The Federal Aviation Administration is collecting inforamtion to determine if a change in emergency flight procedures should be required for all jet airliners, not just the McDonnell Douglas DC10.

The study results from lessons learned in the investigation of the DC10 crash here May 25 that killed 273 people.

Don Armstrong, chief test pilot for the FAA's Los Angeles office, described the study for examiners as a National Transportation Safety Board hearing into the accident.

Also today, the chief of quality assurance for McDonnell Douglass' manufacturing operation, John W. Stillwell, testified that 23 engine support pylons on 22 DC10s were manufactured with shims not called for in design drawings.

American Airlines has contended that one such shim, a thin strip of metal used to fill a gap, contributed to the length of the crack in the aft bulkhead of the pylon on the crashed airplane, and thus was a factor in the crash. The aft bulkhead is believed to be the first part that broke in the crash sequence.

Stillwell said the shims were used on 23 pylons installed on the 15th through the 36th DC10s manufactured. The crashed plane was No. 22.

No other cracked pylon bulkhead -- eight others were found in inspections following the accident -- contained shims, Stillwell said.

The shims were used to fill gaps caused by "a tooling problem," which was corrected, he said.

Stillwell said that McDonnell Douglas' pylon assembly crew has been completely retrained as a result of other difficulties discovered during post-crash inspections. Specifically, 36 pylons were found with cracked or sheared bolts in the upper spar web. The bolt problem was later attributed to a lack of quality control. Stillwell said his investigation of that problem is still underway.

The emergency procedure the FAA has under study is what piloting technique to use if an engine fails on takeoff, as happnned in the crash here. A pilot must instantly choose whether the power available in his remaining engines should be used to climb higher or to go faster.

In the Chicago accident, investigators and test pilots have postulated, American Airlines pilot Walter Lux might have been able to save his plane had he lowered the nose of the DC10 and concentrated on picking up speed instead of altitude. In the wake of the crash, American Airlines and several other U.S. lines have changed their DC10 emergency procedures to stress speed, not altitude, during troubled takeoffs.

An obvious problem with this option, however, is that at some airports, planes might be flying too low to clear obstacles at the ends of runways.

The advantage of extra speed is that it precludes the possibility of an airborne stall -- the loss of sufficient speed to remain aloft. That is what happened to Captain Lux's DC10 because of damage to the flight controls that resulted when the engine and its support pylon fell off the left wing. The loss of those flight controls would not have mattered had the plane accelerated, according to pilots and test results.

In an interview today, Armstrong said that "if there is merit to a procedural change for one aircraft, we wonder if it shouldn't be valid for another. Right now engineers are collecting the data we need to find out."