of reentry. Tiles kept falling off.

Moreover, the main engine, when built, kept blowing up in full tests. Pumps, valves and other equipment failed when complete. Hundreds of millions of dollars more were expended for corrections.

Even since the shuttle started flying in 1981, flights have been delayed and problems have developed in flight as some of the equipment not fully tested failed. At least 7 of the 25 flights have been delayed a total of nearly seven months because of problems with engines, onboard power units or the solid-fuel booster rockets.

In May 1982 NASA's inspector general said, ''Technical difficulties and problems have occurred across the entire spectrum of the Space Transportation System development activity.''

As recently as March 27, NASA's own Office of Inspector General reported that administrative practices at the Marshall Space Flight Center in Huntsville, Ala., were so lax before the Challenger explosion that the center could not properly determine whether defective equipment had been fixed. Marshall officials, who had responsibility for the booster rockets, have declined to comment on the findings. Defective Welds Found on Challenger The link between NASA's administrative and safety problems was apparent in various aspects of the shuttle program, the audits show.

In reports by NASA's Inspector General in 1979 and 1980, the Rockwell International Corporation, prime contractor for the shuttle orbiter, was found to have failed to detect promptly and report defects in items whose malfunction could be threatening to a spaceship or its crew. And the NASA managers at the Johnson Space Center who supervised Rockwell did not correct the problems for at least 14 months, the inspectors said.

Three years later, the auditors reported that a Rockwell subcontractor had routinely submitted falsified X-rays to conceal safety defects in welds on the Challenger ''to avoid the costs of rewelding.'' The auditors said the subcontractor, Jet Air Inc., of El Cajon, Calif., had given its welding work to unapproved machine shops and then falsely certified that it had done the work.

The problem was caught only after former workers of Jet Air informed NASA auditors.

Rockwell, at the auditors' request, reviewed more than 1,800 parts on the Challenger and found a number it could not certify as proper. But the company suggested that the Challenger ''fly as is'' because of the ''high design margin'' and ''noncritical'' function of the parts. It suggested repeated inspections and limits on use of the parts because of possible ''crack growth.'' One part carried the warning, ''May yield under load, but no catastrophic failure.''

The documents of the three inspection agencies contain no specific references to administrative procedures or safety cutbacks related to the rocket joints suspected as the cause of the Challenger's explosion, in which the seven crew members died.

But the head of NASA's shuttle program, Arnold Aldrich, told the Presidential commission investigating the Challenger explosion that a serious management flaw had preceded the tragedy. He was never told, he said, that engineers had complained about troubles with the rocket seals. The complaints, made for several months in 1985, had reached lower-level officials, NASA said.

The Federal auditors warned that reducing tests of safety and design would lead to miscalculations about equipment performance. Indeed, one miscalculation occurred on a system that was designed to blow up the shuttle if it went out of control within range of populated areas.

Air Force personnel, who monitor shuttle launchings, used the range-safety system to blow up the booster rockets that were flying wildly after the Challenger broke apart. Yet in 1974 NASA told the Air Force that such a system would not be needed after the first few flights. ''Adequate safety would be attained through system design reliability'' of the shuttle, NASA argued.

NASA did adopt a system by which the crew could terminate a malfunctioning flight, the Congressional auditors said, but its choice still did not meet Air Force's requirement for a remote control. The Air Force imposed this requirement to protect people on the ground in case disabled astronauts were unable to control the flight.

NASA is said to have continued developing its crew-operated system for more than a year over Air Force objections, while deleting $10 million for testing a remote-controlled system.

Late in 1975 NASA decided to install an Air Force-approved system. It cost millions of dollars more than the agency anticipated.

Among the other development work that suffered was designing equipment for ease of maintenance. In 1975 NASA canceled a contract with American Airlines for a study of maintenance considerations. The agency's inspector general said years later that the lack of such considerations was a major reason for the greatly increased time it was taking for preparations between launchings, 1,240 hours instead of 160. This meant increased pressures on personnel, and an added threat to safety, auditors and space experts alike have said since the Challenger exploded. Congress Was Misled, Audits Assert Audits of the space agency since 1971 allege that NASA consistently misled Congress on projects, whether on costs or feasibility, in violation of various Federal laws.

At one point NASA told Congress that the agency had $2.7 billion to work with when it actually had $4.3 billion, counting surpluses and other funds, the auditors said. The General Accounting Office said that ''NASA's data is inaccurate and incomplete'' and that Congress should not rely on the data. The auditors said NASA's actions violated its own guidelines for providing Congress details about projects. But the dispute continued all through the 1970's, as the General Accounting Office repeatedly said NASA was misleading Congress.

The auditors said the agency's reports to Congress contained some direct misstatements, usually underestimating costs, but auditors and space experts say the reports were also misleading in what they failed to report: that budget cutbacks could not be separated from possible safety problems.

As auditors' criticisms intensified over the last decade, there were some expressions of concern in hearings by the five Congressional subcommittees with jurisdiction over NASA. But Congress kept approving budgets for the shuttle project and usually accepted NASA's explanations or pledges to change.

The committees were too soft on NASA, according to the auditors, various Government experts and even some legislators. One of the auditors, Donald E. Day, senior associate director of the G.A.O.'s Division for National Security and International Affairs, observed, ''The committees have been very supportive of the agency. They have gotten too close to the agency and have been less objective than they should have been.''

Mr. Day singled out Representatives Don Fuqua and Bill Nelson of the House Science and Technology Committee, both Florida Democrats. And he pointed to Senator Jake Garn, whose Utah constituency includes Morton Thiokol Inc., maker of the suspect rocket boosters.

Those and others on their committees, Mr. Day said, ''are NASA's biggest fans when you get right down to it.'' He said that was why auditors' warnings on cost and safety mismanagement had been so consistently ignored. ''Even though we bring things to the Congress, there is this shield,'' he said. ''They say, 'Thank you, G.A.O., for your work,' '' and then ignore it.

Each lawmaker denied vigorously any support of NASA by virtue of their constituencies. They acknowledged certain shortcomings in the process and affirmed their support of NASA.

Auditors said other reasons for ignoring warnings included NASA's remarkable success with the moon landing as well as a public belief in the space program in general.

''The shuttle had a kind of special appeal in this country,'' said Mr. Staats, the former Comptroller General. ''We had many meetings with NASA, but we couldn't convince them to make the changes. And there wasn't enough support for our position in the top levels of Congress or in the Administration.''

Bill Green, Republican of Manhattan and ranking minority member of the House Appropriations Subcommittee on H.U.D. and Independent Agencies, said ''there was an assumption of managerial efficiency'' with NASA.

''Certainly the success of the programs helped the Congress give NASA'' what it wanted, Mr. Green said. ''But the shuttle program has very much eroded that. The oversight will be a lot tighter.''

Walter F. Mondale, who was a vocal opponent of the shuttle when he was a Senator, agreed. ''The information, the statistics, the facts simply didn't matter to NASA's sponsors in Congress,'' the former Democratic Presidential candidate said in an interview. ''There was this whole empire of people left over from the Apollo program with nothing to do. And to sustain their efforts, they needed show business. And manned flight was the drama.''

Mr. Fuqua, a Democrat who has headed the House Committee on Science and Technology since 1979, said he and others did raise questions about cost and testing, but generally let the agency have its way. ''We raised questions. But we said, 'If it works, you look good. If it doesn't, we all are going to have egg on our face.'

''We are strong proponents of the space program,'' he said. ''I think we wanted NASA to do well. We had to trust their judgment. They have some very competent people.''

Now, he said, the agency must make its management more centralized. ''I don't think we have to legislate it,'' he said, ''but there will be a lot closer scrutiny.''

Mr. Garn, who heads the Senate's counterpart to Mr. Green's subcommittee, has been on Senate science committees since 1974 and flew aboard the shuttle Discovery in April 1985.

''Of course NASA has troubles in management,'' he said. ''Of course there are things that could have been done better over the years. There were inefficiencies in the way money was spent. NASA has been wrong in its estimates.''

