|Columbia Accident Investigation Board Press Briefing
Tuesday, May 6, 2003
3000 NASA Road 1
LT. COL. WOODYARD: Good afternoon. Welcome to the Columbia
Accident Investigation Board Press Briefing. As always, we'll
begin the press briefing with opening remarks from our chairman,
Admiral Hal Gehman.
ADM. GEHMAN: We're going to change the system a little
bit here. I'm going to speak last. So I'll go right down the
aisle here and ask for a brief introductory statement from
my colleagues on my all-Air-Force panel here today, including
GEN. DEAL: Thank you for the opportunity. I'm going
to give you a very quick update because the main event, of
course, will be Admiral Gehman when he talks to you.
Our panel. General Barry's in town. He's got a management
focus this week and working on pulling the entire board story
together. Admiral Turcotte was at Kennedy last week and will
be joining us again this week.
Our three subgroups. Our first one is Maintenance and Sustainment.
Everything they've been looking at, we've briefed you before.
But a very heightened focus now on contracts. Multiple items
from the Space Flight Operations Contract at Kennedy to the
external tank contract operations at the Michoud assembly
facility. They're also getting a lot more fidelity to the
matrix ability, looking at past reports, what was recommended
and also what has continued throughout NASA.
Our Management and Human Factors Group did a trip to Huntington
Beach last week, examining not just the engineering organization
there, the so-called brain drain, but also talking with the
engineers that work the Crater program, that predictive tool
for estimating the impact damage to the orbiter.
This week the team's also going to be bringing in two internationally-recognized
experts to talk organizational issues with us. Dr. Karlene
Roberts from UC Berkeley will be here to discuss high reliability
organizations and managing high-risk technology; and joining
us also will be Dr. Howard McCurdy of American University,
who's the author of four books on space policy. Then next
week we have Dr. Carl Weick from the University of Michigan
to continue discussion on HROs.
Our Materials and Structures Group, particularly Larry Butkus
and Lieutenant Commander Johnny Wolf, have been looking at
the external tank and its testing both at Marshall Space Flight
Facility, that I'll be showing you some slides on very shortly,
and also Michoud, and also been working with Group 3 on the
foam gun test parameters that Dr. Widnall will probably talk
Then another lingering focus for us is still quality assurance.
It remains a key area of examination. Yesterday, for example,
we held a half-day session dealing with safety and mission
assurance at Kennedy Space Center. As you have reported yourselves,
inspections are down across the entire shuttle program by
the quality assurance aspect, though most of those have been
shifted to the contractor; but we've got extensive interviews
and documentation that will reflect a likely need to update
and fine-tune the steps that are examined and also the methods
of examination. Simply said, some things that work great for
an airline industry that has thousands of flights per week
may not be right for a research-and-development manned space
flight system with 113 fights under its belt.
So I'll show you a few slides before I get off the stage.
Tom, could you bring up the next slide, please.
One of our favorite topics, of course, is the famous bipod
ramp; and you see that there with the minus Y and plus Y.
Next. We've finished, as you know, 120, the cousin tank. I've
already out-briefed that to you. Next week we'll have the
ET94 sister tank brief for you.
Next. These are the pictures of what is under test now, up
at Marshall Space Flight Center in Huntsville. These are some
panels that they've built with bipod ramps in order to try
to induce the cryopumping that we've talked about, and they're
going to be doing that inside that 12-foot vacuum chamber
Next. We're going to re-address and correct the record the
best that we know now with the known ET bipod ramp losses.
We are now counting six, with an addition of one of STS 52
there, with the enhanced photography that we have. You can
see when they all occurred. We talked about 7, 32, 50, 112;
and then what we discussed in detail this morning was 107.
Next. Just a snippet for you. If you look at 50 and 52, they
were separated by two flights in a four-month time period;
and then 112 and 107 separated by one flight in a three-month
time period. Those are things that Group 2 is looking at the
Flight Readiness Review to see how the disposition of those
foam events was handled.
Next. Just as a summary slide to you. 13 shuttle launches.
We've got photo verification of the bipod condition on 74
of those. We don't have it on 39, due to a number of factors.
It could have been a night launch and we had absolutely zero
photo confirmation, or it could have been the orientation,
such as happened on 107. By the time they were able to take
pictures of the external tank floating away, it had already
rolled over and you couldn't even see the bipod area.
Next. So we have six known foam losses. You could statistically
extrapolate -- you could never prove it -- but you can statistically
extrapolate three unseen losses, based upon those that you
have not seen. So you might conclude that you might expect
some degree of bipod loss every 12 to 13 flights, and how
this became an acceptable risk is one of the things that we're
So thank you very much.
GEN. HESS: I must hurry so I can cede the remainder
of my time to the chairman.
It's good to be here. I'll try to describe for you today the
process that we're all kind of going through as we get to
end game and how these various lines of investigation between
the groups are going to attach.
Steve Wallace is leading the portion of our investigation
that's going to take a look at the disposition of foam events,
primarily across the six missions that General Deal highlighted
to you on that previous chart. In doing that, he's going to
look at the Flight Readiness Review process as well as the
PRCB process that took decisions by the program as to whether
to fly succeeding missions and how they handled the foam particularly.
Dr. Sally Ride, as you know, is taking a look at various things
like the Crater analysis as well as the E-mail thing that
was going on with the STS 107. So she will connect the on-orbit
management decisions of that particular mission as they relate
to the work that Steve Wallace is conducting across the top
almost operational/strategic level of the program.
For my part, I'm taking the same walk; but I'm going down
the safety chain -- I'm saying program-level safety, not trips-and-falls
kind of thing -- to take a look at overall program and policy
as well as the risk management and risk integration on those
various steps and to see how NASA used trend analysis from
within the various data bases that track safety information
to help predict and work risk management problems.
Now, all these lines of investigation should intersect here
as we get to the end of our report and start doing our analysis
level of what all that means in terms of how NASA manages
these on-orbit decisions that they make.
