Columbia Accident Investigation Board
Home Board Charter Board Members News Photos Events Contact The CAIB
Columbia Accident Investigation Board Press Briefing
Tuesday, April 15, 2003

1:00 p.m.
Center for Advanced Space Studies
Lunar and Planetary Institute
600 Bay Area Boulevard
Clear Lake, Texas

LT. COL. WOODYARD: Good afternoon and welcome to the Column Accident Investigation Board weekly press briefing. We'll begin our press briefing today, as we always do, with opening remarks from the Chairman, Admiral Gehman.

ADM. GEHMAN: Good afternoon. I have just three or four short announcements to make. The next public hearing will be next Wednesday, the 23rd, and we have other ones after that scheduled, which we'll announce later.

The board this week got briefed on and approved a plan to terminate the debris searching. The debris searching, as you probably are aware, is probably about 80 percent complete of the ground-walking of the primary search area. It expects to be finished by the 30th of April, and we have agreed with that. There are a couple of individual areas which will be searched after the 30th of April on an individual basis, looking for specific pieces of debris; but the massive effort by between 5 and 6 thousand people a day will be completed by about the 30th of April.

The board is extremely grateful to the many organizations that have stuck with this now for ten weeks. The Texas Forest Service, the U.S. Forest Service, FEMA, and local agencies that have helped us have been enormously helpful; and the board is very grateful. We continue to find interesting pieces of debris that help the investigation. So this is very important to us.

Just some interesting facts. As I say, we have searched about 80 percent of the ground search area that we intended to search; and we expect to be nearly at 100 percent of the primary search area when we finish on the 30th of April. We picked up about 70,000 pieces of debris, weighing about 78,000 pounds, or 36 point something percent of the orbiter's dry weight.

In the area of debris, we have also asked NASA -- and the work is well under way -- to create a three-dimensional computer-assisted design, a CAD reconstruction of the left wing. They are scanning the pieces with a very sophisticated imagery scanning device, several sophisticated imagery scanning devices, and then that scanned representation of the damaged debris is being placed on a three-dimensional model of the left wing. It is our hope that this added reconstruction effort will help us better understand what happened.

The board has received two NASA fault-tree closeout briefings, one of which we briefed last week. We received the second one this week. It will be spoken to by my colleagues here a little bit more, and we have agreed with the closeout of two of the fault trees. We had some comments about them; but, generally speaking, we've agreed with the closeout of those fault trees.

Group No. 4, which you don't see represented here, a group led by Dr. Logsdon, is now up and running. They have a staff, and they have proposed to me several outlines of what they're going to cover. I think that you all understand from our previous press conferences, in general, what that group is going to do is they're going to put this accident and this accident investigation in several contexts -- in the context of the history of manned space flight, in the context of previous investigations and reviews, in the context of budget pressures, changing administrations and changing priorities, the American public's expectations of the space flight program and space exploration, and other contexts like that. That's quite a task, and they're now up and running and working on that.

That's all I have. In accordance with our normal procedures, what we'll do is I'll introduce the panel members here. Each of them has an opening statement, and we will be glad to take your questions. We'll go with Group No. 1, which is working on the materiel and maintenance issues.

ADM. TURCOTTE: Good morning. I'd like to talk to you today a little bit about one of our major projects. One, follow the foam, as the admiral is fond of saying. Two, we'll talk about a lot of what you heard recently or have seen recently in the press about the RCC, the reinforced carbon-carbon which makes up the LESS, the leading edge of the shuttles. We'll talk a little bit about some of the radar tests that are ongoing at Wright Pat and then talk about one of our issues we closed out this week.

First of all, General Deal, my counterpart, is at Michoud. I'm not sure how he got Michoud and I got Houston, but he's enjoying himself there looking at this week. I'm going to show you a little video here shortly of the dissection of the famous bipod tank. I'll show it one more time. I think this thing has probably got more TV time than any of us all put together. This is what it looks like, beginning to end. It's about 2 1/2 pounds or so.

Next slide, please. First off, here is ET 94. This is the sister tank of the ET 93 which was on the STS 107. The importance of this is they were basically identical twins. I talked to you a couple of weeks ago about dissection of the bipod ramp on ET 120, of which we found a few voids and some anomalies. This has been impounded by the board since the very beginning. We released it this week on Thursday; and it's being rolled out from its impound hangar, where we took it away from everybody to look or touch, and put it in the inspection bay.

Next slide. It kind of looks a little bit like open-heart surgery, and pretty close. You could probably assimilate that pretty closely. A very delicate, tender process of putting this and taking this apart and dissecting. A lot of science goes into this and a lot of time.

Next. If you could play the movie. Stay with me here. We sped it up to keep the process going. Here's a process. That's a pull type plug. It takes a circular hole and then pulls it out for later analysis. We carefully dissected. It's going a little quickly but it shows you exactly the process, layer by layer by layer, looking at detections and looking for minor flaws. There's an area.

