Top officials of the aerospace firm that built the space shuttle Challenger testified yesterday that they told NASA launch managers two hours before liftoff last month that it was unsafe to fly because heavy ice on the pad could damage the spacecraft.

The Rockwell International Corp. witnesses said they strongly recommended against the Jan. 28 launch, but that senior NASA officials disregarded them and proceeded to the flight that ended in the explosion that killed the crew of seven.

This account was disputed by the senior NASA official who presided over the prelaunch discussions. In testimony before the presidential commission investigating the explosion, Arnold D. Aldrich, manager of NASA's space transportation programs, termed the Rockwell advice "a concern, not an objection to launch."

The Rockwell testimony was the first clear assertion since the disaster that senior space agency managers might have received last-minute advice from a major contractor not to launch the shuttle. Engineers at Morton Thiokol Inc., manufacturer of the solid rocket boosters that were supposed to help lift the shuttle to orbit, have testified that they initially recommended against launch because of different concerns about the cold temperatures but that their company's management overrode their objections and approved the liftoff.

Commission Chairman William P. Rogers, who has been visibly exasperated in recent days with NASA testimony, said he was concerned that NASA officials had not told the commission of Rockwell's objections earlier, and he suggested for the first time in the hearings that agency officials might not have been entirely "forthright" with the commission. "I'm really surprised and disappointed," he said.

In other testimony yesterday:

*A NASA engineer said he had been swayed by arguments made the night before launch by Morton Thiokol engineers that the abnormal cold gripping Cape Canaveral could endanger the boosters.

The testimony by Ben Powers, an engineer at NASA's Marshall Spaceflight Center in Huntsville, Ala., was the first acknowledgment that anyone within the National Aeronautics and Space Administration accepted the evaluation of the Morton Thiokol engineers. Higher level agency officials had said the data supporting the evaluation were not conclusive. However, Powers testified that the information "was clear to me."

*A senior NASA executive contradicted a subordinate's earlier claim that the agency could not, at the last minute, change its criteria for approving a launch. The earlier witness had said that one of the reasons he was upset by the Thiokol attempt to block a launch in temperatures below 53 degrees was that such a temperature limit was not among the approved "launch commit criteria."

*Two photographs entered as evidence showed more clearly than other pictures the puff of black smoke from the right booster just after liftoff that may have been the first sign the rocket had malfunctioned, leading to the explosion 73 seconds later. A NASA specialist said he thinks that the smoke is evidence of a leak in a lower joint of the segmented booster, the suspected cause of the possible booster failure.

*A senior NASA temperature expert who led an ice inspection team to the pad rejected a theory that some aerospace experts have advanced that a leak from the shuttle's supercooled external tank could have touched off the disaster.

Most of the session yesterday was taken up with carefully stated -- and markedly conflicting -- testimony from Rockwell and NASA executives who set forth their versions of the events leading to Challenger's final countdown.

Rockwell's lead witness was Rocco Petrone, a former Apollo moon rocket flight director who heads the firm's Space Operations Division, which built the four shuttle orbiters and advises NASA on the complex spacecraft, which cost $1.2 billion each.

Petrone recalled arriving at Rockwell's Downey, Calif., launch support center early Jan. 28 to find that NASA officials at Kennedy Space Center in Cape Canaveral wanted to consult about possible dangers to the orbiter's fragile thermal protection tiles from heavy ice that had accumulated on the steel superstructure of Pad 39B, where Challenger stood.

Icicles up to four feet long hung from the launch tower, and pools a half-inch deep had frozen on various service floors above the ground.

Petrone's engineers concluded after a quick study that ice could be knocked from the tower as the shuttle blasted off with 4.6 million pounds of thrust and could ricochet into the orbiter, possibly with dire results for the tiles.

Petrone said he called Rockwell managers at Kennedy Space Center and told them, "We cannot recommend launch. From what we see, we think the tiles could be in danger. Let's make sure that NASA understands that we at Rockwell feel it is not safe to launch."

Petrone's call went to a Rockwell group headed by Bob Glaysher, vice president of orbiter operations at the cape. Glaysher testified he had been at the space center about an hour, conferring with his assistants and "developed a position Rockwell would take" at an upcoming NASA-chaired meeting on the ice problem.

As the meeting progressed, he recalled, "I was asked Rockwell's position." He said he "reiterated" concerns Rockwell had voiced about dangers of ice. There had never been such heavy ice accumulations at the 24 previous shuttle flights.

While company engineers at Downey had hurriedly computed possible trajectories of icicles breaking away from the launch tower, he said, "We had no data base on which to have a judgment. This was the first time that [such conditions] had occurred. We therefore felt that since we were in an unknown condition, unable through any analytical techniques to predict where the ice would go or the degree of damage should it strike the orbiter. . . .

"I then gave the following recommendation to NASA, in which I said Rockwell could not assure the safety of operations. Rockwell cannot assure it is safe to fly."

Rockwell launch site director Al Martin and orbiter systems vice president Martin Cioffoletti testified they voiced similar, if more muted views at the meeting.

Aldrich said other NASA senior managers had informed him their own calculations indicated possible damage to the orbiter would not be "serious." NASA officials at Johnson Space Center in Houston and at Huntsville, he said, "all recommended we proceed."

Rockwell's Glaysher, said Aldrich, "had been listening to this entire discussion, as best I can reconstruct it, and report, and while he did not disagree with the analysis at JSC and KSC, they [Rockwell] could not give an unqualified go to launch as ice on the pad and complex was a condition which had not previously been experieced, and thus this posed a small, additional, but unquestionable risk. Glaysher did not ask or insist that we not launch. . . . "

Commission Chairman Rogers expressed "personal unhappiness" at not having been told until "two or three days ago" of the Rockwell objection. The commission, he said, "asked everyone to be forthright. . . . I'm a little surprised it wasn't volunteered," he told Aldrich.

Aldrich maintained that he had told the commission in a closed-door session of Rockwell's opposition. Rogers then read from that session's transcript, quoting Aldrich as saying in part that Rockwell "had some concern about the possibility of ice damage to the orbiter although it was a minor concern. . . . "

Jesse W. Moore, who at the time of the launch was associate NASA administrator for space flight, yesterday disputed the testimony of solid rocket booster manager Lawrence Mulloy Wednesday that it was too late by the night before the launch for Thiokol engineers to change the criteria defining acceptable launch conditions in order to rule out flights in weather below 53 degrees Fahrenheit.

"Until it's T-minus-zero [the moment of liftoff], you can raise an objection to it," Moore told the commssion.

And Challenger launch director Gene Thomas recalled several occasions when last-minute changes were decreed in the "launch commit criteria" that Mulloy had invoked in barring the Thiokol opposition. Thomas said changes can be made "as late as the day before" launch.

NASA engineer Powers testified that he was "easily persuaded" by statements from Morton Thiokol engineers that rubbery O-rings could be too stiff to seal booster segments properly, a problem that could allow hot exhaust to burn through the rockets' steel sides, as investigators strongly suspect happened.

Powers' comments run counter to previous testimony by other NASA officials who have maintained that the Thiokol engineers' concerns were based on inconclusive data and, thus, not worthy of being passed up the chain of command to key launch officials.