DR. WIDNALL: My group, Group 3, has been following
up engineering and technical analysis, behavior of the shuttle
during flight, as well as the analysis of debris.
Just to mention what some of the group members are doing.
Roger Tetrault has been down at Kennedy; and he's, of course,
focusing on debris. He's also sitting in the back of the room.
He's been focusing on the chemical analysis of debris to tell
us what is it really telling us. He mentioned that the last
scheduled truckload of debris was unloaded at Kennedy Space
Center this morning. Clearly, the amount of debris that's
been collected on the shuttle is really unprecedented.
Jim Hallock is working on constructing various scenarios,
analyzing them in detail. He's also looking at the behavior
of sensors, trying to look at the ones that failed as well
as the ones that are giving us good data. He's our man to
watch over the closeout of the various fault tree analyses
that are being done.
Scott Hubbard is working on the foam tests. He's actually
in town, but he all last week was down at Southwest Research
Institute. He said that the foam impact tests began last Thursday.
The original tests, I think, are more to test out the test
setup, shooting small pieces of foam at an instrumented main
landing gear door; and then they'll move up towards changing
angles and moving up to larger pieces of foam and getting
data on tile damage due to foam impact. Obviously there's
a lot of focus on the testing of the leading-edge RCC. The
test setup should be ready around the end of May. It's a fairly
complicated test fixture, and he anticipates that the testing
on the RCC or configuration will begin in early June.
As for myself, I'm sort of monitoring and participating in
the aerothermal analysis of the various phases of the flight.
So that's what's going on.
ADM. GEHMAN: Thank you very much.
Now, for my part, I'm going to walk you through what the board
calls a working hypothesis or a working scenario. As we had
previously announced, the board felt that we now know enough
and we understand enough that it's time that we change the
board's official policy, which was we don't believe in scenarios
of the day or the most popular scenario, that everything was
on the table, to the point now where we believe that we should
focus our efforts; and that's what the purpose of developing
a working hypothesis is.
We still have our antennas out, our senses sharpened for any
indication whatsoever that we may have it wrong or we may
have overlooked something; but by building a working hypothesis,
as I will show you in my last couple of viewgraphs here, it
really enables us to focus the testing. It enables us to much
more narrowly direct the analysis. It really produces a very
synergistic effect on a lot of different disciplines so that
they now all point to the same direction.
So I'll go through what we believe we know and what we don't
know. A lot of stuff is not in this working hypothesis. This
is just a broad sketch or a broad scenario that encompasses
a scenario which is wide enough and broad enough to capture
most all of the data that we've developed.
So if I could have the first viewgraph, please.
First of all, this scenario is called a working scenario.
The reason we say it's a working scenario is that we reserve
the right to change any part of it at anytime without notice
because we have to have something to go on to conclude this
investigation but at the same time I don't want to suggest
that the board has voted on every one of these statements
and that we're willing to say that this is the end of it.
It certainly isn't.
I would also like to acknowledge that this was done in conjunction
with the whole investigation team. It's labeled as a Columbia
Accident Investigation Board working scenario, but it represents
the work of the entire team and I really want to thank the
NASA part of this team for their hard work on this. The reason
why we wanted to be sure that we had a complete understanding
with everybody in this investigation is because I didn't think
it would be helpful if there were other people out there proposing
various versions of this. So we believe that this is an agreed
position by most of the people that work on this.
The two members of the board who are most responsible for
the work on this are Dr. Jim Hallock, who is unfortunately
under the weather today, and Roger Tetrault and their group.
On the NASA part, Randy Stone and Fran Benz and Jim Kennedy
are the leaders of that group.
All right. Starting through here now. Based on the six items,
six lines that you see indicated there, based on the analysis
that we've done in every discipline, we have reached some
agreement on what happened in both the ascent, on-orbit, and
Next viewgraph, please. I'll just go through this in chronologic
order and tell you a little bit about how we came to this
conclusion. Obviously at seconds after launch -- 81.7 seconds,
to be exact -- we have ample evidence that several pieces
of debris came off of the left bipod ramp, one of which struck
the orbiter. You'll notice that someplace in the vicinity
between Panels 5 and 9. Later on in this presentation I will
talk about other evidence that suggests different panel numbers,
but we've just got to let the evidence speak for itself. You'll
notice that there's nothing in this statement which makes
any kind of a declaration about whether or not any damage
Next. We have just samples of some of the evidence that led
us to this conclusion. Here we have a still from some of the
video loop that you saw earlier this morning of some of the
foam coming off.
Next. We have been able to track the path of the foam, using
the two cameras that saw this; and we're fairly certain we
know the vicinity that the foam came from and we know the
vicinity under the wing that the foam struck.
Next. We've got this trajectory analysis down, even though
it continues to move around a little bit a couple of inches
left and right, but we're down to the area -- this analysis
narrows down the foam strike from what I said before, from
5 to 9, to in the vicinity of Panels 7 and 8, based on this
analysis. There is a little tiny clipping on the corner of
Panel 9 and there's a little tiny clipping on the edge of
Panel 6, but the most likely area is in the area of Panels
7 and 8 and we think that this has been analyzed by so many
different people that we're fairly content to say that we're
in the ballpark. That's ascent.
Next. On orbit. We've not been able to turn up any evidence
whatsoever that leads us to the conclusion that there's any
way that we can detect on orbit any damage to the orbiter.
There's no photography. There's no telemetry. There are no
instruments like inertial navigation systems or sensitive
accelerometers, all of which have been checked, that indicate
that the orbiter was struck or in any way did not perform
perfectly while on orbit.
On the other hand, surveillance of the orbiter's flight path
-- just routine surveillance, nobody was tracking the orbiter
-- but just because the orbiter is something in space and
we track everything that moves in space, there was an observation
that on Flight Day 2 that there was an object in company with
the orbiter. As the testimony this morning indicated, if you
heard that, there was an observation of the orbiter at a certain
time. The next observation took place a couple of hours later
and there was an object there that wasn't there before. That
object was tracked for about 2 1/2 days. We know approximately
the size, weight, and mass of this object; and we are attempting
to match up what it might be.