There's also a graphic display that we use to look for anomalies. We found about six anomalies on the first bipod so far, being the non-destructive. This is an analysis to look for other debris that may have been in there. That's a surface.

The next is a scanning electron microscope, which looks, again, to find out if we've got any foreign debris which should not be there and what the exact surface looks like.

We'll stop there. Okay.

Next slide. There's an example from ET 120 of a void, a small void that we found; and on the right we can see the porosity at the microscopic level of what the foam, the cellular structure of the foam looks like.

Next slide, please. Another issue that we've been looking at very closely is the stringer valley cracks that were found at ET 91 and subsequent versions. This ET 90 was the last regular tank. ET 91 was the first super-lightweight tank that we flew. Prior to that, we had very few stringer cracks. This is an anomaly we're looking at. We don't know the full ramifications of it. The next slide will show you what we found since then, from ET 91 and on.

If you look, the first anomaly here is ET 91. Those are the amount of cracks that we have found to date on the super-lightweight tanks. So again, we don't know what this is telling us. We do know it is an anomaly. Those are cracks that leave open a number of scenarios, from gas creeping in, et cetera, et cetera. Again, this is just another foam anomaly that we're looking at very heavily.

Next. The RCC. As you know, this was designed primarily to bring the shuttle back and forth about 100 times. As it's aged, that has changed slightly. Some are anywhere down from 30 to 40 flights all the way up to 50 flights, just depending on the location of where it is. As this craft has flown or the several craft have flown, we found a lot of anomalies. These wear.

We found some cracks based upon pinholes that we suggest were from oxidation that was found due to the zinc contaminants on the shuttle platform. Also, elongation of a hole. They're called craze cracks; and as the silicon carbide expands and contracts, a small crack appears.

Here's an example of one we found on Discovery. Next slide. Again, much has been written about this; and I wanted to explain it a little further. Here's another one that was found on Panel 10L. On the left it shows it's pretty small, but on the right is a blown-up version. This is, again, the oxidation and through -- and our worry here is the heat is designed to be carried by the silicon carbide outside coating. Our worry is that the load is not being carried appropriately or as healthy as it would on the subsurface carbon substrate which is designed to carry the load.

Next slide. Here's an example. This came off of Atlantis, I believe, of a pinhole. That was shown.

Then here's the example. There's the pinhole going down through, and there's an example of the subsurface oxidation at a highly magnified level. As you can see, we're not sure exactly what that is telling us. The panels were previously inspected visually, and the admiral is going to be recommending this week that we change that process and move forward a process to get a better non-destructive evaluation program forward to figure out exactly what we have underneath the shuttle.

Next. Lastly, some of the research that's going on at Wright Pat. We submitted 26 articles to a 433-hertz radar. This is an example of a T seal. This is the last article submitted. Out of those 26 articles that we submitted for UHF testing, 14 did not have high enough radar cross-section. In other words, what we observed in flight did not meet the threshold, so it could not have been. So we eliminated those 14. Out of the 11 remaining, 10, if you look at the area-to-mass ratio or the ballistics, if you run it backwards, it did not match the radar cross-section.

So we're down to this last article, but that's not the last. The last article that they have is this T seal, which they're going to be looking at this week for further evaluation, to see if it meets both the ballistic coefficient, the mass-to-area ratio, and it matches what we saw in the radar cross-section. More to follow on that. There are going to be other articles submitted, but that's the results of the 26 that have been submitted to the UHF inspector.

Last but not least, this week, early on, we fanned out in many directions. One of the directions was taking a look at Morton Thiokol and the assembly and the management and the programs associated with the reusable solid rocket motor. This week we received a brief from NASA, and their recommendation was to close this out as a causal factor in this mishap. We concurred with their recommendation, with a couple of minor suggestions -- one of them being that we wait until we get all of the OEX data back to make sure that anything anomalous did not come up.

ADM. GEHMAN: Thank you very much.

Group 2, which is looking at the operational issues. General Ken Hess.

GEN. HESS: Good afternoon. This last week and going on into the weeks in the future, Group 2 will continue to take a look at decision processes and decision-making as they relate to several of the subitems we've talked about before, particularly looking at E-mail and requests for DOD imagery as well as taking a look at the disposition of foam events. Dr. Sally Ride is leading our effort, takes a look at the E-mails and requests for imagery and the decisions that surrounded that. My other partner, Steve Wallace, is taking a look at the disposition of foam events, the decision-making level that involves the PRCB and the MMT once the missions are on orbit.

This week we have finished our work, with the exception of just a few minor touchup matters, into things that were involving the medical status of the crew, their training and preparation for the mission, taking a look at payload and its integration onto the shuttle as well as its operation once we're on orbit and also the aspect of crew performance on orbit; and we've eliminated any causal factors in those particular examinations.

