ON THE THIRD Saturday in January, with the crash of Air Florida Flight 90 commanding the attention of the aviation world, senior Federal Aviation Administration experts and Douglas Aircraft Co. engineers met in Long Beach to close the books on another disaster: the DC10 accident in The meeting was congenial, in sharp contrast to the suspicion and animosity that attended dozens of sessions the summer of 1979 between most of the same people. Everyone was relaxed because, despite the fact technical changes were still being made to the jumbo jet, everyone knew -- as the aviation community as a whole now knows -- that the DC10 has a much smaller share of the blame than American Airlines for the nation's worst crash.

In fact, it is time to stop picking on the DC10.

That three-engined, wide-bodied jetliner has recaptured the high level of professional respect Douglas airplanes have traditionally held. The new attitude results from discoveries in both the official investigation of the Chicago crash and the later, largely unpublicized civil liability suits brought by the relatives of those killed in the crash.

However, answers to some of the questions about the Chicago crash will never be known because McDonnell Douglas and American Airlines reached agreement in private on what percentage of damages each would pay in the dozens of legal suits that followed the crash; McDonnell Douglas and American never contested that they were the liable parties. Their agreement is sealed forever by court order, but in the weeks and months before it was struck last April, 23 months after the crash, these curious things happened:

* American Airlines officials ordered an internal investigation of the accident, then destroyed all copies of the report of that investigation on the advice of their own counsel and in violation of a court order, according to court findings.

* Documents obtained during legal discovery make it posssible to infer -- but not prove -- that some American Airlines maintenance personnel knew they had cracked the engine mounting on the crashed airplane, but, thinking the crack unimportant, decided to postpone fixing it and returned the plane to service. American Airlines officials deny the inference, claiming that a key document pointing to that inference was misdated. There is no indication senior American officials were aware of the problem.

* The crew chief who worked on N110AA, the crashed plane, when it received major maintenance between March 29 and 31, 1979, committed suicide at his home near American's maintenance base in Tulsa. He had been scheduled to meet with attorneys the next day to discuss a deposition he was to give about maintenance on the plane.

* Business Insurance, a publication that follows the liability insurance field, reported that the sealed court agreement shows that American Airlines' insurers agreed to pay 75 percent and McDonnell Douglas 25 percent of the damages awarded families of the victims of the crash. Neither American nor McDonnell Douglas will confirm or deny that report, but aviation sources say it is accurate.

Business Insurance also reported that Douglas had agreed to pay 85 percent of the damages in a tentative arrangement worked out shortly after the accident, before investigation and legal discovery began to point the finger the other way. In other words, there was a remarkable turnaround in the perceptions of who was to blame.

Presiding over this entire tale is the ghost of an earlier DC10 disaster, a crash near Paris in March 1974 which killed 346 people and remains the worst single-plane accident in history. Both the FAA, the regulator, and the National Transportation Safety Board, the investigator, had taken hard knocks in the press and Congress for not having prevented the Paris crash, and neither body was of a mind to appear it was not doing its job after Chicago.

The Paris air crash occurred when a door to a Turkish Airlines DC10 cargo compartment blew open explosively as the plane climbed to 12,000 feet. The floor of the passenger cabin buckled, severing control cables and sending the plane into a fatal dive.

The possibility of explosive decompression occurring because of a malfunctioning lock on the cargo door had been discovered in an earlier, nonfatal DC10 accident near Detroit. The question was: Why hadn't the FAA ordered McDonnell Douglas to make sure the door was changed so it wouldn't happen again? As a result of the Paris crash, the floors of all wide-body aircraft have been strengthened so they can withstand explosive decompression.

Five years later, American Airlines Flight 191, nonstop from Chicago's O'Hare International Airport to Los Angeles, took off. Just as the nose lifted off the runway, one engine and the pylon that held it under the left wing ripped up and over the wing and crashed to the ground. Electrical power was cut to cockpit warning instruments; hydraulic lines that powered the control surfaces on the left wing were slashed.

The surfaces, called slats, extend from the front of both wings during takeoff to give a jetliner added lift. With the hydraulic lines cut, the left wing slats retracted while the right wing slats remained extended. That unbalanced the plane's controls, a circumstance called "asymmetrical slats," and resulted in the plane having more lift on the right wing than on the left. The plane climbed briefly, rolled to the left and crashed into a field just north of O'Hare.

It took only two days for engineers to discover that the aft pylon bulkhead, a major support plate in the pylon, had catastrophically sheared in two. All DC10s were grounded for a quick check of the pylon, then permitted to fly again. Another check, and two American Airlines planes that had been checked the first time were found to have cracks in the aft pylon bulkhead.

FAA Administrator Langhorne M. Bond felt he had no choice (and Paris presumably was in the back of his mind). He took the unprecedented action of grounding the entire DC10 fleet for 37 days while his experts pored over the engineering data that, Douglas insisted, proved the pylon to be a sturdy, carefully designed structure.

