Engineers hope to begin analyzing data this weekend from a recovered tape recorder that stored readings from some 721 sensors throughout the shuttle Columbia during its final 45 minutes of flight. Analysts hope the data will help them precisely map out the flow of hot gas through the doomed ship's left wing to confirm and refine - or possibly modify - current theories about where the initial breach occurred and how the deadly plume then worked its way through the interior of the wing.

The OEX recorder appears to have survived in good condition. Photo: CAIB

The orbiter payload experiments, or OEX, recorder was found during a grid search March 19 near Hemphill, Texas. The device was remarkably intact with one edge buried about three inches in soft soil. Over the weekend, engineers with Imation Corp. in Minnesota cleaned and stabilized the 1-inch-wide, 28-track tape and sent it on to the Kennedy Space Center for duplication. Engineers at the Johnson Space Center in Houston hope to begin analyzing whatever data was recorded this weekend.

"In a perfect world, we will have 721 measurements, sensor outputs," said Scott Hubbard, a member of the Columbia Accident Investigation Board. "This covers the wings, the fuselage and the vertical tail surfaces. Of particular interest are going to be 182 pressure measurements, 53 temperature measurements and 447 loads, dynamics and stress measurements.

"Among these will include a series of measurements - if we get all the data back - on the left wing: strains, stresses and strains, temperature and pressures.

"Now we have to note, this has been through a very severe environment," he cautioned. "We don't know if the tape has been demagnetized, we don't know yet the quality of the data in there."

But if the tape is, in fact, readable, "this would be a gold mine of information that could lead us to a much, much better understanding of what happened throughout the vehicle and particularly on the left wing."

During a news conference following a public hearing by the CAIB, Hubbard said additional study of enhanced photography of foam debris slamming into Columbia's left wing shows the material hit the underside of the ship's leading edge in a roughly two-square-foot region centered on a specific panel of protective carbon composite material just forward of the left main landing gear wheel well.

Illustrations show where the foam impact is believed to have occurred during launch. Photo: CAIB

Ongoing analysis of recovered debris, meanwhile, shows the concentration of metallic slag splattered on the inner surface of the leading edge panels increases as one approaches the region where the foam impact occurred.

The two independent lines of evidence are consistent with a breach at or near reinforced carbon carbon - RCC - panel No. 6, or an adjacent closeout panel, that allowed a deadly plume of superheated air to burn its way into the wing, triggering a catastrophic chain of events that led to Columbia's destruction 200,000 feet above Texas.

At today's hearing, two CAIB investigators revealed yet another piece of evidence that supports a burn through near the suspected foam impact point at RCC-6: A partially melted RCC support strut from just behind RCC panel 11, located outboard of RCC-6, that shows hot gas in a cavity enclosed by the wing's RCC panels was moving outboard, as one would expect from a breach located closer inboard to the fuselage.

"There's actually an open space inside the leading edge," said Gregory Kovacs, a professor at Stanford University. "So it's possible if there was a breach, something could have flowed down that open space and then eroded the components in there. So on that score, I think we are thinking about flows down there, where they're concentrated maybe closer to the breach. That's the kind of thinking we're going through."

As it turns out, two temperature sensors monitored by the OEX recorder were mounted underneath insulation on the wing spar behind RCC panels 9 and 10. If a breach really did occur inboard of there, near RCC-6, for example, those sensors may have recorded the initial intrusion of hot gas before it ate its way into the interior of the wing.

"It's expected that if this was a normal mission, (the OEX recorder) would have about 30 minutes of data from ascent and in a nominal mission, an hour of entry data," Hubbard said.

But Columbia's re-entry ended 45 minutes after it began. In addition, the wiring from many sensors in the left wing was routed around the left main landing gear wheel well. Many of those wires may have been severed by the super-heated plume that ultimately burned its way into the wheel well. Even so, that would not affect data from other sensors expected to shed light on the thermal environment and the aerodynamic forces acting on the ship in its final moments.

To put all of that in perspective, here is a brief recap of what investigators believe may have happened to Columbia:

• A large piece of foam insulation weighing about one pound and measuring three inches thick, 12 inches wide and 24 inches long pulled away from the external tank bipod ramp area 81 seconds after launch. The bipod ramp is just that, an aerodynamic ramp of shaped foam just in front of one of the two struts that held Columbia's nose to the tank.
  
  
• The foam, moving at roughly 500 mph in the slipstream between the orbiter and tank, hit the lower side of the left wing leading edge in a two-foot-wide footprint centered on RCC panel 6. It may have damaged parts of RCC panels 5, 6 and 7 or, more likely some believe, one or two so-called carrier panels that form a flush surface between the RCC panels and heat-shield tiles permanently bonded to the lower surface of the wing.
  