Coming up in the near future to match our efforts in decision-making and decision processes, we're going to begin to take a look seriously at NASA's safety program; and one of the inaugural efforts of that will be on the 28th of April. We're going to have a safety seminar day, if you will, for all the members of the board, where we're going to take a look at safety organizations, management, culture and theories. We're going to do a seminar piece on risk analysis and risk management. We're going to take a look at accident investigation models that have been out in the public domain and take a look at safety metrics, as well as looking specifically at leading and lagging indicators and how you decide what the metric should be so that you can make the decisions from them.

To help us in this effort, we have invited a panel of people from industry and academia to help us. We'll have Dr. Sam Mannan from Texas A&M University. He is the director of the Mary Kay O'Connor Center for Process Safety there. Deb Grubbe from Dupont Safety and Environmental Health. James Wick from the Intel Corporation. Dr. Nancy Leveson from MIT will be part of the panel. Dr. David Woods from Ohio State University, as well as Mr. Allen McMillan, who is the president of the National Safety Council, will be here. We will spend the better part of one afternoon, Sunday, talking to them about our investigation and all day on a Monday going through the various subjects that we talked about here so that the board can raise its awareness of all these issues. That should be a very profitable opportunity for us all.

ADM. GEHMAN: Group 3, which is doing the engineering and technical analysis. Scott Hubbard.

MR. HUBBARD: Thank you. Good afternoon. The other members of Group 3 are off pursuing their particular lines of investigation. Jim Hallock is working on sensor data and the fault-tree closeouts that were described early. Doug Osheroff has been looking into the evaluation of chemical analysis for the debris samples that we're taking, also looking at foam physics. Roger Tetrault is at Kennedy Space Center today, reviewing the latest items that have come in on the debris floor. Sheila Widnall is here in Houston, looking at the aerothermodynamic data.

What I'd like to do this afternoon is to give you a snapshot in time of our integration approach.

If we go to the first slide, what I want to show is how we are combining data from three different streams of investigation, the temperature data, the other sensor data, the visual data from the visual record, and what the debris is telling us, in an effort to understand where did the heat enter the wing. The challenge that we've discovered is to try to integrate and reconcile differences between these different data streams.

Also on this figure, you see how we're approaching the impact testing that will be occurring over the next six weeks or so. I'll come back to that later. In order to understand what happened when the foam hit the left wing, we need to come up with the best possible impact tests. So we're looking at this data to help inform and structure that test. As you'll see in a moment, currently the data are not all pointing in the same direction. So part of what we're having to do is to be as flexible as we can in structuring the test article to reflect how in perhaps a month or so these different data streams will come together. Out of this will come the test results which ultimately will contribute to establishing the most probable cause.

Now I'll go through the three different data tracks. Next slide. You recall a few weeks ago the calculation that showed the footprint of the impact region in the leading edge, the so-called RCC panels, the most likely strike zone around Panel 6. This analysis has been reviewed and continues to be reviewed, based on work done by five different NASA centers, other agencies, as well as industrial contractors.

New analysis, which is shown in the next slide, has moved the possible strike zone outboard a bit from 5, 6, 7 to Panels 7 and 8. What they did was to go and look at all of the possible data points from the visual record -- this was all visual analysis -- and add in a few more points to the trajectory. These are observations that they've been able to make now by looking very carefully at the tumbling, falling object as it moves down and strikes the wing, and draw a new trajectory line through that. In so doing, the impact point has moved slightly, about so far, from Panels 5, 6, 7, with the center of impact at 6, out to Panels 7 and 8, as you can see here.

Now, this is probably not the end of the story. In looking very carefully at what's been done with the image analysis team at Johnson Space Center that is leading this group of multiple teams, it's clear that there is more work to be done in evaluating the errors associated with this analysis. So it's not going to be probably until the end of April. There is a consensus among the image analysis team -- that's all these different groups -- about the most probable impact point and the error associated with that. So stay tuned as this all evolves.

The next line of investigation and next slide is a piece of data from the OEX, the orbiter experiment box that was recovered. I'm happy to say that we now have readout from 622 channels out of the 721. So given that at least 50 or so of these channels are not expected to yield data from the beginning, we're getting data -- that is to say, reduced bits and bytes -- of quite a few of the initial data channels. NASA has not yet converted all of these pieces of information into understanding and analysis. There's a major effort going on today and tomorrow to do exactly that. So I would expect that within a week or so we will have an even greater understanding of what that box is telling us. Nevertheless, the point to be made here from one of the early pieces of analysis, from two sensors, one just behind Panel 9 and one at the Panel 9-10 interface is that a substantial heating event occurred at about 500 seconds after the orbiter entered the atmosphere and then the sensors went off line low, as if the wires had been cut or perhaps burned through.

This data, then, says something interesting or disastrous happened at the 9-10 or Panel 9 interface, further outboard from what we showed you from the visual data record. Now, does this say it occurred here? Not necessarily. It could be just this is the place, since this is the only sensor of this type in the wing leading edge, where it was detected; but it's another piece of data and more to come.