We have eliminated many, many things and we have a couple
of parts of the orbiter that are remaining on the table and
this viewgraph indicates that. A part of the a T-seal or a
part of a RCC panel does pass the radar test as well as the
mass-to-area ratio tests, and they are candidates for what
this object might be. It does not prove that that's what this
object was. It just says that those objects pass the test,
whereas another 20 or 24 objects fail the test, and the testing
guess on. We're not done with this yet, but we have some candidates
which meet the tests and obviously since we have just discussed
that the orbiter was struck in the area of T-seals and RCC
panels, it becomes a possibility that we lost part of an RCC
panel or part of a T-seal which came off on orbit.
Next. These are the pieces that we're talking about. This
is what a full T-seal looks like from the inside. One of the
reasons why we build this working scenario is because now
it focuses on the debris. Obviously, if we end up with all
the T-seals down at KSC, then it can't be a T-seal. So this
is a very nice focusing thing.
Next. So that's what we're willing to say about on orbit.
Now for the entry. The de-orbit burn was conducted properly.
We found no anomalies. Everything worked as it should. The
orbiter performed as it should, and the profile seemed to
be nominal to us. Nevertheless, on the morning of 1 February
at the time that the Columbia entered the atmosphere, we believe
that she entered with a preexisting damage of some sort in
the leading edge system and that damage was in the area of
the RCC panels which we previously have mentioned.
The most likely area of damage is in the area of either RCC
Panel 8 or 9 and the T-seal between them. This is based upon
many, many measurements that are outlined in our scenario
here, particularly the temperature sensors which rise very
early and the strain gauges which indicate unusual strain
very, very early in the re-entry.
Furthermore, forensic evidence of the debris further suggests
to us that the pieces of RCC No. 8 and 9 and the surrounding
pieces which we have recovered do show some patterns which
suggest to us that the entry of heat started in the area of
RCC No. 8 or 9 or one of the T-seals. We've listed some of
the issues here. Extreme heating, the deposit of metal, the
fact that some of the RCC material which we've recovered has
been flame-eroded to a knife-edge sharp corner on it indicates
that there was a heat flux past the RCC.
Next, please. We've drawn this out for you here just to show
what we're talking about. We put a little blob around here
to indicate that I'm not sure that we know exactly where it
is. We haven't figured it out to the inch yet, but we're getting
pretty close. It's RCC Panel No. 8 or 9 or the T-seal in between
and we've brought along a model here, too, so you can see
as we describe how the heat gets in there. Sometimes a model
helps. What we're suggesting is that the heat entry started
right around in there -- and this is just a blowup of it --
someplace around RCC Nos. 8 or 9.
Next. This graph here shows the deposit of slag on the RCC
panels. This leads us to believe that because the slag deposits
are extraordinarily heavy in the area of 8 and 9, and just
plain heavy in kind of 7 and 10, that this is the area that
was predominantly affected by the highest temperatures for
the longest period of time and that metal from the internal
parts of the leading edge was melted -- probably more melted.
It may even have been vaporized to the point where it was
actually flame-sprayed in some of these cases. A lot of it
is just melted. It's slag. It's melted metal of several different
types. This again is another piece of evidence which leads
us to focus on RCCs Nos. 8 and 9.
Next. This is a view of Panel No. 9 on the left, Panel No.
8 on the right, as if you were standing inside the wing, looking
forward. In other words, the flanges, you might say, are even
with the screen and the apex is away from you. In other words,
we're looking into the RCC panels. Of course, the T-seal goes
in between 8 and 9. So this flange here and this flange here
would be about 1 1/8 inches apart from each other, with a
T-seal in between. These are the pieces which are eroded to
knife-edge thin, with a sharp edge, which indicates that --
and what we don't have any color on here, we don't have. We
don't have that piece of debris. It's gone and we can't find
So this is leading us to believe that perhaps on the lower
side right here that heat got in here someplace either from
the RCC itself, the T-seal, or this RCC, and as it blew by
these edges here, it sharpened them to knife-edge sharpness.
We have pictures of all that. Pictures don't show up very
well, so we didn't bring them. But we have the debris and
we have pictures of that.
Next. As far as the slag is concerned, this is a piece of
the RCC, a typical piece, and what we wanted to show here
is the composition of the slag. What we're suggesting is that
it's not just aluminum. We have metallic particles from all
of the parts, Inconel, stainless steel. Everything that's
in there, we have some deposits from chemical analysis. So
that indicates to us that the entire leading edge structure
was impacted by heat long enough for it to work its way through,
take all the pieces apart, vaporize them and melt them, and
then spray them around. And that has a certain connotation
of time as well as temperature because each of these things
melts at a different temperature. So it tells quite an interesting
Next. Okay. Entry interface, of course, was 44 minutes after
whatever hour you want to use. Eastern Standard is what we
use here. GMT is what we got briefed on this morning. But
at 44 minutes and seconds after the hour was interface. Then
we now go through this time line of indicating when the temperature
and strain sensors started to tell us that something happened
in the leading edge, and it all kind of leads us to this same
At Time 49:00 or about five minutes later, we have the first
temperature sensors starting to rise. This is even before
the time of peak heating, as we indicated before, and so we
feel that it's safe for to us develop a scenario in which
that Time 49 is when heat started to enter into the cavity.
By Time 52, three minutes later, temperature readings inside
the wing indicate that the leading edge of the spar was breached.
It took about three minutes for the heat to get inside the
spar. Now, this is significant because we are hoping that
with thermal analysis we will be able to determine what kind
of a hole, how big a hole it would take for heat to eat its
way through the leading edge of the spar. So this is significant.
Next. So what we are hypothesizing is that -- and we've drawn
it in Panel 8 but we don't have any proof that it's Panel
8 -- it could be Panel 8 or 9 -- that there was a breach on
the lower end of either Panel 8 or Panel 9 or the T-seal in
between and that heat got into the area behind the leading
edge system for about three minutes and then started to eat
its way through the spar.
