Engineers have completed the replacement of a nitrogen regulator in the shuttle Endeavour's left-side rocket pod and topped off the ship's on-board supplies of liquid oxygen and hydrogen, clearing the way for another launch attempt Wednesday afternoon (NASA will release the exact launch time later today).

The weather remains a major question mark, however, with forecasters now predicting a 60 percent chance of thick clouds and local storms that could delay liftoff yet again. The outlook improves to 60 percent "go" on Thursday and falls back to 60 percent "no go" on Friday.

NASA test director Steve Altemus said today the agency's strategy will be to make three launch attempts over the next four days. As it now stands, NASA has booked June 5 and 6 with the Air Force Eastern Range, the agency that provides required radar and tracking support for all Florida launchings. June 7 and 8 currently are reserved for tests of an Atlas rocket at the nearby Cape Canaveral Air Force Station. But Altemus said NASA managers are optimistic they can work with the Range and Atlas-builder Lockheed Martin to adjust the schedule to support additional shuttle launch attempts if necessary.

NASA has been hoping for a break in the weather since last Thursday when Endeaovur's first launch try was scrubbed due to approaching storms. During the initial countdown, engineers noticed problems with a valve in a regulator that supplies the pressurized nitrogen gas needed to operate propellant valves in the shuttle's left-side orbital maneuvering system rocket pod. Engineers ultimately resolved the problem only to see launch scrubbed by the weather.

Over the weekend, during subsequent testing, engineers noticed more unusual behavior in the suspect regulator. NASA managers then ordered its replacement, triggering what turned into another two days of delay.

"We had to take off the heat shield and go up by the OMS nozzle and actually reach in there and remove three bolts to get that regulator out," Altemus said. "The regulator is about nine inches long, maybe two or three inches in diameter. The test setup to get the nitrogen flow and leak checks done, that was the most challenging part of the whole operation. But we got through those successfully and we're pleased with the results.

"I can't say enough about the team at the launch pad," he added. "They're heroes. Those folks who work out at the launch pad do amazing things with that vehicle, they're ace mechanics, they know that vehicle inside and out and when we get in a tough spot, they always seem to pull us out of it."

The regulator supplies nitrogen gas to an accumulator, or tank, that is "charged" with pressurized gas for launch. The gas in the accumulator is routed to a pair of propellant valves that allow the OMS engine to fire. Altemus said the accumulator was fully charged last Thursday and even if the regulator had failed during ascent, Endeavour's left-side OMS engine would have worked normally to help put the shuttle in orbit. Had problems developed after that, the astronauts would have been able to manage the on-board nitrogen supply to complete a normal mission.

"We felt confident we could manage it on Thursday," he said. "In the turnaround, when we went to do some regulator flow checks on Friday, we saw a different signature. So not understanding, then, how that regulator would behave during the next terminal countdown, and not sure we would be able to get a sufficient accumulator charge to activate those valves during ascent, we opted to change out the hardware.

"If we had that failure during ascent, we would essentially lose our GN2 (gaseous nitrogen) pressure to activate those bi-propellant valves and we would have a less than nominal OMS assist to orbit. (But) we could have managed with a single OMS assist to orbit, so really there was no safety of flight kind of consideration in this failure."