The third line of investigation and the third line of data is on the next slide, which is what we're learning from the debris field, what the debris is telling us. Data is emerging which shows a very significant and unusual damage at the interface between Panels 8 and 9. For example, the splattered metal on Panel 8 is much heavier than elsewhere and there is erosion of the reinforced carbon-carbon elements, the ribs, what's called the lug, where the attachment occurs. That's down in these regions here. This is what a normal RCC panel looks like. This is after it's gone through this heating event. This is the transition event here.

This is not what it looks like. This is, I think, a software problem. Pay no attention to that one. Instead let's go to the next slide.

Go to the next slide. What we see is something that's not seen thus far anywhere else on the wing leading edge or, indeed, in the orbital debris, which is pieces of a very tough material, this reinforced carbon-carbon, eroded to knife edges. Where a normal piece is a half an inch, it's been eroded to about the thickness of a dime. This kind of heating event indicates long duration, very extreme heating.

For example, by comparison, this is what Panels 1 and 2 look like. You can see the half-inch thickness there, and you can see how this has been worn away in this area. We don't know quite what to make of this yet other than what I said, a very severe heating event in the intersection between Panels 8 and 9.

The carrier panel -- that is, the piece of material that goes between the reinforced carbon-carbon and all the tiles on the bottom of the orbiter -- that carrier panel also shows severe heating. It's slumped, like you overheated styrofoam. That's another indication of some very severe heating events.

Next slide. What do we do with this? We have lines of evidence, if you recall, that began with some type of an impact that may have occurred in Panels 5, 6, 7, which new analysis has moved that to Panels 7, 8. We have temperature data indicating a heating event in Panel 9 and 10, and we have the debris speaking to us and saying something severe happened in the interface between Panel 8 and 9.

As was said earlier, the data collection and the debris collection is going to go on until the end of the month. The data analysis from the visual record will probably continue to the end of the month; and other forms of analysis like the OEX box will go on for weeks, as well. So this story of exactly what happened and where it happened, while we're closing in on it, is going to continue to evolve. Meanwhile, we have to prepare to test the impact of the foam to understand what could have happened when the foam hit the leading edge, was it the source of this breach.

We don't have the exact position nailed down yet. We thought, going over to the slide here, that originally it was around here. It could be across here. Maybe it's even out here. Nevertheless, this region seems to bound it.

From the beginning of a go-ahead to building up an article like this takes about five weeks. If we wait until all of the data is in and then build this, we'll be doing testing in July and August; and that's not in anybody's interest, with people on orbit and with all the other needs of the program. So what we've done is to put together what you see here, which is a test structure. This has now been agreed to between NASA and the board, the test structure that will give us maximum flexibility when we begin the testing of the reinforced carbon-carbon leading edge, which is our principal target and where we think the breach occurred.

Other testing is already under way to calibrate the gun at Southwest Research Institute that we're going to use; and because six weeks ago we thought that the impact region might have been in the door -- that is to say, the main landing gear door -- that piece will be ready by the end of April, around April 25th or so. So they'll start doing tile testing at that point. However, as soon as this piece is ready, which will be around mid-May, all that other testing will stop and testing on this will begin.

So at this point, materials have been purchased. Fabrication of this unit is moving ahead. We estimate the first impact tests on this, wherever it turns out is the most likely region or regions, we could take several shots at this without contaminating the results. Those tests will probably take place in late May and will continue for several weeks.

That concludes my briefing.

ADM. GEHMAN: Thank you very much.

LT. COL. WOODYARD: Okay. We'll begin questions here in Houston. We'll start right at the back. Please identify your newspaper, as well.

A REPORTER: Houston Chronicle. My question's for Mr. Hubbard. Have you decided how you're going to characterize your test articles and whether they're going to comprise new and flown components, and how are you going to characterize them after each test so that you can sort of subtract what damage has actually occurred and fold that into your investigation?

MR. HUBBARD: Okay. Good question. First of all, how will we characterize the test articles? There is a whole matrix of non-destructive evaluation that has been proposed by NASA and tentatively agreed to by Group 3, although I hasten to add we don't have a completed, signed-off test plan yet. Those non-destructive evaluation methods will be used before the article ever sees any impact and then again after, to gauge how much damage has occurred. There will be things like X-ray studies and what's known as a CAT scan, computer-aided tomography, where millimeter by millimeter an X-ray is taken of each section of the leading edge and then all those are put together, much as you would do with a human patient, and other things like thermography and shearography. These are techniques to look inside of the panel and see what's going on in there.

Now, in terms of the test articles, we want to end up with articles that are as much like the leading edge of Columbia as possible. We've taken panels off of Discovery. We're looking at a complete inventory. We have agreed with NASA to look at the complete inventory of all of the spares that are available to match what I showed a few moments ago of the areas of 5 through 10. We will use those last, obviously.

One of the panels that is of interest -- both Panels 6 and then Panels 7 and 8 -- we've identified some that have about 30 flights on them. That would make them very comparable to Columbia. In between, though, before we get to that, we will first use fiberglass. Fiberglass has about the same toughness, the same stiffness as RCC, and we will also use some panels that have not seen flight but were taken off of Enterprise and leading up to the final article because, once you hit a flight panel with a pound or two of foam traveling at 500 miles an hour, it's not clear that you haven't contaminated any other test that you want to do. So we want to be sure that we lead up to the test and do it the right way.