But he argued that many Government agencies wasted money and that NASA had done a ''remarkable'' job, given the complexities of its projects.

''Someone has to stand up for this agency when everyone else looks for the negative,'' Mr. Garn said. But he, too, added that there would have to be management changes. Auditors and Experts Offer Conclusions The space shuttle, it is clear from the audits, is far less re-usable than planned. The main engines, which were supposed to last 55 flights without major overhaul, for one example, barely last five flights before major components have to be replaced. Vibration is causing premature wear of a main pump, including blades and bearings. The engines are operating at 5 percent less power than their design goal.

And some of the rocket boosters have lasted less than their predicted 20 uses due to various problems such as damage from impact with the ocean.

The result is far greater cost for replacements, lower safety margins and a nearly eightfold increase in turnaround time. Major multimillion-dollar development work for the program is still continuing even as the orbiters are considered ''operational.''

The shuttle will cost at least $2.4 billion a year to operate and another $1.9 billion a year to amortize the investment, the Congressional Budget Office says. Those costs mean the shuttle is significantly more costly than the expendable rockets it was to replace, according to shuttle economic experts, and is not even close to paying for itself.

Thus, nearly everyone interviewed, both inside and outside NASA, agreed that the original goal of commercial competitiveness for the shuttle should be abandoned. Mr. Newman, the NASA comptroller, said, ''I doubt the shuttle itself is a commercial venture.''

But most agreed, too, that the shuttle was an extremely valuable scientific and military asset. And they contended that if the commercial pressures for the frequent launchings were eliminated and the shuttle was used more selectively when needed, it would become a safer and more valuable national enterprise.

By William J. Broad

The key solution proposed by the space agency to the problem that caused the Challenger explosion was devised years before the disaster and was being systematically applied to the entire space shuttle fleet, interviews, documents and testimony show.

The record shows longstanding worry about seals on the shuttle and application of an innovative remedy to generations of booster rockets. It also shows why the solution was unavailable for the Challenger flight of Jan. 28, in which a leaky booster seal touched off an explosion that destroyed the craft and killed its seven astronauts.

To critics, this new history suggests duplicity on the part of the National Aeronautics and Space Administration: While publicly declaring the shuttle safe to fly, these critics say, NASA officials knew the flaw was grave and had quietly embarked on a major program to fix it.

NASA officials vigorously deny this charge, insisting they were unaware of the seriousness of the flaw and were simply trying to increase the booster's margin of safety.

No matter what prompted past action, the fact that NASA had almost five years of experience with the remedy explains how the agency was able to propose a $300 million seal redesign, unveiled Aug. 12, as fast as it did. Both critics and NASA officials agree that the space agency is now following the basic course it set for itself long before the Challenger disaster.

The key element of the redesign was invented in 1981 to lock seals tightly in place by latching down an element of the joint. This capture feature was applied to the design of booster rockets in 1982 and again in 1985.

In July 1985, amid growing fears of seal failure, NASA officials ordered 72 new steel cases so they could install capture features in boosters already in use. These new cases were in production at the time of the Challenger disaster and today they are the basis for the creation of improved boosters.

This history, gleaned from dozens of interviews, hundreds of newly released documents and thousands of pages of testimony, much of it originally kept secret, is sharply at odds with information previously made public. It had been thought that NASA officials were concerned about the booster seals before the Challenger disaster and were searching for a way to fix them. In fact ''the solution,'' as NASA engineers called the capture feature, was already being systematically if slowly applied.

The Presidential commission that investigated the Challenger disaster steered away from this history and the issues it raises. Among them are: Who knew replacement casings had been ordered? Did officials who knew about these casings press to continue flights? Why was the new design being applied slowly?

The commission's chairman, former Secretary of State William P. Rogers, said the panel should deal only with events leading directly to the disaster. But others are investigating deeply in other directions.

Representative James H. Scheuer, for instance, sits on a Congressional committee that recently reviewed the Rogers commission's findings and held hearings on the disaster. ''It's remarkable how much NASA knew and how little they did about it,'' said Mr. Scheuer, a Queens Democrat. ''There was no sense of urgency. There was no order or direction that went out. There was no mobilization of forces.''

Richard C. Cook, a former NASA budget analyst who warned in 1985 that problems with the shuttle's booster joints might lead to a catastrophe, said: ''The capture feature was to be the solution. It was clearly a big step. But it was happening very slowly.'' ''The engineers knew about the big fix,'' he said, but until it was carried out, ''they held their breath every time the shuttle took off.''

John E. Pike, director of space policy at the Federation of American Scientists, called the situation at NASA ''shocking.''

''They were clearly covering up the magnitude of the problem,'' Mr. Pike said. ''It raises the question of who knew what when. It's one thing to say this was a dinky little problem that a few engineers worried about, and another to embark on a major redesign of the system.''

The fact that the space agency applied the ''solution'' as early as 1982 and again in 1985 could affect litigation over the Challenger disaster, according to a lawyer representing Cheryl McNair, widow of Dr. Ronald E. McNair, one of the Challenger astronauts.

''Why fix the booster if it wasn't broken?'' asked Ronald D. Krist, a Houston lawyer who has filed suit in Texas in behalf of Mrs. McNair.

''It makes the fact that they forced this launch all the more reprehensible,'' Mr. Krist said. ''It shows a conscious indifference to the welfare and safety of the astronauts. Rather than postpone a launch, they pressed ahead.''

Mrs. McNair has sued the booster maker. In addition, Jane Smith, widow of the Challenger's pilot, Comdr. Michael J. Smith of the Navy, has filed an administrative claim against the Government.

In defending past actions, NASA officials insist the depth of the booster problem was unknown before the Challenger disaster. Casing design improvements, they assert, were part of a wide-ranging effort to make all flight equipment as safe as possible.

''Enhancement of that joint was just one several things we were working on,'' said James E. Kingsbury, director of science and engineering at the Marshall Space Flight Center in Huntsville, Ala., which has responsibility for design of the booster rocket. ''No one recognized the weakness it contained.''

Thomas L. Moser, deputy associate administrator (management) for space flight at NASA headquarters in Washington, agreed. ''In fairness I don't think you can say NASA was flying with a system it regarded as inadequate,'' he said. ''The design was seen as marginal but adequate. It's just that there was an opportunity to increase the margin of safety.'' History of Joint Is a Complex One The history of the capture feature begins in 1974 when the Thiokol Corporation, now part of Morton Thiokol Inc., won the contract to design and build the shuttle's solid-fuel booster rockets. They were to be 12 feet in diameter and 149 feet long. For ease of assembly, the boosters were built in parts, including four steel-walled, 27-foot-long fuel segments. These were cast in Utah and shipped to Florida, where they were bolted to one another with 177 steel pins at each seam. For strength at each joint, a three-inch tongue from one steel case fit snugly into a corresponding groove on the next, locking the segments tightly together. Or that was the plan.

In testing in 1977, the joint was found to be seriously flawed. As the rocket ignited, explosive pressures puffed out its half-inch steel walls. At the same time, the joint, where metal was thicker, resisted this motion. The result was a slight shift or rotation of tongue and groove in relation to each other, creating a gap between them. The design called for this gap to be filled by two rubbery O rings meant to prevent gases burning at 5,900 degrees Fahrenheit from leaking out the rocket. The wider the gap, the greater the danger that the joint would not seal.

The discovery of the rotation prompted a series of memorandums at Thiokol in Utah and Marshall in Alabama. The existing tongue-and-groove joint was nonetheless deemed safe enough to fly. Tests showed that at least one of the O rings would hold in each joint. And they did, with few problems, in the first few shuttle flights.

Soon, however, two design changes to improve booster casings made joint rotation worse.

In April 1981, as the first shuttle flew into orbit, NASA engineers were planning two new generations of booster rockets. The first was to have lightweight steel cases so shuttles could carry heavier payloads into space, making the winged spaceships more cost-competitive. Shaving a few hundredths of an inch of steel from casings would lighten them by 4,000 pounds, or about 4 percent.