A case was also beginning to build against American. Within 10 days of the accident, safety board investigators made a major discovery: American was not using the Douglas- recommended maintenance procedure to remove the engine and pylon from the plane while replacing a worn part.

Douglas recommended the engine be removed separately from the pylon, a process that involves unscrewing many bolts, hoses and connections. It saves a lot of time (and money) to remove both the engine and the pylon as one large unit, and that was a procedure American (and Continental) began to use. The problem was that the single-unit procedure placed enormous stress on the aft pylon bulkhead after it was disconnected from the wing because the weight of the engine, about 5 tons, smacked the pylon against metal in the wing. As mechanics moved the assembly, it was possible to crack the pylon bulkhead, safety board tests proved.

The FAA moved quickly to outlaw the single-unit procedure and punish those who had used it. American Airlines paid a $500,000 civil penalty "under protest" to the FAA in November 1979 to settle all claims concerning the maintenance procedure; Continental paid a $100,000 civil penalty.

The next month, the safety board formally closed its investigation of the Chicago crash by ruling that "The probable cause . . . was the asymmetrical stall and the ensuing roll of the aircraft because of the uncommanded retraction of the left wing outboard leading edge slats and the loss of stall warning and slat disagreement indication systems resulting from maintenance-induced damage leading to the separation of the No. 1 engine and pylon assembly at a critical point during takeoff. The separation resulted from damage by she improper maintenance procedures which led to failure of the pylon structure."

The key words in that paragraph are "maintenance-induced damage." American Airlines did the maintenance, but it was McDonnell Douglas that was running an extraordinary television advertising campaign with former astronaut Pete Conrad telling us how much he recommended the DC10.

The board went on in a second paragraph of its "probable cause" section to give some blame to McDonnell Douglas for the vulnerability to damage of the pylon attach points and of the slat system. The FAA was blamed for deficiencies in its surveillance and reporting system, which failed to disseminate widely the fact that in December 1978 and again in February 1979, pylon bulkhads on Continental DC10s were cracked during maintenance when mechanics used the same one-step shortcut later blamed in the American case. Continental caught the problem and fixed the pylons before putting the planes back in service.

While the safety board was ending its investigation, lawyers for the families of those killed in the crash were just getting started. Safety board findings, under federal law, cannot be used as the basis for legal claims, although the technical work of the board's staff and the facts it uncovers help lawyers build their discovery lists.

It appeared initially that the legal task would be relatively simple and that American Airlines and McDonnell Douglas would work out how to divide responsibility for paying the settlements. That proved to be a futile hope after a few months, however, and federal district judges Edwin Robson and Hubert Will permitted discovery to begin in April 1980 in anticipation of a full-scale trial on the question of division of responsibility. That's when it began to get interesting.

In due course, attorneys discovered that Donald Lloyd-Jones, American's senior vice president for operations, had ordered American's senior safety expert, Mac Eastburn, to do an internal investigation of the accident. Such a report, of course, would be of extraordinary interest to lawyers seeking to establish liability so, naturally, they asked the court to order American to produce it. American attorneys, according to court papers, represented on several occasions that an "Eastburn report" did not exist.

Robson and Will, in a subsequent opinion, tell the story best:

"Lloyd-Jones testified that Eastburn informed him in late August 1979 that he had been instructed by counsel not to retain copies of notes or drafts of the report. He testified that some time in September 1979 he confirmed with (Richard) Malahowski, in- house counsel for American, that Malahowski had given such instructions to Eastburn. Some months later, he stated, Eastburn informed him that he had not retained any notes or drafts . . . Lloyd-Jones stated that at the time he discarded his copy of the report, he was not aware that it was the sole remaining copy, but that Malahowski informed him that no copies remained some time in the summer, 1980."

American then contended that the report was not subject to an Illinois state court order issued a few days after the crash and requiring that all evidence concerning the accident be preserved.

"This contention," Robson and Will wrote, "is untenable. A party may not destroy documents where a preservation order has been entered, conceal that destruction for almost one year, then claim that the preservation order never applied . . . moreover it is inconceivable that the Eastburn report did not deal with matters relevant to the issue of liability and the cause of the May 25 accident."

Robson and Will ordered American to pay all costs and fees relating to depositions, court appearances or motions dealing with the Eastburn report "or which might have been unnecessary had the Eastburn report not been destroyed." It could not be learned what that cost American, although spokesman Art Jackson said it was "Not much." Donald W. Madole, a Washington attorney who specializes in aviation litigation and who filed the motion seeking the legal fees on behalf of the families, said the matter "has been settled to my satisfaction."

The Eastburn report, Jackson said in an interview, was incorporated into an "action plan" for the DC10 that American Airlines prepared.

In his motion, Madole argued that the Eastburn report must have contained information showing that American employes knew the pylon was cracked. Madole cited documents obtained during discovery showing that American Airlines had sent out a "fleet campaign direction" to its maintenance bases directing full-scale inspections of the DC10 pylon area. The specific instructions for that inspection were written in Tulsa, and the first signature on those instructions is dated May 25, 1979.

