NASA yesterday began assembling two teams of investigators to study the test failure last week of a nozzle component in the space shuttle booster as speculation about the cause focused on an extreme maneuver of the nozzle that was made 110 seconds into the two-minute test firing.

Discovery of the failure forced officials of the National Aeronautics and Space Administration to announce Tuesday the postponement of the next shuttle flight, which had been scheduled for June 2.

NASA officials said they won't be able to determine the length of the delay until they can take apart the booster and begin to analyze the problem, a process expected to take several more days. Estimates by agency and industry officials of the potential delay ranged from a few weeks to several months.

The complete failure analysis is likely to take "three to four weeks," according to John Thirkill, vice president for space operations at booster manufacturer Morton Thiokol Inc.'s Wasatch, Utah, plant.

Shuttle launch director Bob Sieck said at a briefing in Florida that Kennedy Space Center workers will be ready to launch the shuttle Discovery five months after the first booster nozzle segments are delivered. Shipment of the hardware had been scheduled for this weekend until it was postponed as a result of the test failure.

Engineers who crawled through the burnt-out test booster with flashlights late Monday found that a piece 3 1/2 feet long and 9 inches wide was missing from the ring that anchors the booster's nozzle to a flexible rubbery "boot" that allows the nozzle to swivel. The part is called the "outer boot ring."

Reviewing films of the test firing, which took place Dec. 23 in Utah, engineers thought they could see "a few little pieces coming out" in the last 10 seconds of the burn, according to Arnold Aldrich, head of the shuttle program, in a telephone interview from Houston.

This was about the point in the test that technicians swiveled the nozzle to an extreme position that would only be used in an emergency in actual flight, according to Russell Bardos, a propulsion official at NASA headquarters. "Some officials believe we've never tested to that level before," he said.

A two-foot piece of a part bonded to the boot ring, called the cowl, was found up a hillside in the flame trench dug by the blast of the test firing, engineers said. Some missing parts of the boot ring were found inside the booster. Other pieces are still missing, Aldrich said.

In a normal shuttle launch, the booster nozzles are swiveled only a few degrees to guide the vehicle. In the last test -- the second of four full-scale booster tests scheduled before the next shuttle flight -- extreme seven-degree "slewing," as it is called, was required as a result of more stringent post-Challenger safety standards in order to check changes in the nozzle design.

The pieces missing from the outer boot ring had been positioned from about 2 o'clock to 6 o'clock, looking at the cylindrical booster head-on, and that same area, at around 3 o'clock, "is the direction where it would have been slewed," Bardos said.

However, he emphasized, "whether or not that is a factor is yet to be determined. It's something significant we're looking at."

Members of the NASA teams, headed by officials of Marshall Space Flight Center in Huntsville, Ala., which manages the booster program, began arriving yesterday at Morton Thiokol's Brigham City, Utah, booster plant, where the hardware is built and assembled.

One team will investigate the failure of the booster nozzle. And the other will assess its impact on the shuttle schedule. By Tuesday, it is hoped, engineers will have the nozzle off and a better idea of what went wrong, officials said.

Since the shuttle Challenger exploded 23 months ago, killing its crew of seven, a NASA-Thiokol team has redesigned many parts of the solid fuel rocket booster. A new design for the field joint that caused the disaster appears to have performed well in tests, officials said.

The eight-foot diameter outer boot ring was one of the parts that was redesigned, as a precaution, to correct deficiencies. The Dec. 23 test was the first full-scale assessment of all the changes to be incorporated in the next shuttle flight. It was also the first test of the new boot ring design.

The next shuttle flight would have been delayed, even without the test failure, by the explosive fire Tuesday that destroyed a building at the Morton Thiokol plant where fuel is poured into casings for MX missiles. Five workers were killed.

The fueling of MX segments is the same as that for the shuttle boosters, Bardos said. The fire started as workers were removing a metal molding, called a mandrel, from the hardened fuel inside the rocket casing, leaving a patterned "doughnut hole" through the fuel's center, which determines its burn rate.

After the fire, NASA officials ordered Thiokol not to "pull the mandrels" for shuttle boosters until the cause of the tragedy is determined.

Some of those involved in the booster program said the delay is devastating. But NASA officials sought to put an optimistic face on the setbacks, saying that uncovering problems is what the test program is supposed to do.

"The team is looking at this as a normal event that comes out of the testing process," Aldrich said.

He and others stressed that the failed part had "no effect" on the proper performance of the booster in the test, coming as it apparently did late in the burn sequence. And he added, "There is great potential for other things that will cause us some delay as we move forward."

White House spokesman Marlin Fitzwater, at a briefing in Palm Springs, Calif., expressed disappointment at the test failure and the resulting flight delay but added that "safety was the prime factor . . . . We will continue in a methodical way to correct that program and continue the space effort."