The second generation of casings was to replace steel altogether with lightweight carbon filaments and epoxy resins. These filament casings were viewed as critical for impending flights from Vandenberg Air Force Base, Calif., where shuttles would be sent southward into a polar orbit more difficult to achieve than the more equatorial orbit flown by shuttles launched from Kennedy. Vandenberg shuttles had to be as light as possible; filament casings made each booster 33,000 pounds lighter.

But both new casings were more elastic than the old ones, and their walls ballooned even more on rocket ignition, increasing joint rotation.

''We worried about the lightweight steel case,'' said Kenneth W. Jones, a Marshall booster engineer. ''It aggravated the rotation. And the filament would aggravate it even more. So we were looking for a way to reduce it.'' Alternatives for the lightweight steel casings were limited because it was simply a pared-down version of the old one. No fundamental change could easily be made in its joint. So attention focused instead on the filament casing and its joint, which were altogether new. Another Company Enters the Picture In 1981, engineers from Marshall discussed the rotation problem with representatives of the aerospace division of Hercules Inc., in Murray, Utah, and Thiokol, the two companies seeking the contract for the filament casing. Later, in preliminary design studies, both companies sketched joints nearly identical to the steel-case joints. But on Dec. 4, 1981, a Hercules engineer hit on an altogether new idea.

''I'd been mulling over the problem for months,'' Frederick Policelli, the inventor, recalled in a recent interview. ''Finally it just popped into my head. I was on my way out the door, so I sat down and sketched it out.'' At the time, Mr. Policelli was technical manager of the filament booster program at Hercules. His idea, which he called a ''capture lip,'' came to him as he was leaving a meeting with Marshall officials in Alabama.

The capture feature was envisioned as a one-inch appendage on the tongue that inhibited movement of the opposing groove, which held the pair of critical O rings. The whole joint was to be machined from one solid piece of steel to be tightly bonded to the filament case. The capture feature would give the wobbly joint stability not unlike that of firm handshake.

In early 1982, Thiokol and Hercules submitted rival bids and designs for the filament-cased boosters. Thiokol proposed the same joint as on the steel case. Hercules proposed the capture feature.

''I jumped on it and championed it,'' said William L. Ray, an engineer at Marshall who had long criticized the booster joint as unacceptable. ''There was a lot of opposition.''

''People didn't understand'' the joint or its problems, Mr. Ray said.

Influenced by the strong advocacy of Mr. Ray and other Marshall engineers, the space agency in May 1982 awarded Hercules the contract to develop filament casings. ''The capture feature wasn't the only reason they won it,'' Mr. Ray said, ''but it clearly was a plus.'' Bitter Fighting Between Competitors The capture feature received no patents for originality or awards for elegance. But it did turn a flawed joint into one that was much less susceptible to rotation.

It also injected an element of corporate rivalry into the manufacture of shuttle boosters, one that flared into bitter fights between Thiokol and Hercules. The new joint, after all, could be viewed as an indictment of the old one. Tension ran especially high because Thiokol was still to be in charge of filling the filament casings with fuel and readying them for flight. At one point in late 1982, Thiokol, exerting its influence, stopped Hercules from releasing plans for the filament joint to a forger, the Ladish Company in Cudahy, Wis.

''It came to a showdown,'' Mr. Policelli of Hercules recalled. ''We went to NASA and they supported our side.''

A spokesman at Thiokol's facility in Brigham City, Utah, defended the company's unwillingness to add a capture feature, saying that even today there are questions about its effectiveness. ''It takes time to learn what changes are necessary,'' said Rocky Raab, the Thiokol spokesman.

At the time, the future of the filament casings was additionally unclear because NASA officials were unsure whether the new design might become the mainstay of the shuttle fleet, replacing steel casings altogether. Use of the lightweight filament casings would greatly increase the weight of the payload each shuttle could carry into space.

In 1982, the space agency's master plan was to begin using filament casings in Vandenberg launchings by October 1985. At first, these filament casings were to be used once and discarded, unlike steel casings used at Kennedy that were meant to be cleaned, refilled and reflown up to 20 times. But over time, with more research and development, filament casings, too, were to become re-usable, which would make them practical for use throughout the shuttle program.

Re-useability ''was a goal, not a firm requirement,'' said a Hercules official in Utah, who spoke on the condition of anonymity. ''We were to see if it would work for the program as a whole. And one of the beauties was that it had the capture feature.'' Putting the Design Into Practice In 1982 and 1983, however, space agency engineers were loath to gamble the program's future on the successful development of reuseable filament boosters. Hedging their bets, they discussed the possibility of creating steel casings with a capture feature, according to Mr. Ray. All it would take was casting new steel casings with an additional three inches of steel around the inside of one end, into which a capture lip could be machined.

Each two-part, 27-foot fuel segment would then comprise one new casing with the capture feature and one old casing, the pair sturdily joined at the factory in Utah. These segments, each with a capture feature at one end, could then be pinned together at the Kennedy Space Center. Alternating new and old casings would mean that, as the shuttle fleet expanded, all of the old metal casings could be used.

But there would be drawbacks as well. Capture features would add about 200 pounds each, or a total of 600 pounds per booster rocket - a small but significant amount. Assembly of segments with a capture feature would also be more complex. Moreover, engineers estimated it would take 27 months for new steel casings to be forged, machined, filled with fuel and readied for flight. Finally, in 1982 and 1983, no one was sure the capture feature was the best solution.

''There was great interest in incorporating the capture feature in the steel case,'' Mr. Ray of Marshall recalled. ''The reason it wasn't done quicker is that we wanted to be sure it worked. That joint looks simple, but it's very complex. You don't want to accidentally make it worse.'' Marshall decided to wait for results from the testing of the filament case, and then make a decision about using it in new steel casings.

Meanwhile, steel booster performance problems increased, in part apparently because of adoption of the lightweight steel casings. These were first used to boost the shuttle aloft in April 1983 and thereafter became a standard feature of all flights. In 1984, O rings in lightweight steel boosters were found to have been damaged by hot gases on successive shuttle flights in February, April and August.

Fortunately, a rich harvest of data had been gathered starting in late 1984 on the effectiveness of the capture feature. The first full-scale filament test firing took place in Utah in October 1984, and others quickly followed. Moreover, laboratory tests were begun to investigate the dynamics of the capture feature. This testing program ended in the spring of 1985. ''We were convinced the capture feature was a good thing,'' Mr. Kingsbury of Marshall recalled. Problem Is Seen With the O Rings Then, in June 1985, engineers discovered for the first time that hot gases had gone past a primary O ring on a shuttle flight and had started to erode a secondary ring.

A month later, Marshall officials responded by putting a ''launch constraint'' on all shuttle flights; from then on, the latest available information on joint problems would have to be discussed in readiness reviews before each flight.

The steel booster program had reached a turning point. In July 1985, more than three years after NASA decided to put the capture feature in the filament casing, Marshall and Thiokol officials acknowledged that the steel boosters needed them, too. They ordered the Ladish Company to begin forging 72 new steel casings for the booster rockets, according to secret testimony recently made public by the Rogers commission and in Congressional hearings.

On Aug. 19, 1985, Thiokol and Marshall officials briefed officials at NASA headquarters in Washington on the erosion of the O rings. Documents prepared for the briefing made reference to the ''potential long-term solution,'' the capture feature. The documents said it could have its ''earliest possible implementation'' with a shuttle mission then scheduled for August 1988.

The Rogers commission later concluded that O ring erosion information presented at this August meeting had been ''sufficiently detailed to require corrective action prior to the next flight.'' But nothing was altered in the following shuttle missions.

On Jan. 16 and 17 of this year, less than two weeks before the Challenger disaster, NASA engineers and those from Thiokol met at Marshall to review the history of O ring problems and possible solutions. The first item on the agenda was the capture feature, its status in the filament casings, and its future in the steel ones. A briefing paper prepared for the meeting noted that steel casings with a capture feature were due for delivery by February 1987.