A REPORTER: For Admiral Turcotte. Dallas News. We're still wondering if this carrier panel or a T seal, in the radar testing. Have you ruled out the carrier panel now?

ADM. TURCOTTE: The carrier panel is one of the devices that was ruled out, yes.

A REPORTER: It's ruled out.

ADM. TURCOTTE: In the form that it was submitted. We haven't gone through in its entirety. It's gone in different sections, with and without metal. It's not completely gone; but in the form that was submitted, it is gone.

A REPORTER: So now you're leaning toward what?

ADM. TURCOTTE: The RCC panel, which we're looking at very closely. The T seal. By the way, right here is the cross-section. What I showed you was a full T seal that we submitted. Later on, smaller pieces of this will be submitted.

A REPORTER: Just to follow up on that, can you explain what the process was that first made the carrier panel seem likely and then later --

ADM. TURCOTTE: Its shape, its location, its probability. That looked very closely like that; and I'm not sure which tests it was, either the ballistic test or the radar cross-section test, that had a fallout. I'll get back to you on that.

A REPORTER: Associated Press. For Mr. Hubbard. Now with the new analysis and the new zone, the strike zone, how does that compare to the possibility that a T seal did float away on Flight Day 2? I mean, how does this all work together, in your mind, of building this analysis that you're doing now?

MR. HUBBARD: Good question because integrating and synthesizing all of this, I think, is the key next step. Given that we have found pieces of all of the carrier panels in this region, it's looking less likely that that was a drift-away object, although it could have been an impact zone. The space between Panels 8 and 9 happens to line up with where the T seal is. So that has become a suspicious region; and in the preliminary planning for the impact tests, we're putting strong consideration on hitting not the middle of a panel but that intersection where the T seal is and seeing what happens, what kind of damage would be caused.

A REPORTER: Would it be quite easy, if a T seal is struck, for it to loosen enough to just pop out all by itself?

MR. HUBBARD: We're looking at whether the material's properties would lend itself to that. There's a couple of things you can imagine. You can imagine the T seal actually being broken. We're finding a lot of the pieces of the tops of RCCs but not the bottoms of the panel. A number of them seem to be broken in the middle. So breaking is a possibility. Knocking the attach points loose is another possibility. Before we do any of these tests, we will be doing, first of all, qualitative analysis to see if the strike could have broken a bolt and then caused it to come loose. That's part of the test planning.

ADM. GEHMAN: The T seals are attached to the wing spar independent from the RCC panels. They have two bolts at each end, and it's possible to damage or break off a T seal and not break off an RCC panel. I think that's what you were getting at.

MR. HUBBARD: Yes. Exactly.

A REPORTER: ABC News. Where will you keep searching for debris; and will resume a debris search, say, next fall? What are your primary debris areas that you will still search?

ADM. GEHMAN: As I indicated in my opening remarks, the search teams are going to complete the search area that NASA laid out; and this is the primary zone. When the board was briefed on their plan, we made a minor modification to their plan to search west of Interstate 45, essentially Corsicana, Texas, towards Interstate 35 East. The amount of debris found out there is less than we are finding in the primary zone. Therefore, if you were measuring it by how much debris you were picking up per hour or per acre, it would come out to be less debris; but it's also more interesting because it's all left wing debris. It's all left wing RCC and all left wing tiles. Exclusively left wing debris.

So we are obviously much more interested in left wing debris, even though the search per man per day is a little bit frustrating because you're not finding very much of it. So what we asked them to do and they are doing right now is extending their search perhaps about 40 miles, 30 miles farther west than they had briefed to us. There are four other particular areas that are relatively large areas in which we have a pretty good radar-tracking data on debris which fell off the orbiter in Nevada, Utah, and Western Texas; and these areas will be also searched after the big, massive search ends. They've already been searched to some degree by helicopters and some small teams and things like that. One of them was, up until recently, covered with snow; so we obviously couldn't see anything in there. The trajectory analysis of these pieces continues to this day with the help of the FAA and the NTSB, who do this very well. We'd like to narrow down the search areas before we send people back out again. So that kind of summarizes the additional areas other than the major footprint that goes from the Louisiana border to Corsicana, Texas.

A REPORTER: What do all those pieces, the left wing pieces tell you? What would flow from that?

ADM. GEHMAN: Actually, it's too early for me to make a judgment about that; but in rough order merit, not much. I mean, we already know that the damage was initiated in the left wing and we know the left wing came apart before the right wing. This is just confirming that; but as you have heard from these other of my colleagues here, every little piece of RCC or every little piece of tile contributes to the story and we felt that if there's good left wing debris out there, we can't leave it out there. Now, as the spring and summer goes on and people are out in these areas, as farmers work in their fields, as hunters and backpackers and vacationers and picnickers are out in these areas, we anticipate that additional debris will be picked up.

