• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Design Solution Created Problem with Shuttle Columbia

Posted on: Wednesday, 16 April 2003, 06:00 CDT

Orlando Sentinel -- Like all NASA engineers of his generation, Jim Odom welcomed a challenge.

A ranking manager in the earliest days of America's space-shuttle program, he embodied NASA's "can do" attitude.

Back then, in 1972 and 1973, Odom was in charge of developing a 154-foot external fuel tank; his was one of four projects that were simultaneously designing the major components of what would ultimately come together to launch NASA's first shuttle -- Columbia.

Under Odom's oversight, engineers had designed a huge aluminum vessel to be filled with 535,277 gallons of super-cold liquid hydrogen and oxygen that would power the shuttle into orbit. It looked good -- until the designers of the orbiter noticed a problem.

Because the fuel tank was largely bare metal, they said, it would immediately ice up when it was filled with propellant. At liftoff, that ice would cascade off the tank -- and shred the delicate, silicon-based tiles the orbiter's design team had devised to protect the spacecraft against the heat of re-entry.

Change the tank design, they said. Can do, Odom said -- never realizing that the fix would become the likely cause for the loss of Columbia 30 years later.

"Whatever the problem is, you go and solve it," said Odom, now retired from NASA's Marshall Space Flight Center and living in Decatur, Ala. "You don't have a debating society because you're all on the same team."

The engineers' solution seemed simple enough: coat the entire tank with spray-on polyurethane foam insulation. And it worked, largely eliminating ice formation on the tank's surface.

But it created a second problem: Pieces, and sometimes chunks, of foam flew off the tank at liftoff. In a stunning display of unintended consequences, the very element added to prevent debris damage on the orbiter became a persistent debris source itself.

Foam has dropped or popped off the fuel tank since the start of the program. In the course of 113 missions, falling foam has gouged sensitive heat-insulating tiles, banged off the solid-rocket boosters and knocked against the wings -- despite rules stating that no shuttle elements should strike and damage the orbiter's thermal protection system.

And, on Jan. 16, a chunk weighing two pounds or more, moving at a relative speed of nearly 500 mph, slammed into the leading edge of Columbia's left wing. That collision, investigators suspect, breached a hole in the orbiter's thermal armor, causing it to disintegrate during re-entry Feb. 1.

The Columbia Accident Investigation Board, charged with finding out what happened to the doomed spacecraft, is now asking why NASA accepted foam loss for years, and why it was considered just an annoying and costly maintenance issue, not a risk to flight.

It's a question that even NASA officials are now acknowledging should have been asked earlier.

"In retrospect, I look back and say, 'Wow, I wish we'd spent more time on that. Maybe the program didn't put enough emphasis on that,' " said Bryan D. O'Connor, head of safety and mission assurance at NASA's headquarters.

"That's easy to do in retrospect. But I can tell you, at the time it didn't register as an immediate concern. It's a long-term concern, because it's a maintenance issue."

It's not as if NASA engineers didn't try to fix the problem. Through the years, they tried several solutions -- poking holes in the foam to ventilate air pockets; shaving it to remove hardened outer layers; changing how it's applied -- but had little success. The number of debris hits averaged more than 140 per flight, with more than 25 longer than an inch. And NASA has attributed 90 percent of that damage to foam.

The origin of that debris, engineers conclude, goes back 30 years, to how Jim Odom and engineers from what was then Martin Marietta Corp. tackled a problem not of their making, but one that they had to solve.

In the early days of America's manned space program, ice debris wasn't an issue. The stages of the giant Atlas and Saturn rockets that lifted astronauts into space and then sent them to the moon were often covered with ice at liftoff; it crashed away harmlessly as the rockets rose above the launch pad.

The shuttle's design, however, had the orbiter riding astride the much larger external tank.

By the early 1970s, when Martin Marietta was competing for NASA's external-tank contract, engineers focused primarily on insulating the tank's propellants from extreme heat that would rush over the aluminum surface as the shuttle accelerated to 17,500 mph during its ascent.

Foam would be sprayed on parts of the liquid hydrogen tank, which dropped to temperatures below minus 420 degrees Fahrenheit. But it was not expected to prevent the creation of ice on the tank's surface.

"Large areas of ice and frost would be expected to be formed, as was seen on Saturn, but this had not been identified as a concern," wrote Myron "Mike" A. Pessin, the external tank's former chief engineer, whose technical history of the tank was published last year.

Martin Marietta's earliest designs came in during late 1972 and 1973. They were rated the most cost-effective, received the highest technical grades and, Odom said, won the contract hands down.

But that was before the tank's designers learned that the orbiter's underbelly would be lined with the fragile tiles.

Odom, who managed the external-tank project at the Marshall flight center, said the decision to use foam over large areas of the tank evolved from 1972 to 1974.

"As the tile development progressed and we determined that they'd be relatively fragile, we realized we'd have to reduce the ice [on the tank's exterior]," Odom said. "That was an evolving change. That took place over some period of time. Ice was a different problem."

The choice -- adding insulation to the tank or leaving the orbiter's thermal protection system open to ice damage -- was obvious, he said.

