Shuttle leaders decide to remove fuel tank foam ramps
BY WILLIAM HARWOOD
STORY WRITTEN FOR CBS NEWS "SPACE PLACE" & USED WITH PERMISSION
Posted: December 15, 2005
NASA engineers and managers have recommended the removal of a protective foam air deflector from the shuttle's external tank to eliminate a major source of potentially dangerous launch debris, a top agency official said today. While NASA has not given up launching the next flight in May, additional work to implement and certify other changes required by the deflector removal could push the launch to later next summer.
Bill Gerstenmaier, NASA associate administrator for space operations, would not discuss possible launch targets during an afternoon teleconference with reporters, saying such talk was premature and put unnecessary "schedule pressure" on the team. Instead, he said, engineers were going to focus on thoroughly assessing the possible causes of foam shedding based on recent insights into cracks found in a tank slated for the flight after next.
"We're going to do the right thing and let the data drive us where we need to go," he said.
During a Program Requirements Control Board meeting at the Johnson Space Center in Houston today, engineers recommended removing the external tank protuberance air-load - PAL - ramps from the shuttle's external tank. Last month, shuttle program manager Wayne Hale said engineers did not believe NASA would be ready to launch a PAL-free tank until the third mission in the upcoming sequence.
But computer modeling now shows the structures the ramps were designed to protect - two pressurization lines, support brackets and a critical cable tray - are beefy enough to withstand any expected aerodynamic forces, or loads, they might encounter during ascent. Wind tunnel tests are planned for February to verify the results of the computer modeling.
But engineers are still assessing what changes will be required to so-called ice-frost ramps, areas of foam around the brackets supporting the pressurization lines that are intended to prevent ice formation before blastoff.
"Based on the data we've reviewed, we don't need to do any redesign of the cable tray or the press lines," Gerstenmeir said. "So there's no redesign of the basic bracketry or the basic tank structure. That all looks fine. We'll confirm that with the wind tunnel tests. The margins look very good in that region. ... So that's a very positive thing, it takes a lot of that work out of the critical path.
"From an overall standpoint, depending on what engineering solution gets picked (to address the ice-frost ramp issue), May is still very viable. Some of the other engineering solutions that would require a more detailed certification process of foam applications, etc., may move us somewhere else."
When the shuttle was designed in the 1970s, engineers believed shock waves during the vehicle's transition to supersonic speeds could cause potentially catastrophic damage to the pressurization lines and/or the cable tray. As a result, the tank was equipped with two PAL ramps, one running along the upper section of the hydrogen tank and the other along the outside of the oxygen tank.
During Discovery's launching last July on shuttle mission STS-114 - the first post-Columbia flight - a one-pound chunk of foam ripped away from the hydrogen PAL ramp just after the ship's solid-fuel boosters were jettisoned two minutes and five seconds into flight.
Earlier this fall, Hale ordered all existing PAL ramps dissected and removed. Engineers were assessing new fabrication techniques as a possible solution when numerous small cracks were discovered prior to the removal of the PAL ramp of a tank slated for the third post-Columbia mission.
Engineers now believe the cracks almost certainly are related to the thermal stress the tank undergoes when it's loaded with super cold rocket fuel. The tank in question - ET-120 - originally was to be flown by Discovery last July and it was fueled for two pre-launch tests. The tank later was replaced for unrelated reasons, but the cracks could have been caused by thermal stress.
If that theory is correct, cracks would be a constant threat for shuttles that were fueled for launch and then delayed for other reasons. Whether similar cracks played a role in the foam loss experienced by Discovery's tank in July is an open question.
But given the small size of the cracks, the difficulty in detecting them at the launch pad and the lack of any on-pad repair procedures, NASA managers opted to forego attempting to fix the PAL ramps and to recommend their removal instead.