When we close down the level of effort, what I call the level-of-effort search -- this is the 6,000 people a day out there -- we'll move the headquarters of the search organization from Lufkin, Texas, to here in JSC. The headquarters will remain open every day. There will be telephone numbers and web sites and places where you can call in if you find things. If they do find things, we have teams that are ready to race out there and pick them up and properly identify them. So the picking-up of debris will go on all summer.

A REPORTER: New York Times. Admiral, you've spoken in the past about the possibility -- or Mr. Hubbard -- of Inconel bolts having a deterioration problem in salt air and we're now looking at possible deterioration of a bolt or some sort of loosening. Is this still an open issue, or have y'all pushed it aside?

ADM. GEHMAN: It is still an open issue, and the test articles that Scott described include not only the RCC panels but the attachment hardware also. In addition to that, the attachment hardware is going to be looked at very carefully in separate tests.

MR. HUBBARD: In fact, if they can go back to my slide that shows that, I'll talk a little bit about it; but, in essence, we are looking at the possibility here of not only something breaking in the RCC or in the panel or the T seal, we're looking at all the structure underneath it. So the attempt will be, as NASA builds this up, to duplicate where the bolts are, the same kind of bolts, and all this will be sitting on what's called load cells; and they'll measure how much force was put on there and even to the point of saying could that force, if it didn't break a T seal, have deformed something, actually pushed it out of shape so that you open up a crack, a gap between these things. As part of doing this, all the material properties are being examined. The bolts that hold on the carrier panels, for example, there's only two of those; and they're quarter 20. The neck of those is very small. So all of this is being looked at and will be considered in the testing.

A REPORTER: Florida Today. For Admiral Turcotte. How would you characterize NASA's ability to gauge damage on RCC panels right now with their current inspection processes; and in your recommendations will you be recommending specific types of additional testing and, if so, what technology is available to give them better insight into maybe damage that's lying below the surface, if you would?

ADM. TURCOTTE: Well, right now it's more what we don't know. The testing is a visual. It is done based upon looking at a 10X magnifier. If it fails the visual inspection, then it goes back for refurbishment. Beyond that, the issue is we don't know what we don't know. That's the issue. NASA has been, in the last several years, pushing forward, looking at different technologies to get better at that. Based upon our discussions, they are working up a series of recommendations to define that. And they include CT. They include shearography. They include thermography. They include line thermography. Eddie current. The massive hand-held Eddie current or big-plate Eddie current. There's several good technologies out there that can be adapted from other industry uses. Ultrasound is another one that I threw in there. They are developing, in concert with several of the people on the board, a recommended set of tests to be applied to the rest of the orbiter fleet.

ADM. GEHMAN: But the answer to the second part of your question is that the board will not specify a specific test methodology. We will just specify that they must, through NDE processes, learn how to characterize the hidden conditions of the RCC so that they can measure ageing.

LT. COL. WOODYARD: Okay. We'll go to the phone bridge.

A REPORTER: Washington Times. With the area in question spreading out to include perhaps as far as Panel 10, does that say anything about the size of the breach, that it could be bigger than what was first thought?

MR. HUBBARD: No. What you're seeing is the result of different lines of data, not all at this point quite agreeing with each other, and with the ongoing effort to understand how much error there is in the measurement. I'll give you the visual as an example. I'm sure you've seen the film and the videotape that shows this tumbling, twirling object. When they freeze that frame by frame, there's a lot of blur in that image. So they're using some fairly sophisticated techniques to figure out what the center of that is; but given the fact that there is a lot of atmosphere going by and that it's difficult to pin down exactly the edge of a blurry object, you're just seeing the ongoing process of sorting out the errors. I would expect that in a month or by the end of the month or so, the visual record will hone in on a recommended footprint. What we have to reconcile -- what NASA has to reconcile and then what the board has to reconcile is if the debris is saying there's possibly an event in the T seal area between 8 and 9 and the visual record is saying maybe it's at 7, 8, then what we have to do is figure out you've got a strike at 7, 8, could it have affected something further downstream. And there's mechanical engineering calculations going on to look at these possibilities.

A REPORTER: NBC. I'd like to ask the board. With the deterioration, with the age of the RCC panels and with the impact point being moved farther out that you're looking at now with an accident and considering that the shuttle went through seven more minutes of shattering and twisting to get into orbit after the foam impacted, could it just be that the foam impact had nothing to do with it all and that you're really looking at age here?

ADM. GEHMAN: The answer is absolutely yes. We are examining this foam impact in extraordinary detail, and that's what the Southwest Research testing is all about, to determine whether or not 2 pounds of foam at 500 miles per hour even has the possibility of damaging the leading edge system. There is a possibility that the foam had nothing to do with it. Absolutely. We remain open to all scenarios, but we have to run everything down because this is obviously an anomalous event that has to be explained. So we have to run it down to earth. But the board is keeping an open mind on this and one of the things the board is going to have to struggle with in about a month or six weeks from now is what language do we want to use to describe the relationship between the foam impact and the loss of the orbiter as far as our degree of confidence that we know what this is. And we haven't even discussed that yet.