On Jan. 28, the Challenger exploded, killing all seven crew members. According to the Rogers commission, the disaster was triggered when hot gases leaked past O rings through the lower field joint of the right booster rocket. Work Is Accelerated After Explosion On Feb. 4, Marshall officials met with shuttle contractors to discuss ''acceleration'' of work on new steel casings, according to documents released by the Rogers commission. The February 1987 arrival date was abandoned for more urgent application of the new design.

Lawrence B. Mulloy, who then headed the booster program at Marshall, told the Rogers panel six days later, ''We now will have the first of those parts delivered in 1986, in August of 1986.''

Mr. Mulloy's testimony, at the commission's first closed session, contained his first reference to the new booster casings. His remarks troubled the commissioners.

Neil A. Armstrong, former astronaut and panel vice chairman, said he thought the public would not understand why the proposed solution to the booster joint problem had not already been applied.

Mr. Armstrong said: ''I am concerned, Mr. Chairman, in looking ahead at the perspective of the audience and the people listening, that we will have demonstrated that there was a concern in this particular technical area, but it wasn't deemed dangerous to fly, it was deemed safe to fly, but it was also deemed that it ought to be fixed and be better, and work was going on, which makes an understandable story to me, but I am not sure that it will be understandable to everybody else.''

'' 'Why was it safe?' '' he said people would wonder. '' 'And if it was, why were you fixing it?' And that concerns me a little bit.''

''It concerns us as well,'' said Jesse Moore, then director of the shuttle program.

Dr. Richard F. Feynman, a Nobel Prize winner in physics and one of the panel's most outspoken members, agreed. ''It does look terrible,'' he said.

Mr. Rogers, the chairman, turned his colleagues away from this line of questioning, saying their investigation should focus on events leading up to shuttle mission 51-L, Challenger's last, rather than on planning that was under way at NASA. Mr. Rogers said: ''If you went through all of the things that Larry has gone through and then the last question is, does any of this relate to 51-L, and the answer is no, everybody is going to say, well, why did you spend all of that time on it?''

Although the commission's final report did not deal with the capture feature's history, much of it is detailed in more than 122,000 pages of documents and 12,000 pages of testimony the panel made public.

In May the first of the new steel cases was delivered to a subcontractor in California for machining, including cutting the capture feature.

On June 6 the Presidential commission issued its final report. The text of the 256-page document made no mention of the capture feature or the 72 new steel casings.

On June 12, L. Michael Weeks, NASA's deputy associate administrator (technical) for space flight, told a Congressional hearing, ''The 72 casings were ordered because we knew the rotation problem was serious.'' The fact that the work got under way in July 1985 ''is now saving us six to nine months in the implementation program,'' he said.

Asked if shuttle flights should have continued in 1985 despite the program to fix the boosters, Mr. Weeks responded that NASA officials believed continued flights were reasonable given their knowledge at the time.

On Aug. 12 at a news conference in Alabama, Marshall unveiled its redesign of the booster joint. The main element was the capture feature.

By William J. Broad

Dr. James C. Fletcher, the NASA Administrator who awarded the contract for the space shuttle rockets that led to the Challenger disaster, says that he will ''seriously consider'' removing himself from all future determinations on shuttle booster contracts.

His statement came in an interview on Friday after he heard a review of accusations, made by Senator Ernest F. Hollings, competitive contractors and various aerospace experts, that his original decision could appear to have been biased in favor of a Utah company.

Dr. Fletcher insisted that the decision was ''squeaky clean'' and was made on its technical merits, containing no hint of personal favoritism. He said he did not recall any questions of bias being raised in 1973, when the contract for the shuttle booster rockets was awarded to Morton Thiokol Inc. of Brigham City, Utah. Review of Booster Decision

Dr. Fletcher, appointed again this year to head the National Aeronautics and Space Administration, spoke in reaction, also, to information contained in newly disclosed or re-examined documents and provided in dozens of interviews, part of an investigation by The New York Times over the past several months into the origins and loss of the Challenger.

Those documents and interviews yielded voluminous detail, kept secret for years, about the deliberations and doubts, charges and countercharges, that swirled about NASA when the fateful contract for the shuttle boosters was awarded. 'Suggestion of Impropriety'

While denying any bias, Dr. Fletcher said he would consider disqualifying himself from new contract decisions because such charges are now being brought up in the aftermath of the Challenger disaster on Jan. 28. Seven astronauts died when a joint failed in one of the giant rockets, destroying the craft after it left the launching pad at Cape Canaveral.

''We didn't have a ripple back then on the Utah connection,'' Dr. Fletcher said of the 1973 decision. ''But now we've had a big failure in the solid rocket, which was a Morton Thiokol rocket, and the suggestion of impropriety has been raised. So with all that going on, I would seriously consider recusing myself from decision making.''

Dr. Fletcher's critics maintain that through his close ties to Utah he had a personal, although not financial, interest in the award of the contract to Thiokol, saying that he sat on the executive committee of a Utah industrial development corporation at the time and that two former Thiokol employees held top positions on the NASA panel that evaluated the rocket proposals. In giving the contract to Thiokol, the critics say, Dr. Fletcher ignored a rival design that in crucial respects was inherently safer in the area of the booster rocket where the Challenger explosion occurred.

Dr. Fletcher said in the interview that what is currently troubling about the charges against him is the ''perception'' of bias, insisting several times that he believed none, in fact, existed.

A two-decade-old Presidential order forbids Federal officials from actions that might result in, or create the appearance of, ''giving preferential treatment to any organization or person.''

Dr. Fletcher's critics say their real concern is not so much what happened in the past as the possibility of mistakes being repeated as the space agency weighs whether to hire a new company to build corrected rockets. NASA is to complete studies on such plans and report to Congress early next year. Fletcher Reappointed After Shuttle Failure Dr. Fletcher, a 67-year-old physicist, directed NASA from 1971 to 1977. President Reagan reappointed him as administrator last March, after the Challenger accident, to pick up the pieces of a shattered program and return the nation to the forefront of space exploration.

Some key new points that have come to light now in the saga of the rocket contract are these:

* A rival company, worried about astronaut safety, in 1973 specifically warned that ''integrity of the O ring is largely a matter of faith.'' Thirteen years later, it was the rubbery O-shaped seals that failed between segments of a Thiokol booster, triggering the Challenger disaster.

* By the early 1970's NASA researchers had already conducted a separate 10-year, $150 million program to develop very large booster rockets, and they considered only those with no segments, joints or troublesome O rings. Contrary to customary agency procedures, officials of that program were excluded from deliberations on the final choice of booster rockets for the space shuttle.

* A principal advantage of segmented boosters, which Thiokol and two other bidders proposed, was that they could be made far from Cape Canaveral and shipped to the launching site in pieces. Thiokol had no nearby plant, its Utah facility being thousands of miles away. As it happened, the 1973 Arab oil embargo, which got under way several weeks before the Thiokol contract was awarded, was already beginning to drive up transportation costs.

The newly disclosed documents show, too, that a president of the Mormon Church, of which Dr. Fletcher is a lifelong, active member, lobbied vigorously for Utah to win the $800 million NASA contract. Critics Cite a Web Of Utah Connections Nothing in the story as it is now known suggests that the Thiokol award resulted in personal financial gain for Dr. Fletcher, who was a rich man before the decision and whose financial disclosures to Congress now put his total at $2.3 million to $4.8 million.

Critics charge that it was his complex web of Utah connections that resulted in favoritism for Thiokol, or in his failing to follow procedures to avoid any appearance of it, and in his rejection of a rocket design that excluded the potential for an O ring catastrophe.

''The only reason we selected that terrible design to begin with was that Fletcher was a Mormon and from Utah,'' William C. Bush, an engineer recently retired from the Marshall Space Flight Center in Huntsville, Ala., said in an interview. The Marshall center has responsibility for the booster contract.