Next. Once hot gas has got into the spar, it immediately began
to attack the spar structure, which is all aluminum, of which
we have not found so much as a shred, and it began attacking
the wire bundles that were in there. The wire bundles go to
hundreds and hundreds of sensors. We know at what time those
sensors went off line, whether the sensors rose because of
heat or whether it failed because its wire was cut, so that
we can paint a fairly complete picture, a fairly confident
story, because we have hundreds of sensors that are agreeing
here. So we're pretty confident about what happened.
We hypothesize, as you may recall from two viewgraphs ago,
that the leading edge of the spar, the spar front bulkhead
was penetrated at about Time 52, 52:00. The first wire fails
at 52:16. The first sensor fails at 52:16 -- 16 seconds later.
Then over the next four minutes, over 164 other sensors fail
because their wires were cut, most of which failed almost
immediately but some of which hung on for a while.
Next. So our scenario then is that at Time 52 we breach the
Next. And we got into the interior of the wing. This is the
spar breach right here and heat now has filled this entire
cavity in here. Many of the wire bundles run along this bulkhead
right here, and some of them run along both this bulkhead
and also a bulkhead which runs right here which we'll show
you a picture of in a second. Once the heat got in here, it
started cutting wires here and cutting wires right along here.
And what we're talking about is, of course, right in here
on the model, is what we're talking about there.
Now, we showed that the heat -- there is a vent hole. There
are vents in here, and we don't know if the flame was able
to breach this bulkhead and get in here or whether or not
it just went in through the vent hole. We're not ready to
make a positive statement about that, but there's no question
that then heat got into the wheel well, because we have a
lot of sensors that tell us this. Of course, it vents out.
There are large vents in here. You see that square right there?
That's that square right there. Once the heat gets in here,
it can go all the way through the wing.
Next. So at 52:16 we have heat in here. This is the outboard
wall of the wheel well. This is forward. This is top. This
is bottom. This is out. This is left, out on the wing. So
on this diagram here, we would be looking at this bulkhead
right there. There are the wire bundles that we're talking
about, some of the wire bundles. They begin to burn.
Next. About this time, because of all this heat inside the
wing, the wing begins to deform and a whole lot of things
happen which we can't prove but we believe that probably,
because the wing began to deform, tiles started popping off
and losing their adhesive bonds because they aren't bonded
to a flat surface anymore and the debris-shedding events occur,
the first of which is about a minute and a half after the
wing is breached. The first observed one. Now, there may have
been debris-shedding events for a long time before that and
we just didn't know about it, but this is a minute and a half
later now from 52:16 to :46. We see the first piece of debris
come off. But by now damage is extensive inside the wing.
Hundreds of sensor have had their wires cut. Communication
blackouts are occurring because of probably disruption of
-- the hot gas flow around the orbiter.
Again, at Time 52:06, which is back up just a little bit,
is the first significant increase in drag. The aerodynamic
control surfaces begin to work against that aerodynamic drag.
The vehicle has no problem at this point maintaining the proper
attitude, the proper angle of attack, and a proper roll angle;
but it is beginning to work at it.
At Time 54:20, there is a fairly significant aerodynamic change
as the vehicle begins to work considerably harder. There are
debris-shedding events during this period of time, but we
can't relate them directly. By two minutes later, at 56:16,
the wheel well temperatures, temperatures along the wheel
itself begin to rise, and so now we know that the heat is
into the wheel well itself.
Next. 58:09, there is another aerodynamic change in the aerodynamic
surfaces, accompanied by more debris-shedding events. By 58:56,
all of the tire pressures and temperature measurements in
the left main landing gear wheel well are lost, due to damage
in there, mostly to cutting of wires and sensors.
At 59:29 there was a major aerodynamic event which causes
the shuttle to bring its nose down. It wants to roll left
which, when you're at a 70-degree roll angle, that means down.
This causes all four of the RCS jets on the starboard side
to fire to try to counteract that. Of course, the loss of
signal occurred at 59:32. The MADS recorder continued for
another 30 seconds or 40 seconds; but as we indicated before,
there was nothing particularly significant in that additional
So that is the scenario that we've agreed on as to how this
orbiter failed; and as I indicated in my preliminary comments,
the real value of this is that it tells us what to do now
and where to go next.
If I could have the next viewgraph, please. We have now listed
-- and it's in the little handout that we've given you. We've
listed the areas that we're going to focus on, the areas of
analysis and the areas that are in work by both us and NASA
that we want to put some more steam into so that we can get
some of these things answered. We've listed them. I won't
read them all, but they are the foam impact testing that's
very important to us. We have to continue the very, very hard
work on the aerothermal analyses to correlate some of these
events that we have really good data on but yet we can't quite
fully understand everything that's happening, like how long
does it take to burn through Kapton wiring, how long does
it take to create a knife edge in a piece of RCC that's a
third of an inch thick, how long does it take to burn through
aluminum skin, and those kinds of things. So we have really
good data on this, and so we need to work real hard at fully
As was mentioned this morning, there are several skin temperature
anomalies where some of the skin temperatures up on the fuselage
actually get cooler than normal and return to hotter than
normal. That is pretty easily explained by a change in the
flux, the flow around the orbiter to where a hot spot which
is normally located at one place moves; and that is probably
accounted for by a change in the geometry of the wing. We
have a lot more analysis to do before we're able to say anything
specifically about that.
And then continued forensic testing on the debris. We have
all this wonderful debris down there. We are doing several
things with it, including building a 3-D model with the debris
with have. A lot more chemical analysis, metallurgical testing
of the debris, and including more dissecting of foam and a
better understanding of the cryopumping issues.
So that is in viewgraph form, which we have also typed up
in a printed form, and all of these will be posted on our
website and we are now ready to answer questions.
LT. COL. WOODYARD: Okay. We'll begin right here in
Houston. We'll start on the end.
