American Airlines charges that a manufacturing technique may have played a role in the crash that killed 273 persons here May 25 gained credence today in testimony from DC10 builder McDonnell Douglas.

American Airlines Senior Vice President Donald Lloyd-Jones first leveled the charge in an interview with The Washington Post two weeks ago. Its substance is that McDonnell Douglas placed shims - thin metal strips used to fill gaps between metal pieces - at an unusual place in the engine support pylon of the crashed airplane.

Those shims, Lloyd-Jones said, provided added leverage and thus lengthened the pylon's aft bulkhead crack, which is cited as the source of the accident.

In the second day of the National Transportation Safety Board's hearing into the accident, Dale S. Warren, chief design engineer for structural analysis at McDonnell Douglas, said almost the same thing.

In tests, Warren said, McDonnell Douglas has been able to reproduce cracks similar in length to the one on the crashed airplane on aft pylon bulkheads only with shims installed.

Everybody seated at the American Airlines table smiled as Warren spoke. American and McDonnell Douglas have been busy trying to blame each other almost since the accident.

The crack was induced, investigators have said, by a maintenance procedure used by American Airlines mechanics. In the crashed plane, that induced crack was 10 inches long and grew, under the stress of 430 flight hours, to about 13.5 inches.

Then, the aft bulkhead in the pylon broke, the other two main pylon supports also broke, and the pylon and the engine it held off the left wing. That happened just as Flight 191 was lifting off the ground. Vital flight controls and electrical warnings were ripped out with the pylon, and the plane crashed 30 seconds after liftoff at O'Hare International Airport here.

Warren said that test results on pylon strength will be available Thursday. In response to questions from safety board investigators, however, he said that the only cracks of 10 inches or more McDonnell Douglas has been able to produce were on pylons containing the shims. The longest crack produced in testing on a pylon without a shim is 7 to 9 inches, he said.

Warren took the witness stand after technical experts spent most of the day interviewing mechanics from American Airlines' maintenance base in Tulsa. The crashed airplane received a regular maintenance check there in March, and at the time the pylon was removed and serviced. The crack is believed to have been caused at that time.

None of the five American employes who testified remembered seeing a crack or hearing metal break as they worked on the pylon. However, in testimony and in documents released during the hearing, it was learned that:

The pylon was lowered from the wing so a part could be replaced. That part is a complicated bolt, bushing and bearing assembly that attaches the aft bulkhead to the wing. It was beginning to wear out, so McDonnell Douglas had recommended replacement on all DC10s with a sturdier version.

None of the American Airlines employes who testified had received anything other than on-the-job training in a new technique American Airlines was using to remove and reinstall the pylon. That technique involved lowering the engine and pylon as one unit, instead of separating them, as McDonnell Douglas recommended.

When one work crew arrived after a shift change, the aft pylon attachment had already been removed, but the pylon remained attached at two other points. The pylon and engine were supported by an unmanned forklift, and the wing fitting was resting against the aft pylon bulkhead. With the aft attachment removed, the front attachment would act as a pivot point for the weight of the engine if the forklift were just a little off center.

The forklift then had to be moved, because it was "misaligne," mechanic Raymond Lattanzia testified.