Edward Z. Gray, NASA's assistant administrator for industry affairs during the 1973 competition, objected to the two former Thiokol employees' being on NASA's rocket evaluation board: ''That was unusual and unfortunate. It shouldn't have been done that way.'' He called Dr. Fletcher ''a square shooter'' but said there ''probably were some errors in judgment.''

John E. Pike, director of space policy for the Washington-based Federation of American Scientists, said: ''It's difficult to avoid the conclusion that this is bias and favoritism. Certainly the record suggests that Thiokol should not have gotten the contract.''

Malcolm McConnell, a longtime aerospace reporter who is the author of ''Challenger: A Major Malfunction,'' a forthcoming book on the roots of the shuttle disaster and the Mormon influence in the 1973 booster award, said: ''At the very least, Fletcher should have removed himself from the selection process.''

And Mr. Hollings, a South Carolina Democrat who is in line to become chairman of the Senate Committee on Commerce, Science and Technology in the next Congress, said at a Senate hearing in May that he believed a ''tremendous disaster'' centers on the NASA director:

''After all, Dr. Fletcher was from Utah and he was president of the university, and the Mormon crowd out there, and he married a Utah citizen, and the chairman of the space committee in the Senate was from Utah.'' The Senator charged that ''it was a Utah conspiracy that got Morton Thiokol into this,'' and said Thiokol failed to ''pay any attention to the solid-rocket booster. That's why it's segmented. That's why we have an O-ring problem.'' Records Show Fletcher Tied to Utah Lobby Dr. Fletcher, who has been active in church work, and his wife, a Brigham City native, both come from Mormon families with ties to Utah, where the church exerts a powerful influence not only on religious and social life but also on commerce and industry. The university that Senator Hollings referred to is the Mormon-founded University of Utah, where Dr. Fletcher was president from 1964 to 1971.

At Dr. Fletcher's Senate confirmation hearing in March 1971, Utah Senators Frank E. Moss and Wallace F. Bennett hailed his appointment as NASA Administrator, made by President Nixon.

Then Dr. Fletcher went through his finances, noting what aerospace stocks he would sell to avoid any potential conflicts of interest and what advisory groups he would give up. He did not mention Pro-Utah Inc., a lobbying corporation formed by of Utah businessmen and Mormon officials.

According to documents obtained from the Utah State Division of Corporations, Dr. Fletcher served on the board of directors of Pro-Utah beginning in 1965. And the public record at the 1971 confirmation hearing contains a biographical sketch listing Pro-Utah under the heading ''current affiliations: miscellaneous.'' It says Dr. Fletcher was on the corporation's executive committee.

Dr. Fletcher said Friday he does not recall being a part of that corporation. ''I worked with those folks very closely,'' said Dr. Fletcher. ''We were trying to promote Utah as a high-tech state, successfully. But I don't ever remember being a member of that group.''

He added, however that, if he did serve with the group, he does not recall ever taking steps to sever such ties.

According to its articles of incorporation, the group's goal was to ''promote and encourage the development and retention of businesses, industries and commerce within the state of Utah.'' On its board had sat such prominent Mormons as N. Eldon Tanner, one of the church's presidents.

Raymond L. Hixson, Pro-Utah's president in 1973 and a top official before that, said he visited Dr. Fletcher as NASA Administrator to urge him to select Utah as the site for the entire launching and landing facilities of the shuttle, adding that he had no discussions with Dr. Fletcher about the booster award. Nothing ever came of the Utah shuttle station.

Dr. Fletcher said he remembered no Pro-Utah visits while he was in office. An Explosion Before the Explosion In March 1972, Dr. Fletcher decided that NASA would use solid-fuel rockets rather than liquid ones to loft the newly proposed space shuttle into orbit. The boosters, 156 inches in diameter, would be the world's largest solid-fuel rockets. Dr. Fletcher said they were an economy measure, resulting in an overall saving of about 6 percent over liquid-fuel designs.

Four companies steeled themselves for the competition, knowing the financial rewards would be enormous.

Now the key issue was whether the 14-story boosters would be made in one large piece near Florida or cast in segments and shipped to Cape Canaveral for assembly.

NASA's own skills, exclusively with monolithic rockets, had been honed when it aspired to build a giant solid-fuel booster 260 inches in diameter, nearly twice the size of the shuttles' boosters, to loft the upper stages of the Saturn moon rocket for use in advanced space exploration. Two companies had tackled the project: Thiokol set up a plant on a 7,000-acre Georgia tract near waterways, while the Aerojet Solid Propulsion Company did the same on 75,000 acres in Florida. The research project was directed by NASA's Lewis Research Center in Cleveland.

Then, on April 11, 1965, Thiokol's giant rocket casing, filled with pressurized water to test its ability to withstand the forces of a live engine, blew up.

Two months later, NASA canceled Thiokol's booster contract, saying Aerojet had succeeded in similar tests and that it would take Thiokol too long to correct the welding problems in its rocket cases.

''At that point I made up my mind to try to make reliable joints rather than reliable welds,'' recalled Dr. Harold W. Ritchey, then Thiokol's president. It was a decision esepcially important for Thiokol: By casting giant rockets in sections, and using joints to connect them, the company could consolidate its operations in Utah and ship segments anywhere. These segments were also easier to inspect than monolithic rockets. The drawback was that with joints hot gases and flames always threatened to burn through.

James J. Kramer, head of the Lewis effort until 1966, said some experts feared that the monolithic fuel, hard to inspect for bubbles and cracks, might contain hidden paths allowing a dangerous quick burn toward the metal casing. But he said improved techniques and expertise essentially removed those doubts.

Despite the progress, the booster-launched Saturn program was eventually set aside in favor of the reusable space shuttle, seen as yielding long-term savings.. Four Companies Battle For Lucrative Contract Girded in 1972 for the contractual war on the shuttle booster contract, Aerojet, with its Florida facility ready to build large monolithic rockets, suffered two major jolts. First, it learned that NASA's Lewis Research Center, the world expert in large solid rockets, was to have no role in the booster evaluation. Instead, it would be directed by the Marshall Space Flight Center, which had virtually no such experience. Second, NASA let all contractors know it wanted segmented rockets.

''It was a stacked deck,'' said Richard F. Cottrell, then president of Aerojet. ''We saw no reason for it to be exclusively segmented. We fought it. We convinced NASA that their request should be modified so the design could be monolithic or solid.''

Even so, Aerojet's fortunes began to slip. And Thiokol's rose.

After the November 1972 elections, the chairmanship of the Senate Space Committee fell to Senator Moss of Utah. From his new position of power, he lobbied on behalf of Thiokol, including Dr. Fletcher in his rounds.

In a letter to Senator Moss dated Jan. 12, 1973, the NASA Administrator himself raised the issue of cost, which was to plague the agency for the next 15 years and, along with the consequent need to cut back safety and design testing, became a key factor, according to investigators, in the Challenger disaster.

''While I am of the opinion that your state indeed does qualify for some of NASA's work, yet I wonder just how feasible such an arrangement might be in the long run?'' Dr. Fletcher wrote. ''From our way of looking at it, would it be cost-effective enough for us in the long run?'' But Dr. Fletcher went on acknowledge the keen interest Utah had in the contract. ''I know that President Tanner and various of your state officials have manifested an unusual zeal in hopes that NASA would send some of our business your way.''

Though the contract decision was Dr. Fletcher's, the space agency, as was customary, formed an advisory body in June 1973 known as a source evaluation board, which eventually had a staff of 289. NASA headquarters kept its personnel list secret, but it became general knowledge that the Lewis researchers were absent.

Normally, leaders of a major NASA development program would be relied upon heavily to help in the final evaluation, said Carl C. Ciepluch, then head of the solid-fuel rocket program at Lewis. ''In this case we never got asked. I don't know what happened.''

Dr. Fletcher agreed in the interview Friday that the Lewis researchers should have taken part, though he himself was not certain they were never asked about the booster.

Two top officials of the source evaluation board were former Thiokol employees. In response to complaints by competing contractors, Bart Fugler, director of NASA security, was asked to look into the matter.