A REPORTER: Newsday. Can somebody talk a little bit more about the size of the T seal, what it's made of, and the size of the gap that might be left if one of them were to be lost or pushed aside?

ADM. TURCOTTE: Yes, I can. First off, you can't see it in front of you; but essentially we're looking at a T seal. It's about 7 inches wide. Depending on the location, there are 22 RCC panels; and in between each of those is a T seal. So depending on where it is on the wing, that says its size. Its size could be, roughly guessing, as high as 3 feet, down to less than a foot and a half or maybe a little bit more, in a bowed shape, the shape of a U. Essentially where the U fits on the wing and that's the shape of the wing, that's the shape of the T seal. Each of these T seals are custom fit and made for its neighboring RCC components. When you replace one, you replace the other. They're custom cut, custom made, and custom fitted to each part on the wing. The makeup and the composition is essentially, it's made of a carbon fiber --

ADM. GEHMAN: Same as the RCC?

ADM. TURCOTTE: Yes, it is. It's a reinforced carbon fiber. It's the same material as the actual RCC panels, and it's made and then heat-treated several times and the outer surface is treated and becomes a carbon silicate. That's what carries the heat load. In the flow, that's what carries the heat load. The center of the T seal is essentially the load-carrying part. That's the hard part, and that's the carbon substrate. That's what really carries the dynamic pressure and carries the structural integrity of the T seal.

ADM. GEHMAN: I might add that it's called a T seal, of course, because when you cross-section it, it's shaped like the letter T. The people in the room here can see that the two flanges are not the same. This flange here has a little lip on it. The other one is wider and is flat. When you install these things, they kind of lock into the RCC on one side and they're free to float on the other side. The distance over here on this side between the RCC and this flange here is, I think Roger said, about 1.1 --

MR. HUBBARD: It's about an inch.

ADM. GEHMAN: An inch and an eighth gap in there. So if you were then to project this onto the front of the wing and you were to lose, for example, a piece of this side over here, you now have a 1 1/8-inch gap directly into the wing, multiplied by whatever length of whatever you lost. So the question is: What kind of hole would you have? If you were to break off this little piece right here, for example -- let's say you would break it off by the size of this. You would now have a 1 1/8-inch hole, 8 inches long, going directly into the wing, with nothing to stop it. Now you see what the problem is.

A REPORTER: CBS News. I had a couple for Scott Hubbard. I was wondering first of all -- or maybe this is Steve Turcotte -- if you've ruled out an RCC panel fragment. In other words, if you broke off a part of an RCC panel, does the radar testing at Wright Pat rule that out as a possibility?

For Scott, I was just wondering if there was any explanation yet for how a T seal could get broken on ascent, if we assume that's where this happened, and then stay in place for a day before it separates. I didn't know if there were any thoughts or more mature thoughts about that.

My final question for you, as I'm squeezing in three for one here, I've heard a little bit of talk about some data on ascent from the OEX recorder that shows a little bit of temperature and a possible anomaly behind RCC Panel 9 on ascent. I was wondering if you could tell us anything about that.

ADM. TURCOTTE: I'll take the first part. The radar. We have not ruled out a part. Right now we have a full T seal undergoing analysis. When that's complete, we're going to submit bits and pieces. We may even use some flight hardware to do that. But we have not ruled it out. More to follow on that.

ADM. GEHMAN: Same with the RCC panel. The first test was for the full panel, and now we're going to send them a fragment.

ADM. TURCOTTE: That's correct.

MR. HUBBARD: Okay. I'll take the other two, then. So the first question was: If you hit a T seal on ascent, could it have stayed in place all the way to orbit and then perhaps drifted away on the second day? I think the thinking on that is a little bit more mature in terms of understanding that a T seal could have been the impact zone. The analysis hasn't been done yet, though, on the strength of materials to see if the forces involved in the maneuvers that the orbiter incurred could have shaken it loose, so to speak. So I can't give you a definitive answer on that one, but it is in work.

Then finally the anomaly on ascent. The data from the OEX box is being looked at very carefully not only for entry but also ascent. There are a few anomalies there, but I don't think anybody's prepared to say that they are real. Anytime you have sensor data, particularly if you're measuring it at the very low end of its capability, you can get glitches and blibbets that don't have anything to do with real measurements. And on ascent, the sensors, some of it -- the pressure sensors and strain sensors in particular are not operating in their middle of their range; they're way over at one side. In terms of temperature sensors, I don't know of any ascent anomaly. So I would say let the analysts do their job in the next couple of days and we should see something out of that, I would say, in a week or so.

A REPORTER: Washington Post. Could you describe for me, please, your working theory or whatever analysis you have on the possible reason the RCC panels broke into upper and lower.

