In the moments before the space shuttle Challenger exploded, Mission Control computers recorded a series of problems, but these were not displayed on ground controllers' consoles in time to save the astronauts, an authoritative source at Johnson Space Center said yesterday.

This was among new details emerging about Challenger's 74-second flight last Tuesday in which seven astronauts were killed.

Film released over the weekend by the National Aeronautics and Space Administration revealed that, about 15 seconds before the explosion, the starboard solid-rocket booster ruptured and was spewing white-hot flame onto the large external fuel tank. When the rupture occurred, the booster's sensors immediately detected about a 5 percent drop in internal pressure as expanding exhaust gases escaped through the rupture, said the source, who has had years of experience in the NASA Mission Control room and requested anonymity.

The source in Houston described the shuttle's final seconds this way:

Decreasing pressure kept the solid rocket's fuel from burning at its maximum rate, and the booster lost about 100,000 of its 2.5 million pounds of thrust. Attempting to compensate, the booster's nozzle immediately and automatically swiveled to change the angle of thrust.

The three main liquid-fueled engines were suddenly starved of oxygen, a sure sign that the pipe carrying oxygen from the top of the external fuel tank to the main engines had been penetrated by flame from the booster.

Although the source said all of this information was automatically transmitted to Mission Control, it is not clear whether controllers were aware that anything was amiss.

Such information generally takes three to five seconds to appear on controllers' computer screens and about the same amount of time to be comprehended by eye and mind.

Regardless, the source said, flight controllers could do nothing to prevent the catastrophe because, by the time the information was digested and presented, the explosion occurred.

"I'm not sure I'd have told the crew anything," the source said. "Maybe it's better that they died the way they did, flying supersonic and outward-bound from Earth on their mission."

Challenger was destroyed by an explosion of the external tank's liquid hydrogen. Other sources have said the blast may have been triggered by accidental detonation of small explosives attached to the tank to destroy it when the shuttle enters orbit.

According to one source, the detonation of this mechanism may have caused the bright flash seen on television videotapes just under the shuttle's nose where the explosives are housed.

Outside NASA, speculation about the disaster's cause continues to focus on the starboard booster.

Aerospace engineers at several universities said the rupture may have been caused several ways.

One is a defect in the manufacture or assembly of the 149-foot rocket's half-inch stainless-steel casing intended to contain the enormous pressure of the burning rocket fuel, forcing it to escape downward through the nozzle to provide thrust.

"That casing is there to hold the pressure in and, if there's even the slightest leak, it's going to cause problems," said Herman Krier, an expert in solid-fuel rockets at the University of Illinois. "If there's a place those exhaust gases can leak out, they will, and they're hot enough to melt steel."

Krier said that, even if the leak started as a tiny crack in the casing, the heat would soon melt a larger hole. NASA film of the booster shows the leak starting as a small point and quickly growing.

Krier said the rupture could have started as a crack in either the casing or the block of solid fuel just inside the casing.

Challenger's solid fuel has the consistency of hard rubber and can crack if stressed, said Michael Micci, an aerospace engineer from Pennsylvania State University.

"If there was some kind of vibration during the launch, it could have cracked," exposing a fissure that reached to the steel casing, he said. The fissure's surface would have begun burning and brought heat close enough to the steel casing to melt through it. But, he said, "It's possible there was a gap between the segments of the booster."

Shuttle boosters are assembled by stacking cylindrical segments and fastening them with heavy metal pins. Each joint is wrapped with fiberglass tape and sealed with a special rubber band glued to the casing.

If unusual stresses during launch worked any of these loose, they could have started a leak.