''Bart Fugler determined that Mr. R. Brown and Mr. J. Burks had worked for Thiokol before coming to NASA,'' Mr. Gray, the NASA official, wrote in a letter dated Dec. 17, 1973, that went to Dr. Fletcher and other top officials well before he announced his final approval of the award on June 27, 1974.

Richard L. Brown, a former Thiokol engineer, was chairman of the advisory panel's management evaluation team. A. Jack Burks was vice chairman of the overall panel. Years before. Mr. Burks had worked on Thiokol's rocket casing that failed, and subsequently left the company.

William C. Rice, a retired Marshall manager who was then a top official of the evaluation board, said he knew of no other former industry employees among the top six officials. And Mr. Fugler reported finding only former Thiokol employees.

Dr. Fletcher said he did not recall having any knowledge that the former Thiokol employees were on the board. But both he and NASA's general counsel, John E. O'Brien, said that if they were on the panel, space-agency procedures would have forced them to have severed any financial ties to their former companies, such as stock or pension plans, and no one has suggested that these rules were not followed. Mr. O'Brien said the two men's former connection to Thiokol would have made no difference in the overall booster-selection process.

In August 1973, the four companies vying for the contract sent bids to NASA. The designs of Thiokol (then known as the Thiokol Chemical Corporation), the Lockheed Propulsion Company and United Technology Center were for segmented boosters. Aerojet's was monolithic.

The Aerojet document noted that 99.9 percent of the nation's solid rockets were monolithic: ''A one-piece case is always superior in inherent reliability, performance and economy to a segmented solid rocket motor.'' The advantage of segmentation, it went on, ''is solely to satisfy facility location and transportation constraints at the ultimate compromise of crew safety.''

In an unfortunately prescient passage, Aerojet faulted the rubbery seals of segmented boosters: ''Acceptance of the integrity of the O ring is largely a matter of faith.'' The other contractors with segmented designs clearly had confidence that their boosters would work. The Utah Contractor Gets the Prize On Nov. 19, 1973, top officials of the source evaluation board presented their findings to Dr. Fletcher. The competitors had been scored on design, manufacturing, management and cost.

On design, Dr. Fletcher was told that Lockheed ranked first, United Technology second, Aerojet third, and Thiokol last. On manufacturing ability, Lockheed ranked first and Thiokol second.

Thiokol's score rose dramatically on management, on which it ranked first. In charge of management evaluation was Mr. Brown, the former Thiokol engineer. In his hour-and-a-half presentation to Dr. Fletcher, he took the unusual step of presenting cost figures, although another official was in charge of the cost category, according to NASA records marked ''sensitive'' and still never released by the space agency but now made available by the House Science and Technology Committee.

As to cost, Mr. Brown said Thiokol's bid was lowest by far.

The next day, on Nov. 20, 1973, Dr. Fletcher announced his decision.

Rivals were stunned. United Technology hinted darkly at conspiracy, according to a NASA memo of a meeting on Nov. 29: ''Thiokol was overconfident throughout the competition, as if they knew what the outcome would be.''

Lockheed, which had the winning design, took public action. On Dec. 5, it filed a protest with the General Accounting Office, the investigative arm of Congress.

Now, again, costs became the key. Lockheed charged that sizable errors had been made in NASA's evaluation of costs. The contract award was put on hold, and the G.A.O.'s mandate was strictly to evaluate its technical merits; the investigators were not charged with exploring issues of bias.

As controversy swirled, Senator Moss published Dr. Fletcher's long selection statement in the Congressional Record of Dec. 14. It was the first public glimpse of his reasoning. He first weighed mission suitability, which included design, manufacturing and management. Lockheed was the winner, Thiokol and United Technology tied for close second and Aerojet was last.

''The monolithic case design proposed by Aerojet precluded the potential failure modes associated with joints and seals,'' Dr. Fletcher noted, but he went on to fault the company's overall proposal, especially its plan to move from California to Florida while work was in progress.

He concluded that the main criticisms of Thiokol's mission suitability ''were technical in nature, were readily correctable, and the cost to correct did not negate the sizable Thiokol cost advantage. Accordingly, we selected Thiokol.'' Inquiry Cited Flaws In Figuring the Costs The war over the booster award came to a head on April 23, 1974, when officials of Lockheed, Thiokol and NASA met with General Accounting Office analysts. On June 24, the G.A.O. issued its verdict: NASA should weigh a reconsideration because of flawed reasoning on costs.

The bids of Lockheed and Thiokol were both in the $800 million range, G.A.O. said, but a mistake had been made. The source evaluation board erred in finding that Thiokol's bid was $122 million cheaper than Lockheed's because, the G.A.O. said, it allowed a large difference in favor of Thiokol on the cost of ammonium perchlorate, a component of rocket fuel, even though both companies planned to buy this chemical from the same source.

This readjustment lowered Lockheed's bid by $68 million and thus made Thiokol only $54 million cheaper.

The accounting office criticized NASA for not creating a uniform inflation scale for the 15-year contract, noting differences in the assumptions of both companies. Lockheed claimed that a proper adjustment would cut its bid by $60 million and that this would actually make its overall cost lower than Thiokol's.

In closing, the investigators noted that Dr. Fletcher had recently rejected the low bid in a different $800 million competition for the shuttle's main engines, even though the technical merits of rival proposals were close.

On the next day, June 27, NASA headquarters issued a terse news release saying Dr. Fletcher had upheld his previous decision.

''The initial rationale for the selection of Thiokol remains valid, even assuming, but without conceding, the correctness of the position taken by G.A.O. respecting ammonium perchlorate costs,'' the release said.

Events outside NASA's control had already started to erode the award's underlying rationale. The October 1973 Arab oil embargo ended the era of cheap oil, fueling inflation and sending transportation costs in a steady upward spiral. By November 1975, Marshall officials were worried. ''Our major concern was the escalation of rail rates since the award of the contract,'' one official wrote in meeting report, noting a 41 percent rise.

For this and other reasons, Thiokol's bid of $800 million on the 15-year contract proved to be very low indeed. To date, NASA has spent nearly $2 billion on shuttle boosters. In addition, it is now spending some $300 million to correct the flaws in Thiokol's original design.

By William J. Broad

New information warrants a Congressional investigation into the contract for the space shuttle rockets that failed last January, a key senator said yesterday. Another said the NASA Administrator should remove himself from future decisions on such contracts.

But the idea that home-state favoritism had led the official, Dr. James C. Fletcher, to ignore a safer design in giving the rocket contract to a Utah company was dismissed as ''nonsense'' by a third senator, from Utah.

The three lawmakers, who play important roles in space issues, were asked for their reactions to new information from dozens of documents and interviews, reported yesterday in The New York Times, about Dr. Fletcher's close ties to Utah, where Morton Thiokol Inc. makes the rockets. Dr. Fletcher headed the National Aeronautics and Space Administration in the 1970's when the shuttle program was begun and was named to run the agency again this year.

The rocket is currently at the heart of a battle over who will redesign and rebuild it to correct the flaw that led to the Challenger disaster. On Jan. 28, a leaky booster rocket built by Thiokol touched off an explosion that destroyed the spaceship 73 seconds after liftoff and killed its seven astronauts.

At stake now, according to a variety of aerospace experts, is a new generation of rockets that may be cheaper and must be safer. For two key players, rival aerospace contractors, the current battle is a continuation of a 25-year-old competition.

''Dr. Fletcher apparently allowed home-state boosterism to compromise the integrity of the contract selection process,'' said Senator Albert Gore Jr., a Tennessee Democrat who sits on the primary Senate subcommittee on space. ''He should immediately remove himself from any decisions involving future shuttle booster systems.''

The Senator added that he would call for the General Accounting Office, a Congressional investigative agency, to look into the original contract award.

Senator William Proxmire, a Wisconsin Democrat, said, ''In view of Dr. Fletcher's involvement in the initial solid rocket booster decision, as well as his substantial connections with the home state of the present contractor, it would be wise for him to delegate the final decision regarding future booster production.'' Mr. Proxmire is in line to become chairman of the Senate subcommittee overseeing NASA's budget.

