NASA officials debated their options by cross-country teleconference yesterday as engineers at Morton Thiokol's Utah plant worked to disconnect the aft segment of a shuttle booster that holds vital clues to why one of its components failed in a test firing last week.

The discovery this week that a part of the nozzle assembly had come apart during the test forced postponement of the next shuttle flight, which had been scheduled for June 2.

One option under consideration would make it possible to launch the shuttle Discovery in late summer or early fall, according to a NASA propulsion engineer, but it would mean gambling on an "irreversible" plan and might turn out to delay the program further. In the post-Challenger era, it is the sort of close decision that confronts NASA officials, pressured to launch again as soon as possible, but operating under unprecedented safety standards and public scrutiny.

Under this option, engineers would soon begin cutting up and retrofitting the booster hardware currently being held at the Morton Thiokol plant that was scheduled for shipment to Kennedy Space Center in Florida this weekend until the test failure was discovered. It contains redesigned nozzle parts like the one that failed.

The flight hardware, under this option, would be fitted with nozzle components based on an earlier, "interim" design that was included in the first full-scale test firing of the booster three months ago and worked. The segments could be shipped to Florida possibly as early as mid-February. Space center workers there could be ready to launch five months after delivery, officials said.

However, this plan would require a decision to begin by late next week and it is doubtful that a thorough analysis of the failure will be possible by then, according to Russell Bardos, a propulsion official at NASA headquarters. For that reason, he said, some top NASA officials oppose it.

"Some are against doing anything irreversible," he said. "If we did proceed, it would be with the knowledge that it could be a mistake."

Engineers yesterday were focusing on a failure scenario that involved an unusually severe swiveling of the nozzle during the test but with other possible contributing factors that might suggest "doing something very simple" to solve the problem, he said. "But it's not solid yet."

One possible factor involved in the failure, officials said, is the effect of carbon dioxide, which is used to quench the flames as the test ends and could cause brittleness in some components. This procedure does not occur during an actual flight.

A more fundamental factor is the possibility of a design defect. The part that failed is a ring -- called the outer boot ring -- which anchors the booster's nozzle to a flexible boot that allows the nozzle to swivel.

That ring is among many parts of the booster that were redesigned in the wake of the Jan. 28, 1986 Challenger disaster that killed seven astronauts. It is not related to the field joint that directly caused the tragedy. It was one of the few parts on the booster that had never been tested until the Dec. 23 firing.

Late next week, top NASA managers will gather at Marshall Space Flight Center in Huntsville, Ala., which manages the booster redesign, for a series of meetings where they will review the situation and consider what course to take. By then, engineers expect to have taken the nozzle apart and have at least some idea of what the problem is, according to Morton Thiokol spokesman Rocky Raab.