What makes that interesting, Madole's motion argued, is that investigators did not have access to the accident site until May 26, the day after the crash. Therefore, he suggested, there was no way to know that the pylon was a suspect area and no basis for ordering a fleet campaign directive to inspect the pylon area on all other American DC10s unless American employes already knew there was a crack in the pylon.

American denies that inference. Spokesman Jackson said that the person who first signed the fleet campaign directive "mis- dated it. All the other signatures show the proper date." The fleet campaign directive, he said, could not have been prepared "until he got information from the crash site."

Judges Will and Robson said they did not have to deal with Madole's contention that some American employes knew the pylon was cracked because the motion was filed after the division of responsibility for paying damages had been settled. "This is not a trial on liability," the judges wrote, "and the court is not required to find whether American knew of the crack."

Neither is the FAA because, remember, it settled all issues pertaining to the maintenance procedures on the DC10 when American paid the $500,000 civil penalty in November 1979.

A total of 165 different suits arising from the Chicago crash were consolidated in the federal court in Chicago until the division of responsibility was decided and sealed. About 80 families have been compensated either through out-of-court settlements or as the result of trials. Settlements range from as little as $75,000 to as much as $1.8 million. It is all very depressing: a young professional with a brilliant future and three children to educate is worth considerably more than a senior citizen on a lark. Many of the cases have been remanded to other federal or state courts for trials on damages, but about half the families have still received no compensation.

That leaves one other matter. Last March 24, Earl Russell Marshall, 47, a supervisor at the American Airlines maintenance base in Tulsa, left his house near suburban Broken Arrow around 10 p.m. to go to work on the graveyard shift. He returned home almost immediately, according to Wagoner County Sheriff Clyde Watkins, and told his wife he just didn't want to work that night. She persuaded him to go, then went to bed.

Marshall never got to work. He returned home again that evening and backed his car up to a window of the laundry room of the ranch-style home. He attached a hose to the muffler of the car and ran it through the laundry room window and started the car. Then he stuffed clothing around the edge of the door and around the hose coming through the window and "sealed it real good," Watkins said.

Marshall lay down on the floor, where his wife discovered his ody the next morning. His lunch pail was still on the seat of the car. Death was ruled a suicide, although Marshall left no note.

Marshall's wife told Sheriff Watkins that he had been despondent over the air crash because he had been the crew chief when the engine-pylon was removed and repaired on N110AA. "She said he felt he was partly responsible, althoughhad discussed it with him many times and tried to convince him he s in aviatiohad nothing to do with it." Marshall was scheduled to meet at the end of his shift with attorneys.

Several things have been done since the Chicago crash to change all DC10s or how they are flown. New instructions have been given to pilots on how to "fly out" of situations similar to the one the Chicago crew encountered because, it has been learned, it is possible to survive an asymmetric-slat, engine-out condition on takeoff if normal procedures are modified. More redundancy in electrically powered cockpit warning systems has been required. Some minor changes in a couple of bolts in the pylon have been devised to make the aft pylon bulkhead less susceptible to damage.

There have been two major DC10 accidents since the Chicago crash, but neither involved anything so bizarre as a cargo door blowing out or an engine-pylon combination falling off the plane on takeoff. A Western Airlines DC10 crashed on landing in Mexico City, in October 1979, killing 72 people. The incident occurred when the plane landed on a runway parallel to the one it was supposed to use. The wrong runway was undergoing reconstruction at the time.

A New Zealand Airlines DC10 full of sightseers crashed into the side of a mountain in Antarctica in November 1979 and killed all 257 aboard. The board of inquiry placed considerable blame on New Zealand officials for tinkering with navigational aids and instructions to the crew.

There is no indication of mechanical problems with the DC10 that slipped off the end of an icy runway at Logan International Airport in Boston Jan. 22, apparently drowning two people.

There is, however, one other interesting incident, and it led directly to that meeting on the third Sunday in January. The right wing engine of an Air Florida DC10 blew up as the plane was accelerating along the runway on a takeoff from Miami Sept. 22, 1981.

Pieces of the disintegrating engine severed one of the cables that, in combination with the hydraulic system, keeps the leading-edge slats extended. The slats on one wing retracted, just as they had in Chicago, but it made no difference because the pilots knew about the engine problem soon enough to safely abort the takeoff.

So now, almost three years after the accident, the FAA and Douglas were sitting down to agree on a "fix" to make sure (as American Airlines had requested after Chicago) the slats would not retract unexpectedly. Douglas had previously insisted that, since it was possible to continue safe flight with asymmetrical slats, it was not necessary to prevent the condition, a position supported by federal regulations. Thus, the FAA could not order the change because the Chicago crash had been found to be caused by a maintenance- induced crack, not by an inherently dangerous design.

The Air Florida incident, combined with an earlier, similar occurrence in Pakistan, suggested to both FAA and Douglas that a slat modification might be prudent if not required. That modification is now in the pipeline and will be completed on all DC10s by Jan. 31. The time has come to stop picking on the DC10.