Clouds hang over Kennedy Space Center as a cold front moves through the Central Florida region. Photo: NASA TV/Spaceflight Now

NASA managers today cleared the repaired shuttle Discovery for a delayed launch attempt Monday evening to kick off the 100th shuttle flight, the most complex space station assembly mission yet attempted.

While forecasters are predicting a 70 percent chance of high crosswinds at the shuttle's 8:05:17 p.m. launch time that could prompt another delay, NASA managers have decided the shuttle's external tank attachment bolts are healthy and good to go as is.

Discovery's planned launch last Thursday was delayed just before fueling when engineers reviewing film from a flight last month noticed one of the two external tank aft attach bolts protruding about two-and-a-quarter inches after the tank was jettisoned.

It should have retracted all the way into a protective housing when the explosive nut on the bolt detonated to begin the tank separation sequence.

An analysis of film from previous shuttle flights has turned up several instances where an aft attach bolt has failed to fully retract into its housing.

The concern is that the shuttle could impart a load, or force, on an extended bolt as separation begins, causing the tank to pitch, roll or yaw in such a fashion that it could hit and damage the orbiter.

"They have seen a number of times that this has occurred in the past, on the order of a half-dozen or so," NASA test director Steve Altemus told reporters today. "There's never been any (damage) associated with that at all."

After studying the issue over the weekend, engineers concluded the bolt issue poses no threat to flight safety. The only case in which an extended bolt could be a factor, they concluded, was during a return-to-launch-site abort or a trans-Atlantic abort to Spain or Africa.

During such an abort, the shuttle would be positioned atop the tank during the separation sequence.

If the shuttle's main flight computers failed, forcing the crew to rely on a backup flight computer, the tank separation sequence would occur a bit slower than usual, allowing more time for recontact to occur.

But such a series of failures is unlikely and Discovery was cleared for launch as is.

The photo from a camera on Atlantis that showed the bolt sticking out. Photo: NASA

NASA initially ordered a minimum 24-hour launch delay for Discovery to assess the bolt issue.

But the flight was delayed again Thursday night, this time to 8:05:17 p.m. Monday, after NASA managers ordered engineers to replace a sluggish valve in the shuttle's engine compartment that helps minimize vibrations during ascent due to propellant surges.

Engineers noticed the problem while they were recycling the countdown for a possible Friday launch.

Had NASA pressed ahead with launch Thursday, the sluggish valve would have prompted on-board computers to stop the countdown just before main engine ignition.

In any case, the valve has now been replaced and Altemus said Discovery is once again ready for launch on the 100th shuttle mission.

"While the program managers continue to evaluate and come to closure on the ET separation bolt issue, the launch team continues to prepare the vehicle and all systems for launch on Monday evening," he said.

"The bird is in great shape We're not working any technical issues whatsoever and the ground systems are performing as well."

But the weather remains a wild card.

John Weems, shuttle weather officer with the 45th Weather Squadron at the Cape Canaveral Air Force Station, said high winds in the wake of a passing cold front will cause problems.

"The main concern will be the winds behind this front," he said. "They will be from about 350 to 020 degrees. The farther east these winds become, the more likely they'll violate the RTLS crosswind constraint of 15 knots."

During the first few minutes of flight, a main engine failure could force a crew to attempt a risky return to launch site abort, known as an RTLS, to the Kennedy Space Center.

A file image of a night landing at Kennedy Space Center. Photo: NASA-KSC

Flight rules forbid launching a shuttle if the predicted crosswinds at the shuttle runway are higher than 15 knots in daylight or at night. A 12-knot landing limit is used for normal night landings at the end of a mission.

The shuttle runway runs from northwest to southeast, i.e., 330 degrees to 150 degrees. Winds from 60 degrees or 240 degrees are pure crosswind. For other angles, the crosswind component is less.

For Discovery's launching, forecasters are predicting winds from 350 degrees to 20 degrees, resulting in a crosswind right at NASA's limit.

"To give you an idea of how critical (wind direction) can be, if the winds are from 360 degrees, you would have to have about 29 knots peak to violate that 15-knot crosswind component," Weems said.

"If they swing around to around 010 degrees, that peak speed limit drops to about 24 knots and the further east it goes, the less wind it takes to violate that crosswind components."

Weems said computer models tend to err on the side of predicting winds from a more northeasterly direction than what is actually observed.

"If we're incorrect even by 10 degrees in that quadrant, it really buys us a significant amount of wind," he said.

In the meantime, the official forecast calls for a 70 percent chance of high enough winds to cause a delay. The outlook is the same for Tuesday night, but conditions are expected to improve dramtically by Wednesday, resulting in a 70 percent "go" forecast.

NASA must get Discovery off the ground by Wednesday or launch will be delayed to at least Oct. 14 because of an already scheduled Atlas rocket launch from the Cape Canaveral Air Force Station.

NASA normally makes two back-to-back attempts to launch a shuttle before standing down 24 hours to give the launch team a chance to rest. But Altemus said NASA managers could waive that rule and make three attempts in a row next week if necessary.