Dr. Fletcher has denied the existence of any bias on his part. But in an interview Friday, he agreed that his Utah ties might be seen as creating the appearance of favoritism and said he would ''seriously consider'' withdrawing himself from any new contract decisions on the boosters.

Senator Jake Garn of Utah, a Republican, came to Dr. Fletcher's defense when asked to comment on the critics' allegations. The bias charges raised against the NASA director are nonsense, the Utah Senator said through an aide yesterday. Mr. Garn, who flew on a shuttle mission himself last year, insisted that Dr. Fletcher had always bent over backward to avoid any appearance of favoritism. There was, he says, no Utah bias: ''These things are done in the cold light of day.''

The disclosures in the documents and interviews, which form the basis of the concerns of Senators Gore and Proxmire and others, include these key points: that Dr. Fletcher sat on the executive committee of a Utah industrial development corporation at the time he joined the space agency, that two former Thiokol employees were among the six top officials of a NASA booster evaluation board, and that this board contained no officials from a separate 10-year NASA research program to develop large booster rockets.

Within three weeks, NASA is to complete studies of alternative booster designs and early next year is to report to Congress on whether to stay exclusively with Thiokol, whom Dr. Fletcher awarded the original booster contract 13 years ago, or hire a new company to produce corrected shuttle boosters. The coveted contract will be worth hundreds of millions of dollars. Benefit of Competition Is a Primary Question The current push for competition involves two distinct phases, which often blur. The first is to whether immediately acquire a second source of booster rockets to complement those built by Thiokol and to generally resemble them. The second is to switch, in the 1990's, to an altogether new, unspecified type of booster.

In the months immediately after the Challenger catastrophe, NASA resisted any change in the status quo.

''The potential for competition is there but NASA's fighting it,'' said James H. Dodd, who heads the Washington office of the chemical systems division of United Technologies, a potential booster rival.

One reason, NASA critics say, is that over the past 13 years the agency has grown comfortable with its Thiokol relationship and the accident did not change that. ''It's not unusual for the firm that develops a system to be its producer forever and ever,'' said Lee A. Edwards, a Congressional investigator who routinely examines NASA procurements for the General Accounting Office. ''The Government team just gets comfortable. That's not necessarily in the Government's best interest, but it's a fact of life.''

Thiokol officials agree that current pressures may forces some changes, though not on the immediate redesign project.

''Certainly our booster will be on the shuttle when flights resume, assuming we get results with our redesign effort,'' said Thomas S. Russell, vice president for corporate development at Morton Thiokol's main office in Chicago. ''What will happen after that is a hard one. From an economic point of view, it doesn't appear to make sense to go to a second source. Will there be political pressures? Will they make a difference? That's another question.''

Ever since the booster contract went to Thiokol in 1973, NASA has weighed the possibility of having other contractors bid for a piece of the huge production job. The reasons were simple. Competition was viewed as a good way to cut booster costs and to hedge against delays, stoppages and disasters. But for a variety of reasons, NASA never achieved that goal.

Then, in the 1980's, serious accidents began to strike the Thiokol plant in Utah, which a General Accounting Office report cited as pointing up the danger of relying on a single manufacturer. In March 1984, a fire broke out in a casting pit, destroying two booster segments and more than $11 million worth of facilities, including $8.6 million of Government property. In June 1985, another fire broke out in a fuel mixing house.

During this period, rival contractors became more and more vocal on the matter of competition, and Congress joined in. In late 1985, before the Challenger explosion, 80 members of the House of Representatives signed a letter calling for a second source of booster production.

Exactly one week before the shuttle disaster, NASA announced its intention to hold a competition for a second source. It cited no immediate cost saving but rather ''national security'' and a need to broaden the nation's industrial base.

After the explosion, NASA quickly dropped the competition plan and focused instead on aiding Thiokol in a $300 million program to fix the flawed booster.

Root of the Problem: A Segmented Rocket At the root of the problem were the booster rocket's four large steel-walled fuel segments. These 27-foot-long segments had thick insulation between casing and fuel to keep the searing heat of burning propellant from melting the walls of the booster. But no insulation covered, or could cover, the spot where segments fit together. Instead, segments were locked together by a tongue-and-groove joint and further protected from the leakage of hot gases by a pair of rubbery O-shaped seals.

It turned out, however, that pressures in the rocket forced the joint open, allowing hot gasses and flame to impinge upon the O rings. A further problem for the Challenger on the morning of Jan. 28 was exceptionally cold weather, which kept the O rings from expanding properly and allowed flames to leak past them outside the rocket and eventually touch off an explosion in the huge adjacent fuel tank full of liquid hydrogen.

NASA and Thiokol are currently proceeding on a redesign plan that calls for joints with additional O rings, for heaters outside the joints to keep them warm in cold weather, and for new casings with features meant to keep joints locked in place.

But in basic design, the booster rocket remains fundamentally the same. Where fuel segments join together, there is still, of necessity, the gap in insulation and thus a path for hot gas and flames to escape.

On April 25, the Aerojet General Corporation offered NASA a proposal to use existing booster cases in a way that altogether eliminated the flame path and the resulting danger. It proposed connecting the cases, covering their joints and all other interior surfaces with uninterrupted insulation, then loading the solid fuel. The result would be, in effect, a one-piece rocket impervious to the type of failure that struck the Challenger. The goal could quickly be achieved, the company said, since it owned a plant in Florida, not far from Cape Canaveral, that had been used in the 1960's for NASA research on solid-fuel rockets. Carolinian Complains Of a 'Utah Conspiracy' On May 8 Senator Ernest F. Hollings, Democrat of South Carolina, faulted space agency officials for taking no action on the Aerojet proposal and, at a Congressional hearing on the NASA budget, charged that a ''Utah conspiracy'' had influenced the awarding of the original contract to Thiokol and was now holding up corrective measures.

In the context of his ''Utah conspiracy'' remarks, Senator Hollings referred to Dr. Fletcher, who had been president of the University of Utah before heading the agency, and to Frank E. Moss, a Utah Republican who had been chairman of the Senate Space Committee in 1973. Senator Hollings went on to say that the Utah connection was still, in 1986, influencing NASA's work, but he named no names.

''We could reinstitute that particular operation down in Florida,'' Senator Hollings said. ''But the Utah conspiracy will not even let us in. Now, that sounds like fanciful nonsense, but it has got some substance to it.''

Senator Hollings said he wanted part of the millions in booster redesign money to be spent on investigating the merits of the Aerojet proposal. ''You have got a tremendous disaster on your hands,'' he told NASA officials. ''But you tippy-toe all around an obvious, logical solution.''

Adm. Richard H. Truly, the head of NASA's shuttle program, responded that the Aerojet proposal was ''not obvious or logical'' since it required a barge and construction of a facility on Cape Canaveral for erecting large one-piece rockets. The current shuttle booster segments, already fueled, are assembled on the mobile launching pad itself.

''You have got to open up your minds and get a second source and get rid of that joint,'' Senator Hollings replied. ''This country is not going to stand for any more joint problems, I can tell you that right now.''

On June 5, Senator Hollings took the floor of the Senate to demand that NASA study booster alternatives, addressing his remarks to Senator Garn. ''I am concerned,'' Senator Hollings said, ''that a 'quick fix' bodes poorly for the future of the shuttle program.'' He said that ''common sense tells me that all options should be considered.''

The Utah Senator graciously agreed. ''I am open to fencing off some of the monies allotted in the fiscal year 1986 supplemental to pursue alternative fixes and to consider parallel programs,'' he said.

Also on June 5, overall Congressional criticism picked up when 258 members of the House, a majority, wrote Dr. Fletcher to ask for ''a full, fair, and open competition'' on a second source for shuttle boosters. The letter said NASA's plans, developed in late 1985, for finally seeking competition bids from a second booster manufacturer, were themselves skewed in favor of Thiokol.