  
• The day after launch, military radars tracked an object leaving the space shuttle. It is not yet known what that object might have been, but sources close to NASA's investigation believe it may have been part of an RCC panel or one of the carrier panels in question. Tests to determine precise radar signatures of various types of possible debris are not yet complete.
  
  
• During re-entry, hot gas breached the leading edge from the presumed foam impact damage site. The gas presumably first moved through the cavity behind the RCC panels, damaging various support structures, before entering the wing itself. At about this same time - before any realtime sensors detected anything unusual - the shuttle's flight control system began responding to unusual aerodynamic forces. Whatever was wrong with the front of the wing presumably was disrupting the airflow over the wing's upper and lower surfaces enough to prompt a response from Columbia's digital autopilot.
  
  
• The plume began melting leading edge support components and the forward wing spar, splattering molten metal back across the inner surface of the RCC panels. The timing of such splattering is uncertain, but the slag can be seen on recovered debris, concentrated near the region of the presumed breach.
  
  
• At the same time, pieces of the leading edge, support structure or adjacent tiles began falling off. Some components may have fallen into the worsening breach. "If you think about the static pressure loads, certainly at the beginning, there's a lot of force," said Gregory Kovacs, an engineering professor at Stanford University. "So if a piece is sort of flapping around and it's on the underside, if you think about the (40-degree) angle of attack, it's easy to believe a piece could be folded and broken off and end up wedged inside there. It would be driven back as things are melting and if it's RCC (material), it may well survive that and end up stuck there."
  
  
• A brake line temperature sensor in the left main landing gear wheel well recorded a slight but unusual increase as hot gas swirling into the wing entered the wheel well through a vent on its forward face.
  
  
• The breach worsened as Columbia's descent continued until a plume of super-heated air impacting the side or forward face of the wheel well burned through wiring from sensors located toward the rear of the wing. Lost data from those sensors were among the first indications in mission control that anything was amiss.
  
  
• The plume eventually burned through the outboard wall of the wheel well, or perhaps through its forward face depending on the actual breach location, dramatically raising temperatures in the well. A titanium pin used to mount the left landing gear door to the outboard side of the wheel well is partially melted, indicated it was subjected to temperatures as high as 3,400 degrees. Other debris from an area just inboard of the forward corner of the door shows signs of hot gas outflow. Engineers believe the plume of hot gas entering the wheel well likely exited around the corners of the heat-softened landing gear door.
  
  
• Columbia's flight control system struggled to keep the shuttle on course, counteracting increasing aerodynamic drag by adjusting the ship's roll trim and, when that wasn't enough, firing four right-side yaw thrusters. Eventually even that was not enough, and the spacecraft yawed out of control, breaking up at Mach 18.
  
  

"We have focused a lot of our attention on the left wing and particularly the left wing leading edge," Hubbard said. "If the sensors that are there were active and working and got recorded, that information would be extremely valuable. If for some reason those sensors dropped out, the orientation of the vehicle, from some of the other sensors that also help us determine the attitude of the vehicle, might give us a better sense of what was going on."

Starting around April 6, engineers at the Southwest Research Institute in San Antonio, Texas, plan to being firing large pieces of foam insulation at RCC panels and carrier panels at various angles of incidence to determine what sort of damage might have occurred during launch. The foam will be fired from a nitrogen gas canon at impact velocities of around 500 mph.

An RCC panel and upper and lower carrier panels taken from the prototype shuttle Enterprise will be heavily instrumented to determine impact stresses. An RCC panel taken from the shuttle Discovery, which has flown 30 times and closely matches the original-equipment panels that were on Columbia for its 28th flight, also will be subjected to foam impact for a more realistic assessment of what might have happened to Columbia. In addition, foam will be fired at so-called acreage tile, that is, normal heat-shield tiles located in the presumed foam debris footprint just behind the lower carrier panels.

In the meantime, NASA and CAIB investigators are continuing detailed chemical and metallurgical analysis of recovered debris to decipher complex damage patterns in hopes of pinning down the breach location and plume propagation path.

Mark Tanner, an expert in failure and forensic analysis with Mechanical and Materials Engineering, said "what we're trying to do is get an idea from the flow patterns, the splatter patterns and the deposits, where was the breach, where was the origin this whole thing started from. Or origins."

Complicating the picture is the ever present problem of determining when the observed damage occurred.

