Engineers have not yet found any obvious manufacturing problems or defects that might explain the potentially catastrophic foam loss that marred shuttle Discovery's July 26 launching. While a variety of possible fixes are under study, it's not yet known how long it might take to identify root causes and as a result, there is virtually no chance of launching the next shuttle flight in September as originally planned, a top NASA manager said today.

"There is nothing that became very obvious to us from an overall standpoint that said there was anything truly unique about this tank that's different than the other tanks that would easily clear us to go ahead an fly the next tank," said Bill Gerstenmaier, manager of NASA's space station program.

"So if I take the preliminary results I've gotten to date from the team, I think more than likely we're going to have to make some kind of minor engineering modification to the tank and to do that probably impacts the September launch window. We probably won't make the September launch window."

NASA chart shows the location of foam loss from the external fuel tank on STS-114. Image: NASA.

NASA had hoped to follow up Discovery's mission by launching the shuttle Atlantis around Sept. 9. But Discovery's launch was delayed and because Discovery must be ready for takeoff as a rescue vehicle in case of major problems with Atlantis, only four days - Sept. 22 through Sept. 25 - are available for the second post-Columbia flight. And that was before Discovery was diverted to a California landing, which will add a week to its ground processing.

Gerstenmaier discussed initial findings from so-called "tiger teams" of engineers trying to find out what caused relatively large pieces of foam insulation to fall off Discovery's external fuel tank during the climb to space July 26.

The largest and most alarming piece was a 0.9-pound chunk of foam that separated from a so-called protubereance air load - PAL - ramp used to smooth the flow of supersonic air over a cable tray and two pressurization lines. The foam peeled away two minutes and seven seconds after blastoff, just two seconds or so after Discovery's solid-fuel boosters were jettisoned.

Last week, NASA confirmed that during the tank's assembly, engineers had to repair minor damage in the same area of the PAL ramp. But Gerstenmaier said today the damage in question measured just two-tenths of an inch across and "that alone probably wouldn't be enough to cause the foam loss we saw.

"We still need to confirm this (but) there's probably another underlying cause."

The external tank is equipped with two PAL ramps, one that runs along the outside of the upper liquid oxygen section and a 38-foot six-inch-long ramp that runs along the upper part of the hydrogen section. The hydrogen ramp, where Discovery's debris originated, is made up of foam that is sprayed on and shaped by hand. It weighs a total of 22 pounds and measures less than six inches thick.

The PAL ramps were not modified in the wake of the Columbia disaster and engineers were hopeful the foam loss seen in Discovery's launch was caused by a manufacturing defect or some other isolated, non-generic problem that would not require a fleet-wide redesign.

But Gerstenmaier appeared to play down that possibility today, saying "we didn't find any obvious root cause," generic or otherwise.

"Probably the root cause will be some combination of events," he said. "But we didn't see anything that showed us the specific root cause. We just need to keep looking. There's no immediate answer or problem that jumps out at us."

While the PAL ramp foam has received the most attention, foam also separated from four other areas of the tank:

* LH2 tank acreage: Two divots, one measuring about four inches across and the other eight to 10 inches in diameter, were seen in an area where foam is sprayed on by machines.

* Intertank/LH2 flange closeout: An area that was heavily modified in the wake of Columbia, the intertank flange is where a cylindrical section of the tank joins the lower part of the oxygen section and the upper part of the hydrogen section. In past flights, foam shedding from this area was commonplace. In Discovery's flight, two divots were seen, both measuring about three inches across.

* Bipod fitting closeout: Columbia was brought down by a large 1.67-pound piece of foam that broke away from an aerodynamically shaped ramp insulating one of the two struts holding the hose of the shuttle to the tank. The bipod ramps were eliminated and replaced by heaters. But a piece of foam measuring 8.4 by 7.3 inches broke away near one of the exposed bipods 148 seconds after launch.

* LH2 ice/frost ramps: Foam ramps are used to smooth the flow of air over fitting that attach two pressurization lines to the tank. The pressurization lines run adjacent to the PAL ramps. During Discovery's launch three of these ice/frost ramps suffered minor foam loss.

Despite the obvious problems, Gerstenmaier said overall, Discovery's tank performed well.

"When you look at the amount of foam on the tank and the performance overall, the tank really did a pretty phenomenal job," he said. "When you look at it emotionally, you kind of feel one way. But then when you start looking at the hard engineering data, you sure get a different feeling."

The tank was launched with 4,192 pounds of sprayed-on foam insulation. Ascent and post-separation photography showed just 1.2 pounds of foam was liberated.

"So in terms of the aggregate, that's a pretty small number, one out of 4,192," Gerstenmaier said. "It's still not the performance we'd really like, but it gives you a feel for how the tank performed overall."

While the intertank flange area was the subject of a major post-Columbia redesign, "we had two small divots three inches in diameter. But again, that's pretty phenomenal performance compared to what we used to see on previous flights. We had lots of foam loss in that area and this is a dramatic reduction."

Gerstenmaier said a wide range of possible fixes were under discussion, including the removal of current PAL ramps so they can be rebuilt with a "clean spray" in a controlled fashion. Other ideas include using fiber or mesh-like material in specific areas to add strength.

The loss of foam near the bipod struts may have been caused by a post-Columbia heater wire that allowed air to get underneath the foam. Nitrogen in the air then could have liquified due to the ultra low temperature of iiquid hydrogen. During launch, that liquid would have changed back into a gas, possibly blowing out overlying foam.

"We may have some potential for cryo ingestion through a heater wire that runs through there," Gerstenmaier said. "We may make a minor change to the heater wire to try to alleviate that if that appears to be a problem."

Whether such fixes might work remains to be seen.

"We're going to have to really understand why this foam came off," he said. "We're going to have to spend the time to do the root cause analysis and understand with a high probability what the underlying mechanism, or mechanisms, were that contributed to this foam loss. Then the techiques we use to either put the foam back on or to change it have to accommodate and rule out and remove all those leading causes that could have led to the foam loss.

In the end, he said, "I think you're going to find differing mechanisms for different areas."

"There may be a different mechanism for foam loss associated with each of these different areas," he said. "Then the question is can we control that?"