Later in June a House-Senate conference committee, fashioning a NASA appropriation bill, ordered the agency to finance studies of alternative booster designs, including designs with one-piece cases and fuels, and complete the studies by Dec. 31. It also told NASA to develop a second source for boosters if the ''agency determines that such an approach can be adopted with available resources.'' President Reagan signed that bill July 2. 5 Companies Contract To Study New Rocket Early in September, NASA headquarters awarded five companies $500,000 each for studies of a new space shuttle booster rocket. The companies were Aerojet General, the Atlantic Research Corporation of Alexandria, Va., the Hercules Aerospace Company of Salt Lake City, the Wasatch division of Morton Thiokol in Brigham City, Utah, and the chemical systems division of United Technologies in San Jose, Calif.

The only constraint on the contractors was that the new boosters be the general size and shape of the old version, so they fit with other shuttle components. ''With few exceptions,'' the agency said, ''the studies provide the five contractors a free hand to propose a new solid rocket motor.''

NASA's move was quickly applauded. On Oct. 10, the National Research Council, a private group overseeing NASA's booster redesign, wrote Dr. Fletcher: ''We strongly recommend that NASA maintain a program to explore and develop original, possibly quite different designs'' in case the current redesign effort fails to ''offer sufficiently good performance and margin of safety.''

But Aerojet, which believed that NASA had shown favoritism to Thiokol in the past, feared the situation continued. Its officials decided to talk to Dr. Fletcher directly, meeting with him on Nov. 6 in his NASA office. There they outlined their proposal for turning existing segmented cases into the equivalent of one-piece rockets, which inherently preclude the type of accident that destroyed Challenger.

One Aerojet official said later that Dr. Fletcher ''seemed receptive, but I don't think he went across the hall and told his guys to fund us'' - Aerojet people believed that momentum toward Thiokol had not been fundamentally changed.

Mr. Russell of Morton Thiokol, while saying he has no philosophical objection to competition in booster production, says he thinks it will prove economically useless to appoint a second contractor. The cost of making sure that a new contractor's work is safe and reliable, even before production of usable rockets begins, is estimated at $100 million. Ordinarily, NASA would bear these start-up costs but a dispute exists over whether, in this case, the contractors would assume them.

''The marketing rhetoric is all about how much NASA's going to save,'' Mr. Russell said. ''But I don't think $100 million is nearly enough to bring another contractor on line, and the prices they're quoting are unrealistic.''

An overlooked issue, Mr. Russell added, is that fewer boosters are going to be needed in the future because the shuttle's flight rate has been drastically cut in the wake of the Challenger disaster, going from an envisioned 30 flights a year to a maximum of about 15. The cost per booster will rise, he said, if two companies maintain the facilities and personnel needed for such limited production.

Moreover, with the nation worried about the joints, Mr. Russell said, it was unlikely that NASA, after the booster is successfully redesigned, would want to turn over production to an inexperienced firm. ''Are they then going to bring in a second supplier and change everything after the first flight in 1988?'' he asked. ''My guess is no. Why fix it if it's not broken?'' Mr. Russell added that the Aerojet proposal was impractical: ''To my knowledge, none of the infrastructure at the Cape could handle a one-piece booster like that.''

Thiokol does not make large monolithic boosters, which cannot be shipped long distances by rail. In the early 1960's both Thiokol and Aerojet were hired by NASA to develop very large one-piece boosters, but Thiokol's design blew up in a 1965 test, and the company subsequently decided to use the segmented approach in building giant rockets.

The current debate over monolithic vs. segmented boosters echoes one that took place over the original 1973 award, and Fletcher critics charge that Thiokol's location in Utah, far from Cape Canaveral, was a factor in NASA's having settled on a segmented design. Several Factors Cited In Bid for Second Source Today George G. Brown, an Aerojet vice president, argues that price competition between suppliers would clearly make boosters cheaper, cutting the cost from $16 million to as low as $10 million apiece. The low price would be especially feasible if the use of the boosters expanded beyond the manned shuttle program. Several companies, he said, have recently proposed building unmanned ''space trucks'' out of shuttle components. ''As long as there are seven or eight flight sets of boosters per year, it's economically feasible for a second source,'' he said.

Moreover, another factor is a wide range of ''national security'' issues, including whether a single company could be closed by fire, accident, labor problems or natural disaster. ''Right now,'' Mr. Brown said, ''if Thiokol gets shut down, you shut down the whole shuttle program.''

He said that although no equipment at the Cape could currently hoist a single monolithic booster weighing 1.2 million pounds into position, such devices are readily available. He noted that a mobile derrick built by Lampson Inc. can lift up to 1.8 million pounds.

As for launching the rocket into space, there is no weight penalty like that imposed by some of the new steel features in the Thiokol redesign that add 600 pounds per rocket and decrease the amount of payload that can be carried into space, Mr. Brown said.

Aerojet says it can have its boosters ready for flight in about 20 months, but Mr. Brown said the most important point to stress is inherent safety: ''Our approach completely isolates the joint from hot gas exposure and guarantees that the Challenger problem will never happen again,'' he said.

He cited a key recommendation of the Presidential Commission that investigated the Challenger disaster, which advised that in a redesigned booster ''the integrity of the structure and of the seals of all joints should be not less than that of the case walls.'' Mr. Brown said Aerojet has found no way to fulfill this recommendation without the booster's being monolithic. NASA and Morton Thiokol, in jointly reworking the booster, say the segmented design can be made safe.

Representative Bill Nelson recently made a visit to the Aerojet facility in Florida as chairman of the chief House subcommittee on space. ''Considering that it hasn't been used in 19 years, it's in surprisingly in good shape,'' said the Florida Democrat, who flew on a space shuttle mission early this year. ''Their proposal would clearly eliminate the field joint'' and its weaknesses. But as to the economic and political factors that might bear on the plan, ''that I don't know,'' he said, adding that such a decision could be made only after further study.

''The Aerojet proposal is worthy of consideration,'' said a Senate aide who specializes in NASA affairs. ''But a one-piece booster is a big chunk of hardware to move around. And Aerojet doesn't have an operating system right now. I'm a little leery of their optimistic schedule.'' Titan Rocket Failure Fuels Design Debate Advocates of monolithic boosters note that what historically had been the nation's premier example of segmentation success, the Titan unmanned rocket, is no longer so. In April, a $65 million Titan 34D exploded seconds after liftoff when one of its twin segmented boosters ruptured. A long Air Force investigation showed that a critical layer of insulation near a joint, where there is a break in the insulation, had peeled away, allowing flames to burn through the rocket's casing.

Moreover, when a Titan 34D that was to have been launched from Cape Canaveral was recently dismantled, inspection of its segments showed that insulation had peeled away in more than half the segments.

The upshot is that the Air Force is considering a move to one-piece boosters for its next generation of Titan rocket.

''We've had discussions with the Air Force on the possibility of product improvement,'' said a spokesman for Martin Marietta Aerospace in Denver, the prime contractor for the new Titan. Topics have included reducing the number of joints in the solid boosters and eliminating them altogether by adopting a monolithic design, he said. Martin Marietta is currently talking to all major manufacturers of solid-fuel rockets to solicit ideas about new generations of improved boosters.

Twin boosters for the Titan 34D are made by the chemical systems division of United Technologies, which is one of the companies studying new booster designs for the space shuttle. Like Aerojet, it too says it is considering a monolithic design. ''It's one of several options,'' said Mr. Dodd, the Washington official for United Technologies. ''You've got to cover every base.'' Experts say this year's twin segmentation disasters, involving the space shuttle and the Titan rocket, have triggered a fundamental reappraisal of the technology's merits.

Presumably, Senator Gore's call for an investigation of the original Thiokol award will result in an examination of not only the basis for charges of favoritism but also these fundamental questions of booster-rocket design.

For NASA, the added question of competitive procurement, with its complex economics and politics, only makes the process of decision making more difficult, especially as the agency focuses its main attention on picking up the pieces of its shattered space program.