In retrospect, Odom said, he would redesign the external tank to eliminate as many indentations and protuberances as possible -- especially the ribs in the so-called "intertank" area. Foam adheres better to a smooth surface.

"The basic size of the tank had to be what it is," Odom said. "You might have made the outside skins smoother."

But there wasn't time for that. There were tests to be done; components to mate; a shuttle to launch, to show an impatient American public that NASA was ready to take the next step to vault man permanently into space.

So there was no redesign. Just foam.

From the first shuttle flight in 1981, foam routinely popped and chunked and tumbled off the intertank. It wasn't hard to figure out why.

Unlike the rest of the tank, the 22.5-foot-long intertank section isn't smooth. It's ribbed -- corrugated for strength -- with narrow peaks and valleys, not to mention a series of lines, bump-outs and extensions that make foam adhesion a challenge.

The intertank joins the liquid oxygen tank on top to the liquid hydrogen tank below, like a collar, and is also the place where the front of the orbiter mates with the external tank. It needs to be strong.

Because of its location, the intertank is stressed by the forces of the oxygen and hydrogen tanks inside, the weight of the orbiter above and the fast-moving airflow running between the tank and the orbiter. Plus, that section of the tank experiences sub-freezing temperatures at both ends.

"Ribs are a very efficient structure to carry loads," Odom said. "It gets you maximum strength at minimum weight."

The intertank had another problematic area, where the "bipod" struts attach the orbiter's front end to the tank. To insulate those attachment points, and to accommodate aerodynamic forces, engineers designed a "ramp" to cover each attachment.

A foot high and 21 inches to 34 inches long, the ramps are hand-sprayed and hand-sculpted. But it's easy for voids to occur inside the foam. And on ascent, these air pockets can literally burst, dislodging chunks of foam.

Investigators think one ramp tore away from the tank and struck Columbia's left wing. And while the chunk that hit Columbia is the biggest recorded, it wasn't the first to fall from the ramp area.

Neither the intertank nor the bipod attachment area included foam in the original designs. And ever since it was decided foam would be applied to those areas, the tank's insulation history was one of trial and error.

During the first 16 years of the shuttle program, external-tank engineers reduced the angle of the bipod ramp, changed the intertank foaming process repeatedly and even drilled holes in the intertank foam to ventilate air pockets -- all attempts to reduce the amounts of debris striking the orbiter.

But none of the efforts eliminated the problem.

During a November 1997 Columbia flight, the shuttle was peppered with 308 debris hits. Its vulnerable underbelly was scarred by 109 "major" hits larger than an inch. Some divots measured 8 inches to 15 inches long. Two months earlier, Atlantis had also suffered significant damage, including one hit measuring 6.5 inches.

The reason: NASA had again switched the foam process, this time because the propellant used to apply the foam had been banned by new federal environmental rules. NASA suspected the change was causing the increased foam loss.

The damage trend forced engineers to try new measures. Teams reworked the foam on the intertank before Endeavour flew in January 1998.

During the next two years, engineers developed a more precise system to vent holes around the intertank -- and thought they had finally fixed the problem.

Between 2000, when NASA declared success, and 2002, the number of hits on the orbiter each flight averaged 109.

But last October, two flights before Columbia's final mission, NASA discovered significant bipod foam loss on Atlantis. It was the first time in 10 years -- and only the fourth overall -- that foam loss from this area had been reported. However, engineers now say there may have been more instances that weren't noticed, either because the flights were at night or monitoring cameras didn't spot them. NASA proposed more changes to the intertank area, including the removal of a silicon-based insulation that goes beneath the foam.However, a final decision to adopt the change wasn't made before Columbia's launch in mid-January.

Since the Columbia accident, NASA has taken up a complete redesign of the bipod ramp that will likely eliminate the foam there, according to NASA documents.

Last week, J. Scott "Scotty" Sparks, a NASA materials engineer, told members of the Columbia investigating board about NASA's longstanding efforts to reduce debris. But when asked a series of probing questions about the persistent foam problems, he referred to the tank's earliest designs.

"You've really got to go back to the beginning as far as the design of the tank," Sparks said. "I'm not so sure that the TPS [thermal protection system] processors were in the same room when they designed the tank. It's not designed for the TPS to be processed on there."

Sparks singled out the intertank area. "If you were to redesign, completely, a tank, you would make the external a bit smoother," Sparks said. "You'd have those people in the same room."

New tanks being built by Lockheed don't even include the ramps as the company awaits final word on their redesign. And CAIB board member Brig.

Gen. Duane Deal pointed to one ramp this week as if it were a product being recalled.

"You're looking at a piece of history over there because that design is never going to fly again," Deal said.

By Anthony Colarossi and Mary Shanklin

Gwyneth K. Shaw and Robyn Suriano of the Sentinel staff contributed to this report.

-----

On the Net:

NASA

More science, space, and technology from RedNova

To see more of The Orlando Sentinel -- including its homes, jobs, cars and other classified listings -- or to subscribe to the newspaper, go to http://www.OrlandoSentinel.com

(c) 2003. Distributed by Knight Ridder/Tribune Business News.

More News in this Category