MR. HUBBARD: Well, I guess I'll take a shot at that one. I don't know that we have at this point anything other than an observation. You said: What is the debris trying to tell us? In this case until we complete the search, we can't be sure that this observation is a really true one and then, secondly, if it is, if we don't see any evidence within the debris footprint of the parts, the lower parts of the RCC panels, then that's telling us something, but I don't think that we're mature enough in the analysis to say what the precipitating mechanism might have been. I mean, clearly if you have a bent structure, that point there, if you're applying like a wishbone stress to it, that's going to be where the stresses meet. Whether that is a scenario based on physics and engineering is yet to be seen.

ADM. GEHMAN: I would agree with that analysis, and I would add to it that it's not clear that what we saw in the breakup of all these panels may be late in the event and therefore not part of the cause. I mean, it could have been just the way they failed and have nothing to do with the initiation of this event. So stay tuned.

A REPORTER: Aviation Week. A couple of questions for Admiral Turcotte. First, Admiral, you raised stringer cracks, at the start of the presentation. Could you take a little more time and describe how large they are and zero in a little more on their location as you've seen them at Michoud.

Second question follows up on a point that was raised by General Deal back in mid-March. General Deal said back then that the board was looking into reports out of NASA about the potential for water to have perhaps pooled in the leading edge area as Columbia sat, getting ready to go into Palmdale, and was rained upon as it sat horizontally. Has any water issue remained on the table?

ADM. TURCOTTE: Well, first part, I'll talk about the stringer cracks and then the water issue with Palmdale.

If you could bring that slide back up, it would help me a little bit. The stringer valley cracks first started with ET 91, we think as a result of two things. It's the first time we machined, and it's also because it was a super-lightweight tank. You've got less density. In other words, you've got a thinner layer of foam. We don't know the full effects. We do know that the cracks were anywhere from 3 to 6 inches, of the pictures I've seen. There could have been others. Those are the ones I was exposed to. We don't know the full results of whether or not it caused shedding events. We do know that we had large-scale shedding events, but we do know that we did see the cracks and that's what we're looking at. It was not seen prior to that, at least for ten years; and then once we went to the new super-lightweight tank, they emerged. And more to follow on that.

Next. With the Palmdale. There have been several instances throughout the lifetime of many of the orbiters, as they're exposed to the elements, there are occasional times where they've had some water intrusion. Yes, we've looked very extensively at several areas, especially on the Columbia. Somewhere around the forward bulkhead, I believe it's the 582 bulkhead, which is the bulkhead just after the crew compartment, before you hit the shuttle bay, there is some corrosion that we found in there. There was also two reports of some water found at various other places in the shuttle -- it escapes me exactly where at this point -- but were investigated and cleaned and the water was extracted and the ensuing investigation revealed no damage.

This is an ageing aircraft, however; and if you look at all of the shuttles, there is corrosion on all of the shuttles in various forms. Columbia, being one of the oldest, did have some corrosion issues. That's part of what we're addressing in the total program of looking at these aircraft or the shuttles as an ageing spacecraft. So to answer your question, several instances of water and, yes, it was investigated.

LT. COL. WOODYARD: Okay. Final question from the phone bridge.

A REPORTER: Could you help me understand a little bit more, Scott Hubbard, about some of the failure cascade of events here. If I recall correctly, early on in the discussions there was a feeling that the breach would have to be fairly large -- 36 square inches sticks in my mind, but maybe even bigger. A missing T seal certainly doesn't get you up to that volume, yet you also see debris falling off over the Pacific et cetera. Could you help me understand what that debris might be, what your current thinking is about the dimensions of the breach in order to create the kind of heating effects you've seen?

LT. COL. WOODYARD: Could you tell me your organization, please.

A REPORTER: Sure. I'm sorry. National Public Radio.

MR. HUBBARD: What you're referring to is perhaps six weeks ago or even longer, there was a very simple initial analysis that said if you had a 20-square-inch hole -- that is to say, something 4 inches by 5 inches -- this would produce enough heat, if it was in the wheel well, to create the temperature increases that you see. The people that are dealing with this branch of physics and engineering, called aerothermodynamics, have gone well beyond that now and they're starting to look at a plume of heat that would have played around inside the wing, inside the wing structure possibly in somewhat an uncontrolled manner, like water coming out of your fire hose that no one was holding. I think that the jury is still out on how much of a breach you would have to have in order to create the effects that you see. In particular, the understanding or the feeling from the visual data that we may not be dealing with a round hole but instead something that created a long narrow slit is a different type of boundary condition. It's a different initial condition. So the thermodynamics people have gone back to the drawing board -- and they're doing that even as we speak -- to figure out what a long, narrow slit would do, as distinguished from a hole. That's a different set of boundary conditions.

LT. COL. WOODYARD: Thank you. This concludes the press briefing for the day. As always, the board members will remain here and answer a few of your questions. Thank you for coming today.

(End of conference, 2:02 p.m.)

Back to Event
Privacy Policy Disclaimer Accessibility