A REPORTER: ABC News. Admiral Gehman. I'm a little
bit off topic; but because I get one question, I have to go
with this. The missions operation director last week released
their report on what they thought they could have done, if
anything. It was very limited in scope. We had chatted briefly
about what the board thought its involvement would be in those
kind of scenarios. Given that report, are you looking at readdressing
this issue down the road?
ADM. GEHMAN: I have to say that the board is not finished
with that level of inquiry, and I really don't know how the
board is going to express itself on what could have been done.
Anything I said would be getting out in front of the board,
because we have not spoken on that.
I don't have any comment on the study that NASA did. I mean,
it was good that they did this and it would have had to be
done anyway, but I don't really have any comment on it because
we put things like that in kind of the second tier of priorities.
Whether or not we address ourselves to that or not, I'm just
going to duck that question because it hasn't come up yet.
A REPORTER: Orlando Sentinel. For Admiral Gehman also.
The foam impact tests at Southwest Research. If the results
of those show that it's unlikely that damage could have been
done on the scale that you would need for it to have been
done to make the scenario work, what does that do to not only
the scenario but the entire sort of investigation? I mean,
do you go back to Square 1? Where does that leave you?
ADM. GEHMAN: We are very careful in our work to not
have a scenario-dependent investigation unless we can prove
it. We are very careful to make sure that our scenario and
our working hypothesis cannot be contradicted by a confirmed
fact, but we're very careful to make sure that we don't have
a scenario-dependent investigation. If we find areas of safety
or risk assessment or budgets or oversight or quality control
that we think need to be improved in order to make the shuttle
safer to fly, our report will be rich and complete in that
we cover all these subjects and not be dependent upon whether
or not the foam broke the RCC or not.
If we find, for example, that we feel that the fault originated
in the leading edge system in some way and then we start to
look at such things as whether or not the aging of the leading
edge system has been properly measured, we can make a recommendation
that would make the shuttle safer to fly because NASA would
understand the characteristics of the aging process better
and it wouldn't be dependent on what broke -- keeping in mind
that, as I have said before, we may hit the RCC, for example,
a good hard shot and the RCC does fine and the bolts break
or the linkages break or something like that. So we're being
very careful not to have a scenario-dependent investigation,
particularly since at this time we can't prove beyond a reasonable
doubt what initiated this thing.
A REPORTER: Associated Press. Can you describe or characterize
the level of agreement with members of the board on this particular
scenario or, alternatively, could you quantify your level
of certainty so far as your conclusions here?
ADM. GEHMAN: We have been working on developing this
scenario for about ten days and in that ten days we have attempted
to build the scenario general enough and generic enough that
we have embraced everybody's opinion and anytime that we had
a fairly large or fairly strong view one way or another, we
either have accommodated that or we left it out. In other
words, there's all kinds of data and work going on in this
investigation that's not included in this scenario. So I would
say we have a very, very high degree of agreement that this
working scenario is something that everybody can live with.
Remember my caveat at the beginning. The reason we call it
a working hypothesis is because we reserve the right to change
any part of it at any time without any notice. So with that
caveat, it's hard not to agree with it.
A REPORTER: USA Today. Just for the record, Admiral,
it seems to me, in looking through this, that you do not ascribe
a root cause -- that's probably the wrong language, a cause
-- do you not draw any link --
ADM. GEHMAN: Direct cause.
A REPORTER: Yes, direct cause. You do not find any
link between the events on launch and what happened. Am I
right in thinking that; and, if I am right, when did you expect
to have a direct cause?
ADM. GEHMAN: You are correct. We're careful not to
say that the foam knocked a hole in the leading edge of the
orbiter, because we can't prove it. Now, that's not to say
we don't believe that's what happened, but we were careful
here to base our working scenario on agreed facts. And right
now we aren't willing to make that kind of a statement.
It's too early to characterize how the board will form the
words that describe the direct cause. I mean, whether we say
"most likely," "we are sure," "probably," "maybe," "could
have been," "would have been" -- I mean, I don't know how
strongly the board will characterize the cause chain here.
I would say that the board is certainly leaning in the direction
or the board is certainly suspicious that the foam had something
to do with this, but we were very careful in here to not make
a statement that we couldn't live with.
A REPORTER: CBS News. Maybe Dr. Widnall can answer
this. I don't know. But the two things that survive the radar
cross-section and ballistics analysis -- to date, anyway --
are the T-seal or large piece of a RCC fragment. I mean, the
sizes of these things are pretty different in terms of the
resulting hole you would have in the leading edge. I was just
wondering how the date or the timing of sensor failures --
would it be easier to believe it if you had a big hole, like
a 120-square-inch piece of RCC, or this 1.8-inch gap or whatever
it is. I don't understand. I mean, it would seem intuitively
to a non-expert it would be pretty easy to pick which one
of those would explain the data.
DR. WIDNALL: Well, my understanding is that NASA is,
in fact, doing specific analysis for those different shapes,
sort of two-dimensional analysis for the slit and then --
I think my understanding is that the hole sizes they have
been using to date are quite a bit smaller than the RCC panel
that was suggested in the Wright Patterson radar tests. So
I think more analysis is clearly required.
ADM. GEHMAN: I'd expand on that by saying that this
is one of the key areas that we're going to continue to focus
on and the way I like to describe it is that the breach that
was there at the time of entry has to be big enough to cause
the heat scenario that we saw but it also has to be small
enough that permitted the orbiter to get all the way to Texas.
So keeping in mind that we've got some bounds in there and
we've got a very, very rich time line, I believe that we've
got a good chance of achieving the analysis it's going to
take to be more specific about this; but we're talking about
weeks of work here. It's just plain hard work.
A REPORTER: NBC. Admiral, looking at the issue of the
direct cause versus the contributory causes to -- possibly
connecting the foam impact with the breach, can you discuss
for us the various contributory causes to the unobserved weakening
perhaps of structure that could have combined with an impact
to lead to a breach? What sorts of options have you been looking
at and what kind of interesting things have you found?