"They've done a tremendous job of chasing the cracks," Gerstenmaier said. "We saw the cracks in the non-destructive testing, in the X-rays and in the terahertz radar data. What they've done now is, they've carved out the cracks and they've followed the crack as it goes down into the foam and then they see the crack actually break into smaller cracks, into some delamination layers down internal to the foam. Then we see some of the cracks progress slightly underneath the cable tray, we see some of the cracks actually (branch) out into the acreage foam. So we know where the cracks are running.
"What we don't know is really why they're initiating. We have some theories. ... There are two different types of foam. There's the foam that's on the acreage of the tank that's about an inch thick and then we spray this PAL ramp foam on top of it. Those two types of foam have different thermal expansion coefficients, so we now have a math model of the stress field between those two foams as they chill down. That's a key driver. Where we put one type of foam on top of another type of foam, it causes that lower layer of foam to be colder, it's not as strong at cold temperatures so the cracks can initiate down there and then carry up into the upper piece of foam."
Gerstenmeir said a major unrsolved question is how temperature changes and tank pressurization interact to cause cracks. "Exactly how they fit together, we haven't had a chance to pull together," he said. But he provided a relatively detailed summary of how a crack might lead to foam shedding.
"The tank has a one-inch layer of foam all around and then the PAL ramp is about, probably an eight- to 10-inch foam layer (is) sprayed on top of that," he said. "What happens is that foam that's sprayed on top, it provides an insulating layer that allows that one-inch piece of foam to get cold through its entire length, so it is now cold from the surface of the tank all the way to the top of that one-inch layer of foam.
"The fact that that foam is cold, when it gets cold it loses some strength, or it's more brittle, and easier to crack. And then the fact that the foam on top is expanding at a different coefficient of thermal expansion, it causes a shear, or a stress layer, on top of that one-inch layer of foam as well as there's a shear layer, or a stress layer, down where the foam attaches to the tank.
"So then, that can cause a crack to initiate somewhere in that underlying one-inch piece of foam and then that crack can then propagate on up into the foam on top of it. If that crack goes all the way to the surface, that's where the outside air is, and it goes all the way down to the tank, then that provides a path for air to come in, liquefy in that lower region next to the tank."
As the super cold hydrogen in the tank is drained and the structure warms up, "that liquid air then expands, pushes out on the foam and can cause a large piece of foam to come off," Gerstenmaier said. "So that's kind of the theory we're looking at."
Removing the PAL ramp will eliminate the problem with differential contraction and expansion based on the use of two different types of foam. But engineers still must decide what will be needed to eliminate foam shedding around the ice-frost ramp brackets that support the pressurization lines.
"The ice frost ramp sticks under the cable tray a little bit and it sticks into that region where the PAL ramp was and we saw some cracking in that ice frost ramp right next to the cable tray, or actually maybe a little bit under the cable tray," Gerstenmaier said. "What we need to do is understand what caused that cracking and will it be prevented by just removing the PAL ramp? Do we need to do something in that area, do we need to put some kind of other foam over that (to prevent) ice formation in that area? That's the kind of work we need to go understand.
"The engineering teams have many solutions for that. We'll check some of those out analytically over the next couple of weeks, we'll pick an engineering solution, probably in the next week or two, then schedule a little bit of confirmation tests or analysis in the early part of January. And then if we need to make a physical fix on the tank by spraying foam, we would do that in the middle part of January to the later part of January and then we would ship the tank around the first part of February."
If that schedule holds up, NASA would have a theoretical shot at launching the shuttle Discovery in May. Because of a post-Columbia decision to launch at least the first two return-to-flight missions in daylight - and to ensure the external tank separates in daylight half a world away - NASA can only launch during relatively short "windows." The next three such launch windows are:
The tank currently slated for use in that mission, ET-120, has so many small cracks, many of them penetrating all the way to the aluminum skin of the tank itself, it may make more sense to process another tank, ET-118, in its place. Gerstenmaier said today a second tank will, in fact, be ready in time to support Discovery's mission whenever it is cleared to fly.