NASA has decided to send the space shuttle Discovery back to Earth for a post-midnight landing Monday as previously scheduled, without ordering an ast

NASA has decided to send the space shuttle Discovery back to Earth for a post-midnight landing Monday as previously scheduled, without ordering an astronaut to repair a mysteriously damaged thermal blanket outside a cabin window.

Officials announced the decision Thursday after scientific analyses -- including an emergency "huge effort" involving a giant wind tunnel test at NASA´s Ames Research Center in Mountain View -- convinced them the return flight would be the safest of three possible options.

Shuttle program deputy manager Wayne Hale told reporters at a press conference at NASA-Johnson Space Center in Houston on Thursday that those options included:

-- Conducting a potentially risky in-flight repair of the blanket.

-- Transferring the seven astronauts to the international space station, where they´d camp out until another shuttle could be launched to rescue them, while the abandoned, unmanned Discovery is allowed to fall to "a fiery death" on re-entry over an ocean.

-- Sending the Discovery back to Earth without any further repairs.

A select team of about 20 experts debated what to do, Hale said, and finally held a vote, unanimously choosing option three, Hale said. "There was very strong and emphatic unanimity," he said.

By Thursday morning, he said, experts had calculated that there was only a 1.5 percent chance that the damaged piece of shuttle thermal blanket could - - as some had previously feared -- fall off and strike a vital part of the shuttle during re-entry.

"I am not here to tell you that we are 100 percent confident there is no risk during re-entry," Hale said. "That would be untrue and foolish.

"But we have assessed this risk to the very best of our engineering knowledge, and we believe it (the danger) is remote."

Not only would it be dangerous to send an astronaut out to try to fix the blanket, "we don´t know exactly what to do" to fix it, Hale admitted. Worse, a repair job might damage the blanket even more, which, he said, "would violate the ´First, do no harm´ principle" often avowed by physicians -- and, more recently, by NASA.

Located under a cabin window toward the shuttle´s nose, the blanket normally protects its part of the spacecraft from overheating because of friction with the Earth´s atmosphere during re-entry at speeds far faster than the speed of sound. No one is sure what damaged the blanket, although some have speculated it was frayed by atmospheric shock waves during launch, Hale said.

A major part of the NASA experts´ safety analysis included the wind tunnel test at Ames. Early Thursday, after initially being stymied by a three- hour mechanical malfunction, Ames investigators exposed a simulated piece of the "thermal blanket" to extremely high winds inside the tunnel.

The artificial windstorm slightly damaged the fiberglass-like quilted blanket, but not enough to support NASA´s worst-case fear: that during Discovery´s re-entry into Earth´s atmosphere, the blanket might fly off and damage part of the spaceship, perhaps catastrophically.

"The general conclusion of the test was (that the Discovery blanket) will hold together, and if it does come apart (during re-entry), it will come apart in little pieces that will not cause any problem," John Allmen, head of Ames´ Return to Flight division, told The Chronicle. The Ames wind tunnel is 11 by 11 feet and can generate winds close to twice the speed of sound.

During Thursday´s press conference in Houston, Hale cited the Ames wind tunnel tests and showed the media pictures of the whitish, fiberglass-like, rectangular thermal blanket before and after it endured the high winds inside the tunnel.

The pictures show that in the wind tunnel test, "some small pieces (of the blanket) have departed, but in bulk, the blanket is still there," Hale told reporters. "Our indications are (that during re-entry), nothing will come off -- and if it does come off, it will be very small, a few threads."

The wind tunnel test was a rush job. On Wednesday, nervous NASA officials -- already frazzled by several days of anxiety over this shuttle flight´s loss of foam insulation during launch, and briefly delighted by Wednesday´s successful removal by a spacewalking astronaut of two potentially hazardous gap fillers protruding from the spaceship´s bottom -- ordered the emergency wind tunnel test on the blanket.

The clock started ticking as engineers from Johnson Space Center in Houston rushed to Kennedy Space Center in Florida and removed a real shuttle thermal blanket from another shuttle. Then they flew to Ames, arriving at 6:40 p.m. Wednesday; the wind tunnel test was expected to last all night.

However, on Earth as on the shuttle, NASA technology sometimes goes awry. A gadget called a rheostat limit switch, which controls the wind speed of the Ames wind tunnel, "failed -- it broke -- so we had to go in and replace" it, Allmen told The Chronicle. The failure delayed the start of the test by exactly 2.7 hours, he said.

At 4 a.m. Thursday, the wind tunnel team did a "calibration run," or a preliminary test of the wind tunnel to ensure that it could meaningfully mimic the extremely intense winds that the Discovery will encounter during re-entry early Monday morning. Then, at 6:30 a.m., they draped the thermal blanket onto a mockup of the shuttle´s side window and switched on the wind tunnel. During three separate tests, the winds roared for a total of four hours.

The test concluded at 10:40 a.m. The result was a somewhat frayed-looking, but otherwise intact, thermal blanket. "This little blanket turned into a huge effort," Hale said. Even so, he cautioned, the conclusions aren´t absolutely certain, being based on "a lot of ifs" and assumptions, as is typical for cutting-edge scientific research.

"I will not tell you (the re-entry plan) is zero risk," he stressed. "That is clearly not the case."