ADM. GEHMAN: The way the board is going to structure
the report is we're going to attempt to state -- we're going
to use three or four layers of causes. The first level is
going to be at what we're calling a direct cause; and this
would be, for example, if we were to come to the conclusion
that the foam striking the leading edge put a hole in it.
Then that would be either the direct cause or the physical
cause or the mechanical cause of this accident.
At the second level, we're going to use the term "contributing
factors." Not causes. Contributing factors. Those would be
such things as a weakening of the leading edge. They might
be things such as quality control at the shop floor or the
construction processes in building the ramp in the first place
or budget constraints or something like that.
Then at the third level, we're going to have what we're calling
the root cause. This may not be the way the safety experts
do this, but the root causes are the cultural kinds of things
-- the age of the orbiter, the budgets, the manpower, the
climate, the management systems that the program operates
in. The climate. The background kinds of things. Attitudes.
Then at the fourth level, the board is free and is charged
to make what we are calling any significant observations.
This would be any observation that the board members make
that may have nothing to do with the shuttle program. It may
be something that we noticed when we were visiting at one
of the plants or factories or centers and we just felt that
NASA would benefit from us saying that this hasn't got anything
to do with the loss of Columbia but we did notice something.
As far as giving you a list of possible contributing factors
at this time, it would be fairly premature, but I certainly
can say that things that we have under consideration are things
like crew performance, whether they made a mistake or not.
The payload would be a contributing factor. Maybe. Maybe not.
What I'm saying is that they are contributing factors and
we would say that they're non-causal. In other words, the
payload, for example, we will have examined in great detail
to see whether or not it was a contributing factor, and then
in the report we will say whether or not it had anything to
do with this or not. But it will be in the report.
Weather. Space weather. Micrometeorites. I mean, I do have
any list with me, but I'm not going to get any more detailed
than that. But I would suggest that we probably will end up
with , 12, 15 factors that we will address in the report and
we will then adjudicate whether or not we believe they had
anything to do with this or not.
LT. COL. WOODYARD: Any other questions from Houston?
A REPORTER: New York Times. I just wanted to ask whether
there's any significant difference between the working scenario
you're laying out here today and the NASA briefing which you
recently received, that lays out what they came up with as
a working scenario. Where's the daylight here between these
ADM. GEHMAN: There is no daylight, and that is the
purpose of us massaging this for the past ten days. We have
been meeting at the working level, as well as the leadership
level, to make sure that there is only one working scenario.
That's what we did. That's what we've been doing for the last
ten days. As of this afternoon, this is the working scenario.
You know how it is. Every time we announce something, then
we go find another piece of debris and it changes. But there
is no daylight and we have worked hard to make sure that the
whole organization is speaking with the same script at this
A REPORTER: Washington Post. Just sort of a variation
on the earlier question. If the tests show that the foam strike
doesn't turn out to be the cause of the breach, what would
be the strongest backup theory as to what would happen; and
are you determined to continue the investigation until you're
able to say definitively what was the cause?
ADM. GEHMAN: No. I have to answer both questions in
the negative. Even if the foam impact testing does not break
any of the leading edge systems to the degree that it creates
a plausible hole -- or let me take the other side. Even if
the foam impact testing does break the RCC or does break the
leading edge system, that doesn't prove that that's what happened.
It just proves that it could have happened. It's just like
radar cross-section testing of the Flight Day 2 events. If
we conclude that it could have been a part of an RCC panel,
that doesn't prove it was part of an RCC panel. It just proves
it could have been. So the foam impact testing will demonstrate
to us whether or not we have a plausible scenario or not,
but it doesn't seem to me that it will prove anything one
way or another. Therefore, the report doesn't hang on that
-- at least in my mind, it doesn't.
A REPORTER: Houston Chronicle. I guess I'm backing
up a little bit with my question. I just want to make sure
that I sort of understand your thinking on why it's important,
beyond just good investigative analysis, not to reach for
the foam right now if it's not there. I guess I'm thinking
in the context of how NASA would respond to change the way
the orbiter is constructed or prepared for flight or actually
the flights are executed.
ADM. GEHMAN: I don't understand your question about
why don't we latch onto the foam.
A REPORTER: Yes. Why don't you latch onto it?
ADM. GEHMAN: You mean why don't we state that the foam
caused a hole --
A REPORTER: No, I understand that. Beyond good investigative
technique, why would it be bad for the space program to be
premature in taking that position? I guess I'm trying to explain,
to people who are lay, why you're not going there yet.
ADM. GEHMAN: Okay. Let me see if I can answer it this
way. Anything that has to do with an event which the orbiter
is not supposed to be subjected to, which includes being struck
by debris, is being analyzed and examined very, very carefully
by the board, up to and including taking a foam object about
the size that came off and actually firing it at the leading
edge. And we may come to the conclusion that no foam should
strike the orbiter. And I think NASA knows that and, I mean,
that will be a blinding display of the obvious.
We are going after every single aspect of the foam hitting
the orbiter -- and every other anomaly that's happened on
ascent -- with a lot of rigor. I don't believe that it's necessary
for us to state at this time that the foam strike causing
a fatal flaw in the orbiter is necessary to energize the investigation.
As we said, our scenario is based on things we know. The one
thing that we did in our scenario is we made sure that we
have no facts that would contradict our scenario. So that's
still a possibility, but I don't think it's necessary for
us to move forward.
GEN. DEAL: If I can inject on that, sir, there is extensive,
extensive work already going on that the board is closely
monitoring at Marshall Space Center, not just the dissection
of the tanks but also the redesign of the bipod ramp, for
example, looking at the flange and the ramped area on the
tank as well. Plus the most extensive work is in different
types of NDE, non-destructive examination of the tank itself
and the foam. So we're monitoring that very closely, and we
have an entire foam team made up from members of the entire
board that are going to and from Marshall and working with
them on a daily basis.
