Baffling problem could prompt shuttle fueling test
BY WILLIAM HARWOOD
STORY WRITTEN FOR CBS NEWS "SPACE PLACE" & USED WITH PERMISSION
Posted: July 17, 2005
Testing and inspections continued into the night Sunday, but initial hopes of finding an easy-to-fix problem faded with each passing hour. NASA's mission management team, which has met every day since Discovery was grounded, will meet again on Monday to review the weekend's progress, the results of ongoing analyses and to consider possible courses of action.
As of this writing, launch remains targeted for no earlier than late this week. Whether or not that target remains even theoretically possible past Sunday, NASA managers have not given up on getting Discovery off the ground before its current launch window closes July 31, assuming the ongoing tests eventually isolate the problem that grounded Discovery in the first place.
Among a variety of options now on the table, sources said, is another fueling test, either a stand-alone exercise or a tanking test conducted as part of an actual launch countdown, to further isolate the problem. But as of this writing, those are merely options on the table for discussion and no such decisions have been made.
Discovery's launch on the first post-Columbia mission was aborted July 13 a little more than two hours before the planned liftoff when engine cutoff - ECO - sensor No. 2 failed to respond properly to an automated test. The cutoff sensors are part of a backup system that ensures the shuttle's main engines don't run too long, draining the tank dry, if some other problem prevents a normal, on-time shut down. An engine that ran out of fuel while running at full throttle likely would tear itself apart and NASA's launch commit criteria require all four ECO sensors to be operational for a countdown to proceed.
The ECO sensor system is simple in concept but complex in operation. Rather than repeat those details here, please see the CBS News/Spaceflight Now ECO sensor page for graphics, details about the system's operation and a chronology of the sensor problems that have plagued Discovery since its initial tanking test April 14: here.
While engineers are leaving no stone unturned trying to isolate the problem and fix it, NASA managers have directed one of the troubleshooting teams to study the rationale for amending the launch rules to permit a takeoff with just three of four operational ECO sensors. That's assuming, of course, engineers can successfully demonstrate that whatever prevented ECO sensor No. 2 from operating properly during Discovery's countdown won't affect other sensors after launch.
Several NASA sources familiar with the discussions pointed out the ECO sensor system is a backup that only comes into play if some other major problem affects engine performance or fuel usage. On top of that, the ECO system is internally redundant. Even if another problem required the ECO system to work, three of the four sensors would have to fail "wet" - falsely indicating they were still submerged in liquid hydrogen - before the system would allow the engines to drain the tank. On top of all that, the shuttle is launched with more propellant than it needs.
But in the wake of the Columbia disaster and the recommendations of the Columbia Accident Investigation Board, it may be difficult for NASA to change an LCC "on the fly," regardless of how technically sound the engineering rationale might be. And in any case, deputy shuttle program manager Wayne Hale, chairman of NASA's mission management team, is on record saying he likely would oppose such a move.
"Going down the logic path, one of our safety requirements on this vehicle is that we are two-fault tolerant in our electronics," Hale said Friday. "We can take two failures and can continue to keep on flying safely. And anytime you step away from that standard, you incur risk and you'd better make sure you have an airtight story to step away from that posture.
"If we get to the end of all this troubleshooting and everything's working fine, we may come around to the discussion of 'what if.' But we're not ready to go there yet."
Many engineers believe a tanking test of some sort will be required if the team fails to find anything wrong in the electronics aboard Discovery that route data from the fuel tank sensors to the shuttle's flight computers.
If a test ultimately is approved, engineers may swap the wires leading from a control box to ECO sensors 1 and 2. If the No. 2 sensor malfunctioned again, engineers would have evidence the sensor itself, and not the wiring, is suspect. If the No. 1 sensor acted up, it would indicate a wiring problem. It's also possible all four sensors would operate normally. Either way, engineers would have additional data to help make an informed decision about how to proceed and whether or not amending the four-of-four ECO sensor launch constraint could be justified.
It's not yet clear, however, whether a tanking test is possible before Discovery's launch window runs out July 31. The end of the window is defined by the lighting in space when the shuttle's external tank separates. NASA managers want good lighting so cameras on the shuttle can photograph the tank for signs of foam insulation loss.
Extending the launch window to Aug. 4 would result in a bit more shade on parts of the tank as seen by a camera mounted in the belly of the space shuttle. Initial studies indicate the camera could still achieve the desired results, but again, no final decisions have been made.
The window cannot be extended past Aug. 4 under any scenario because the lighting for launch itself would not be sufficient.