"A key question when we're looking at these pieces is with the damage we see, was it caused by something that happened on ascent, on descent, breakup or ground impact," Kovacs asked. "A lot of these pieces, you look at them and there's pine needles embedded in them. So clearly that did not happen in space and we're taking great care to understand the relationships of those issues."

All in all, he said, "there's a lot of jigsaw puzzling going on out there. It is like putting together a multi-thousand piece 3D jigsaw puzzle on a 2D surface."

So far, some 54,000 pounds of shuttle wreckage have been recovered, or about 24 percent of Columbia's 223,000-pound dry weight. That total represents 45,762 different parts or pieces of debris, of which just 1,406 have been laid out on a grid mapping the shuttle's lower surface. Dozens of RCC fragments have been found, but most have not yet been precisely identified as to their original location. Almost nothing from the RCC panel 6 location has been recovered, but sections of adjacent panels, along with RCC support hardware, are under analysis.

Engineers believe between 35 percent and 50 percent of the shuttle actually made it to the ground. The rest presumably burned up in the atmosphere following breakup, including virtually all of the aluminum skin structure of the wings.

"A lot of the parts have come through an aluminum molten rain cloud if you will, where they have aluminum splatter over much of the debris," NASA test director Steve Altemus said of the wreckage recovered to date. "We see that molten aluminum on almost everything we have back. It may also speak to why we don't have upper wing surface on the left wing in that that's very thin aluminum. So yes, we expect a great deal of the aluminum not to have made it to the ground."

Asked if anything unusual leaps out under casual inspection, Altemus said "what did strike me as odd, first of all, is the size of the pieces, how small they actually are, and also there's very little left wing lower surface structure and very little left wing upper surface structure. When you walk the grid, hat's what you notice, what is not present as opposed to what is there."

In any case, with spring blooming fast approaching, NASA expects to conclude its massive ground search in the next four to six weeks.

"As the spring comes on, the undergrowth begins to leaf out and the canopy begins to cover," said Michael Rudolphi, deputy director of NASA's Stennis Space Center. "Obviously, the first area you'll lose search capability will be from the helicopter. And then on the ground, as the brambles and the briars begin to leaf out, it's going to damage our ability to see and to identify any material on the ground.

"The forest service thinks we've got another four to six weeks before that becomes real serious," he said. "But ... it is already starting to green up. And It will impair our ability to find as much as we would like to find. There is a time element to do the thoroughness we would like to do. Obviously, we can keep walking around in the woods after it's greened up and we can keep looking, we just won't be as successful."

Shuttle debris fell in a 4- to 5-mile-wide footprint stretching some 250 miles across central Texas and western Louisiana. The debris field impacts some 169 counties in Texas and 39 in Louisiana. Of that total, 143 Texas counties have been searched, 10 counties are in the process of being searched and 26 remain to be searched. Thirty one Louisiana counties have been cleared with eight left to go.

While some 1,800 potential targets have been found on lake bottoms in the debris field using sonar, no major pieces of wreckage have yet been recovered by divers. And no debris has been found west of Texas.

Two flatbed truckloads of debris, each one carrying about 4,000 pounds of wreckage, arrive at the Kennedy Space Center each week. It is then examined, identified and bar coded and analyzed. A two-dimensional reconstruction of the shuttle's lower surface is being laid out on a grid in a 40,000-square-foot hangar near the shuttle landing facility.

Analysts are in the process of developing a computerized three-dimensional virtual reconstruction that will show recovered debris modeled on an imaginary orbiter much like aircraft wreckage is sometimes reassembled to aid visualization and to map out the actual locations of various pieces of debris.

"There are a couple of hundred small pieces of RCC panels," Tanner said. "This would be like putting together a solid gray puzzle. What complicates this is in some cases the fractures that would mate normally really well have been eroded away. So it is going to be a time consuming process of putting this together."

Just how long it might take is anyone's guess at this point.

"We're doing our very best. But this is the first time in human history anyone's had to analyze a hypersonic space re-entry disintegration of a craft," Kovacs said. "And so there's no one to ask how long will it take. And there's nothing out there on the floor I look at and go, well, it's going to be another month and we're out of here."

Even so, he said, "I'm very optimistic. I think we're going to be able to figure it out. We feel pretty confident."

Video coverage for subscribers only:
   VIDEO: NASA OFFICIALS TESTIFY ON DEBRIS COLLECTION QT
   VIDEO: EXPERTS DISCUSS DEBRIS ANALYSIS, RECONSTRUCTION QT
   VIDEO: COLUMBIA ACCIDENT BOARD NEWS CONFERENCE QT
   SUBSCRIBE NOW