ADM. GEHMAN: That's very pertinent and very relevant
because it illustrates that everything that we think is related
to this damage scenario is being investigated and reviewed
and recommendations are being considered and whether or not
we state that the foam caused this chain, I don't think, in
any way lessens the urgency by which we're going after all
LT. COL. WOODYARD: Okay. We'll go to the phone bridge.
A REPORTER: A question for Admiral Gehman. It would
be very helpful to visualize this, if you wouldn't mind pointing
out on the model of the wing right where Panels 8 and 9 and
the T-seal are and showing where you're theorizing the strike
was and the path of the hot air or plasma.
ADM. GEHMAN: Sure.
(Indicating) All right. This is the under side of the left
wing. This is forward. This is outboard. This is aft obviously.
This is RCC No. 9 right here. This is the wheel well with
the landing gear obviously. This is RCC No. 9. This is RCC
No. 8. This is the T-seal between 8 and 9. This goes 7, 8,
9, , like that. What we are indicating is that whether the
breach is in 8 or 9, right in here, that once you get in here,
you're free to move about. The heat can go through here and
then work its way. As a matter of fact, these things come
Wait a minute. I misspoke. This is 8 right here -- 7, 8, 9.
This is 8 that we were referring to and the idea is that --
these things come off, by the way, if I can do this without
breaking this wonderful model. The idea is that once you get
through here -- they're magnetic -- once you get through here,
then you burn through the spar. Then once you burn through
the spar, you're into here.
What we are hypothesizing is that the breach is on the under
side here, on one of these under sides and then goes through
the spar. That's what we were hypothesizing. Then once it
gets new here, this is all open. You can go through here.
These are just little struts, and the heat can go any way
it wants to and then starts cutting wires. There are wires
that run along here. There's a big wire bundle that runs along
here on the top. In fact, you can see it. Here's the wire
bundle that runs along here. Once the heat gets in here, it
starts cutting wires. I hope that answers your question.
LT. COL. WOODYARD: Next question on the phone bridge.
A REPORTER: Los Angeles Times. Admiral Gehman, do you
have a working scenario for why NASA had a half-dozen foam
strikes that originated in the area of the bipod ramp and
that all apparently violated its launch criteria and yet it
continued not to regard that as a safety problem and do you
have a working hypothesis for why during the mission dozens
of engineers convened a meeting and concluded that the foam
strike was serious enough to warrant photographs to be taken
and that was never executed by management?
ADM. GEHMAN: General Hess and his group are looking
at that very hard. I'll let him answer that.
GEN. HESS: I think the best way I can answer your question
is to talk about it in terms of over time there were many
changes made to the external processes of applying the foam
to include the type of blowing agent that were used as well
as looking at ways to stop the popcorning of foam off as it
was hitting the orbiter. So I wouldn't characterize that NASA
took no action to abate the foam strikes over time.
As we have looked at the Flight Readiness Review in preparing
for the COFR for various flights, we found quite a bit of
detail in terms of how they discussed the foam issues; and
they have a piece of hazard analysis that they conducted that
kind of led them to believe that there would be no safety-of-flight
issue with regards to foam coming off of the orbiter. I'm
not willing to characterize this as just kind of a normalization
at this point but they were working the problem but over time
they became, I think, very used to the fact that foam was
coming off and they never had a disastrous outcome as a result
of it. So that probably led them to believe that they could
continue to try to work the foam problem in some way without
any adverse effects to the shuttle.
Now, with regards to the engineers that were working and the
E-mails that you're talking about and the requests for imagery,
we are fairly certain that the damage assessment team that
was working the problem of this particular mission was attempting
to get DOD imagery because they needed to be able to characterize
the level of damage to the wing in order to complete their
analysis. There were other people inside the NASA organization
who had requested an opportunity for the DOD imagery; and
inside the NASA decision-making apparatus, the MMT and below,
the two requirements kind of ran into one another. When they
did, they canceled each other out because it was never realized
by management that the request for imagery really was coming
from the damage assessment team, because the communication
was not totally clear there. They thought that the request
was coming from kind of outside Johnson Space Center, like
from Kennedy; and we're trying to nail down at the management
level who was asking for the requirement and it was never
made clear to them that the people who wanted the requirement
were actually the damage assessment team altogether.
So as a consequence they said, well, we haven't asked for
an image. That message got back to the damage assessment team
as, no, we're not going to let you get an image; and that
was really the wrong message to be sent. At the end of the
day, what happened was that everybody went back and did the
best they could with the information they had and never moved
forward or pursued down to the root exactly who was asking
for an image and why there was all this pressure that was
down at the working level.
I'd also say that by the time they were briefing the mission
management people, they did brief them in a very short fashion
that there was no safety-of-flight issue and that even though
that there could be pretty serious damage to the orbiter,
it was the consensus of the working group that it was not
a safety-of-flight problem. So all this served to be a pretty
good deal of miscommunication about the problem altogether.
LT. COL. WOODYARD: Next question from our phone bridge.
A REPORTER: Newsday. For Admiral Gehman. If you can't
say for sure that the foam did the damage, what are the implications
for return to flight? I mean will fixing the bipod and doing
better NDE on the RCC panels be sufficient?
ADM. GEHMAN: Those two things alone won't be sufficient,
but we can make a fairly good case of what we think the return-to-flight
criteria should be, with or without any positive knowledge
or positive proof that the foam caused the accident, because
we are going to make return-to-flight recommendations which
are designed to enhance the safety of the orbiter in every
way that we find that it needs to be enhanced. Just fixing
the foam alone won't do it.
So I'm not in the least bit concerned that our inability to
make a positive statement with proof that the foam knocked
a hole in the leading edge of the orbiter in any way slows
us up or in any way restricts what we need to do in order
to come up with criteria for return to flight. Most of our
work or most of our thought processes on return-to-flight
issues have to do with the fact that the safety margins have
been changed over the years, and we're going to try and restore
those safety margins back at least to where we're comfortable
A REPORTER: Aviation Week. I have what you might call
a fairly fundamental question on quality control for General
Deal. Correct me if I'm wrong; but philosophically at least,
the shift to the contractor, United Space Alliance, was to
have freed up NASA personnel for more broader oversight. Now
that you've looked at it for two to three months here, can
you characterize whether this was actually occurring effectively?
