Space agency officials yesterday told the presidential commission investigating the explosion of the space shuttle Challenger that they have long been aware that seals on shuttle booster rockets have been damaged during some previous flights but insisted that the problem was never thought to pose an imminent safety hazard.

Nonetheless, National Aeronautics and Space Administration officials acknowledged that they had carried out "a more and more intensive analysis" of the problem last summer and fall and had developed a possible corrective mechanism that was to have had its first ground test this Thursday.

The seals are designed to prevent white-hot exhaust gases from burning through the walls of the solid rocket boosters that help launch the shuttle. Such a burn-through in the right-hand booster is being studied as a possible cause of the disaster on Jan. 28.

In a related matter the commission heard more testimony about the effect of cold air temperatures at the launch site, a factor that Lawrence B. Mulloy, project manager for the boosters at the Marshall Space Flight Center in Huntsville, Ala., acknowledged can make the seals -- rubber O-rings -- stiffen and, at least theoretically, fail to seat properly in the gap they are supposed to seal.

Mulloy told the commission that the day before Challenger launched, representatives of Morton Thiokol, the company that makes the boosters, had suggested that the shuttle not be flown if the temperature was colder than on any previous shuttle launch. It was 38 degrees when Challenger lifted off, colder than any previous shuttle launch. Earlier in the morning the temperature was 27 degrees.

Mulloy said that, despite the suggestion, it was eventually agreed to proceed with the flight.

At Cape Canaveral yesterday, NASA officials said that in spite of the cold, the Challenger's solid rocket boosters were not examined by any NASA engineer in the 38 days it spent on the launch pad before liftoff. They said such physical inspections are considered unnecessary.

At the presidential commission hearing in Washington, officials said that despite damage, the rubber O-rings fitted between the stacked segments of the booster had always held, preventing burn-throughs.

It is known that Challenger's right-hand booster sustained a burn-through at or near the joints between the two lower segments but the officials refused to say whether this was a cause or an effect of whatever went wrong.

Mulloy said that of 228 joints examined after earlier shuttle flights, 22 showed damage to the primary O-ring and one showed damage to the back-up secondary O-ring.

The hearing was called by commission chairman William P. Rogers in response to a New York Times article Sunday that quoted an internal NASA memorandum from last July that called the seal damage "a potentially major problem affecting both flight safety and program costs" and warning that "failure during launch would certainly be catastrophic."

The memo was written by Richard C. Cook, a NASA budget office employe who had been asked to analyze the seal problem and estimate the cost of fixing it.

After yesterday's hearing NASA released copies of the July memo along with another that Cook submitted six days after the disaster. In that memo Cook said it was the consensus of NASA engineers examining the Challenger explosion that "if such a burn-through occurred, it was probably preventable and that for well over a year the solid rocket boosters have been flying in an unsafe condition."

Cook's boss, Michael Mann, played down Cook's authority, saying he was a new employe still in training and that "the memo overstated the concerns."

While Mulloy discussed the effects of cold on the O-rings, commission member Richard P. Feynman, a Nobel laureate in physics, carried out an impromptu experiment, dunking a sample piece of O-ring in his ice water.

Feynman reported his findings immediately, saying that at 32 degrees -- the temperature of ice water -- the rubber had hardened considerably, losing much of the resilience needed to conform to the shape of the gap it is supposed to seal.

"I believe that has some significance for our problem," Feynman said.

Mulloy said that although the O-rings are supposed to work at temperatures as low as 30 below zero, the data "is refuted by some other test data." Mulloy said NASA was running new tests, simulating the conditions that prevailed during the Challenger launch.

During lunch Feynman repeated his experiment for reporters and said the chilled rubber "doesn't work very well. It's obvious."

Yet another scenario for what might have happened emerged yesterday when Mulloy raised the possibility of a failure involving the "leak test port," a small hole in the booster wall that is supposed to be plugged with a threaded steel bolt.

At every joint in the booster there is a hole from the exterior to the space between the two O-rings. When the booster is being assembled, technicians are supposed to check the reliability of the O-rings by opening the port and pumping in high-pressure gas. If the O-rings work, they prevent any gas from escaping into the rocket. After the test, the hole is closed with a tight-fitting bolt.

If the bolt is left out or works loose, Mulloy said, the hole would provide an outlet for hot gases that get past the primary O-ring. The test port is situated near the spot on Challenger's booster from which the mysterious plume emerged.