HOUSTON – NASA’s next space shuttle flight will not launch before Sept. 22 as engineers struggle to understand and fix the persistent foam debris issues with the launch system’s external tank, agency officials said Aug. 5.

NASA had targeted Sept. 9 to launch the Atlantis orbiter on its STS-121 spaceflight, a second test of fixes made in response to the 2003 Columbia disaster. But the external tank foam shed during Discovery’s launch and other mission-processing activities have eaten away at that flight window, which closes Sept. 26, NASA officials said.

Despite the new concerns that the space shuttle’s external fuel tank is still capable of shedding pieces of foam insulation large enough to cause a serious problem like the one that doomed Columbia and its crew during re-entry Feb. 1, 2003, NASA has not given up on the idea of launching the next mission in September.

“I don’t presume the worst, I don’t assume the best, I like to go where the data takes me,” NASA Administrator Mike Griffin said Aug. 5 in a press conference at Johnson Space Center here.

While acknowledging that the cause and solution of the foam problem would have to become apparent very quickly for the agency to commit to a shuttle launch in September, Bill Gerstenmaier, the space station’s program manager, told reporters Aug. 5 that NASA has not given up on launching the Space Shuttle Atlantis then, but is now targeting the end of the two-week window, or no earlier than Sept. 22.

“Until we run out of lead time to make the September window we will preserve it … because that is what the taxpayers pay us to do,” Griffin said. “When we can no longer preserve the window, we will reset for November.”

Griffin has appointed a tiger team to troubleshoot the cause of the foam shedding witnessed during Discovery’s July 26 launch.

Gerstenmaier said he expects his first briefing from the tiger team Aug. 9. The tiger team spent the week of Aug. 1 at Lockheed Martin’s Michoud Assembly Facility near New Orleans, where the space shuttle external tanks are built.

“If next week, the guys have a eureka moment on the foam and say ‘yes, we understand it’ … then we’ll go forward,” Griffin said.

NASA mission managers decided Aug. 2 that Discovery was ready to return to Earth on Aug. 8, and ruled out the need to conduct a fourth spacewalk to repair a damaged thermal blanket that came loose near one of the orbiter’s windows.

Gerstenmaier said he had not reviewed a 2004 internal NASA memo, first reported Aug. 3 by The New York Times, criticizing quality control on some foam application techniques. The report stated that engineers “did not do a thorough job” of tracking the minute variations in hand-applied foam, The Times reported.

“It’s available, I’m sure, in all the other documentation that the teams are looking at,” Gerstenmaier said. “We’ll take that information and see if there are some things there, again from a technology standpoint or from an engineering standpoint, that we can use and apply.”

During Discovery’s July 26 launch, video from a camera mounted to its external tank recorded several pieces of foam insulation peel away during the ascent. A large, 0.4-kilogram chunk visibly popped free from a ramp previously thought safe from foam shedding. That chunk did not strike the orbiter, but at least three other foam pieces that also separated during the launch were too large to be considered acceptable, shuttle officials said.

The foam debris from Discovery’s external tank disappointed shuttle engineers and Discovery’s astronaut crew, given that NASA has spent two and a half years and about $200 million of the $1.4 billion devoted to its post-Columbia accident work toward revamping orbiter external tanks to prevent harmful foam shedding. Shuttle officials said they will not launch another shuttle until they understand and address the foam issue.

A 0.8-kilogram of foam fell from Columbia’s external tank during its launch and pierced the heat shield panel lining its left wing leading edge. That wing damaged allowed hot atmospheric gases to enter the wing during re-entry on Feb. 1, 2003, leading to Columbia’s destruction and the deaths of all seven astronauts onboard, investigators found.

Gerstenmaier said that all of the imagery collected of Discovery’s launch and subsequent orbital inspections has given engineers a wealth of data. “We learned a lot from this flight,” Gerstenmaier said. “The next step … is to look at the future tanks that are coming and see if there any applications from what we learned.”

Only then will engineers decide whether to modify the external tank for Atlantis, which stands mated to its external tank-solid rocket booster launch stack in the massive Vehicle Assembly Building at Kennedy Space Center, or shift the orbiter to a completely new tank, Gerstenmaier added.

While the Sept. 22 launch date for Atlantis shaves about two weeks from its flight window, there are still multiple opportunities to launch the shuttle within the narrow flight window.

“It’s still gives us four launch attempts toward the end of the window, and still looks good from a planning standpoint,” Gerstenmaier said.

Brian Berger contributed to this story from Washington.