GEN. DEAL: Well, you'd have to look in the bigger context.
That's a great question, because there's been many, many changes
as you go through looking at the history of it and how they
have gone down in the number of GMIPs that were put out, Government
Mandatory Inspections Points, and how they've actually added
on some more since them. You'd have to go even deeper to look
at the manning that they have there and what they are tasked
to do. We're to the point now where we're ready to present
some things to the board to reference that because they have
moved toward some things that are in light of ISO 9000 and
some other types of quality progressions where you're doing
sampling versus doing a continual look at something that only
flies once or twice a year. So we're looking very heavily
at that as to whether that is the right answer for a program
like this that's manned space flight.
Effective? If you look at the number of rejects that they
have and their metrics that they have, it's pretty good. However,
is it up at the level that it should be? That's what we're
looking at. I'd be hesitant to give you an answer today on
that because we still have more analysis to do and more metrics
that we're receiving.
ADM. GEHMAN: But that will be a significant part of
the board's report.
A REPORTER: New York Times. Admiral Gehman, if you
don't conclusively connect the foam strike to the damage,
do you raise a permanent doubt about whether the shuttle with
fixed foam is, in fact, safe to fly in the future?
ADM. GEHMAN: I don't think so. It's the board's view
that our review of this program is so much wider than that,
and this is one area in which we differ from the Challenger
investigation. Our review is considerably wider than foam
striking the leading edge. It has to be, since we can't be
absolutely sure that that's the direct cause. Therefore we
believe that we have a good chance of coming up with recommendations
that will, indeed, make the orbiter safe, the shuttle safe
fly. I mean, that's our goal is to come up with a number of
recommendations that cover a broad area of the program, not
just foam striking the leading edge. That's why, for example,
you heard these comments about quality assurance and committees
and boards and organization.
General Hess, do you want to follow up on that?
GEN. HESS: I think it would be important to add at
this point that the board is never going to make the shuttle
program risk-free. It's always going to be risk-risk trade-offs;
and at the end of the day whenever they decide to fly a mission,
they're going to be accepting a high degree of risk in a lot
of areas. So I think it would be a little bit naive of us
to think that we were going to be, by our proceedings, able
to give them a zero-risk program.
ADM. GEHMAN: That's absolutely right. I know that is
an obscure and complex issue in the sense that we don't have
any concrete, irrefutable proof that the foam striking the
leading edge was the cause of this accident yet and then for
me to say that it really doesn't make much difference to the
board; but that's the truth. The truth is that we are looking
at this program and these shuttles in a very, very broad way.
We have to because we don't have a single-point failure like
the O-rings. Therefore we're going to come up with a broad
range of recommendations which, taken together, we believe,
will make the program safer. The fact that we don't have a
single causal event doesn't bother me in the least -- I don't
know about my fellow board members. It may not be quite so
easy to explain; but practice-wise and function-wise, it doesn't
bother me in the least.
A REPORTER: Discovery Channel. Admiral Gehman, what
would constitute the concrete, irrefutable proof that the
foam strike did cause the breach?
ADM. GEHMAN: We've actually asked ourselves that a
couple of times: What would it take to prove that the foam
did it? Obviously if we had a picture, that would be nice.
That's why this morning when we had the ascent experts here
and the video reconstruction experts going over this with
some detail, we brought this issue out and we dragged this
issue out. That would be nice. If we found a piece of debris,
if we found a piece of leading edge RCC with a lot of foam
embedded in it, that might convince us of something. But the
answer is I'm not sure what it would take to prove that the
foam did it.
As we have said two or three times already today, even in
the foam impact testing, if we do major damage to the leading
edge system, that still doesn't prove it. That just proves
it's plausible. So those are two things that pop into my mind.
I don't know if there's any others, but we would love to find
some evidence like that but it isn't out there, I don't think.
A REPORTER: NASA Watch. A very simple question. You
brought up the issue of the fact that this is never going
to be risk-free, but is the panel going to endorse any specific
safety-of-risk number that NASA uses, encourage NASA to go
back and revalidate that, or make any statement on the specific
level of risk that is acceptable?
ADM. GEHMAN: The board is going to attempt to characterize
the true risk in our own words. Now, whether or not we put
a number on that, the board hasn't decided; but we are going
to attempt to describe for our constituents -- the Congress
and the administration and the astronauts and the people of
the United States -- what the risk is in this enterprise.
As General Hess said, it's not zero. It's not anywhere near
zero. I don't know that the board would be interested in putting
a number on it. Maybe we would. We haven't got that far yet.
And we certainly would not -- well, I better be careful what
I say here because I don't know this to be a fact -- whether
or not we pass judgment on any number that NASA uses remains
to be seen. But we will attempt to characterize the risk in
our own terms. If it differs from NASA's, so be it; but that
will be one of our goals, to restate the risk in terms that
there can be a good public policy debate on whether or not
we should be doing this or not.
LT. COL. WOODYARD: One more here in Houston.
A REPORTER: Stern Magazine, Hamburg. I heard conflicting
versions to which extent the crew was informed about the foam
impact and the discussion that followed. Could you clarify
this for me, exactly what exchange took place and when did
GEN. HESS: Yes. I can't remember exactly which flight
day it was, but there was a scheduled press event with the
crew. In the day prior, there was an E-mail sent up to them
that said they had a foam strike on the orbiter on ascent;
and then they also E-mailed up to them an impact file that
showed the same images that we saw earlier today with the
debris striking the shuttle. They were also advised by the
NASA management that they didn't think that there was a safety-of-flight
problem. So they were very much aware that the strike had
LT. COL. WOODYARD: Any other questions from Houston?
If not, that will conclude today's press briefing. We thank
you for coming; and again, as always, the board members will
remain here to answer a few more questions. Thank you.
(Press conference concluded at 2:15 p.m.)
Back to Event