|Columbia Accident Investigation Board Public Hearing
Monday, March 17, 2003
Hilton Houston - Clear Lake
3000 NASA Road One
Board Members Present:
Admiral Hal Gehman
Rear Admiral Stephen Turcotte
Brigadier General Duane Deal
Mr. Roger E. Tetrault
Dr. Sheila Widnall
Mr. G. Scott Hubbard
Mr. Steven Wallace
Dr. William Ailor
Mr. Paul Hill
Mr. Robert "Doug" White
ADM. GEHMAN: Good afternoon, ladies and
gentlemen. Welcome to our second public hearing. The
subject of this afternoon's hearing is going to be a
discussion of the reentry of the Shuttle Columbia, and
we'll hear from several witnesses this afternoon. The
first one is Dr. William Ailor. Dr. Ailor is the
director of the Center for Orbital Reentry Debris
Studies from Aerospace Corporation.
We are very thankful, Dr. Ailor, for you
for taking time to come down here and help us walk
through this. What the board is interested is, first
of all, a non-NASA view of how things reenter the
atmosphere, which will help us form our questions for
later this afternoon when we get the detailed analysis
of how the Columbia entered the atmosphere, and your
presentation will help us understand to a much greater
degree what we'll hear later.
Dr. Ailor, I would offer to ask you to
give us a short bio or short background, if you please;
and then if you're prepared to start, we are prepared
testified as follows:
DR. AILOR: Okay. Thank you very much.
Just by way of background, I joined Aerospace in 1974 and have been basically working reentries ever since
that time. I'll go over in my presentation a little
bit more detail on some of the ones we've worked on
before, but Aerospace established the Center for
Orbital Reentry Debris Studies back in 1997 in
recognition of the kinds of issues that we expected to
see from both space debris and the hazards posed by
reentry and in recognition that there needed to be a
fair amount of work done to understand the reentry
breakup process. I'll go over some of that in my
So a little bit more background, I did work on the external tank reentry a number of years ago, one of the issues where it was associated with what altitude did that break up. We worked very closely with NASA in resolving those issues. Then I've also been in various capacities on the Interagency Nuclear Safety Review Panel, which reports to the White House on space missions which carry radioactive materials so Cassini, Mars Pathfinder, Mars Exploration Rover. We've worked on all of those.
So if I could have the first chart.
Okay. Go back one. No, that's good, I'm sorry.
What I'm going to talk about is what we
can learn from reentry debris. This is really based on the experience that we've had over the last 25 years in
this area, actually longer than that. Aerospace has
been working in this area for a long time, and our
desire has been really to understand the breakup
process. Again, these things coming down through the
atmosphere can present a hazard to people and property
on the ground. One of our objectives has been to
understand what that hazard is and to be able to model
it and perhaps minimize it as time goes on.
So what I've got here is an overview of the reentry breakup process. This is just for a standard reentry; and as I'll show you in a minute, we see a number of these a year. For a typical satellite reentering, it slowly comes down through the atmosphere, slowly works its way down out of orbit in an orbit decay fashion or, in fact, you can actually drive something into the atmosphere and I'll talk about that in a bit, as well.
Basically for unprotected space hardware, the heating and loads will gradually tear it apart. I'll talk more about that in minute. The kinds of things that we've seen that survive reentry are things that you would probably guess might, things like steel sometimes I'll talk about that glass, titanium, and then parts that are sheltered by other parts.
One of the things about the reentry
breakup process is that the heating is like, in a
sense, cooking an onion. You basically start from the
outside; and then as you heat the pieces up to a point
where the materials will fail, that will expose some
new materials. They'll go through the same process and
the object can be broken apart. We do have objects
that are melted and shedded away, things like aluminum,
solar panels. Things like that come off pretty early.
Mylar sheets. Some satellites are wrapped in Mylar
Once this debris comes off from the
parent body, it follows its own trajectory at that
point. So it will go on about its business, basically,
based on its own properties. If it's a very dense,
heavy piece, for example, it may go further. If it's a
very lightweight piece like a solar panel or something
like that, it will fall early in the trajectory.
Then the debris pieces impact on a
footprint on the ground. I've got an illustration
there that just shows that typically what we see is
initial breakup or shedding of some things like solar
panels that come pretty quickly. And we have
catastrophic breakup. I'll talk more about that but
typically it can be quite a substantial event. There can be secondary breakups that happen when those pieces
come apart. Then you see a footprint where you get
low-mass debris that comes in early; and typically
longer, heavier pieces go late. We'll talk more about
that, as well.
Next chart. Okay. So just some
characteristics of reentry breakup. It's characterized
by intense heating and major fragmentation; and as I
mentioned, fragments are shed as the structure heats
and fails. Typically we see instantaneous high loads.
For example, when an object comes off of a parent body
it now experiences the air stream that exists there;
and it will respond based on its own characteristics.
For example, if you've got a very lightweight piece
that comes off of a heavier object that's coming
through the atmosphere, it's like throwing a piece of
paper out of a car. That will decelerate very quickly,
and the same things happens even at Mach 20. So when
you do that, you see very high loads; and you can also
see very high heating. That can be important if you're
trying to understand what actually happened in the
process, because now you've got an object that's been
separated from a parent body that, just because of its
own interaction with the atmosphere, will have seen a
fairly severe environment.
You can have some events with a moderate
velocity increment. What I mean by that is if you've
got a fuel tank or something like that that explodes,
it's like a balloon. Some of those fragments will pick
up some velocity increment from that. We've measured
as high as a thousand feet per second. And the initial
breakup can be energetic. Basically a typical way for
things to break up when they reenter is that they'll
come down through the atmosphere for a certain amount
of time, they look absolutely fine, we've seen videos
of these things where they just like spacecraft coming
down, and all of sudden they come apart. When they
come apart, they just disintegrate. That altitude
typically is about 42 nautical miles, plus or minus a
few nautical miles; but that's a pretty good guess. So
just as a rule of thumb, it seems like a critical point
for space vehicle reentry and breakup is around
42 miles. We have never had any measurements internal
to a spacecraft during this breakup process and that's
something that we would like to see. It would really
help us understand the process better.
Next chart. Survivability depends on a numbers of factors. The material. For example, the melting point of the material, the heat capacity. Just by example, it's very rare to find aluminum on the ground from a standard spacecraft reentry; and finding aluminum on the ground would basically mean that that aluminum was somehow protected as it came down. Steel can survive. It doesn't have to, though. We have cases for example, there was a Russian satellite that came down in Canada, had steel, a reactor case. That reactor case basically disintegrated during the reentry, but also I'll show you some pictures of steel that did.
Size, shape, and weight. An empty fuel
tank, in a sense it's a lightweight object relative to
its size. That will affect its survivability, and that
can be very important. For example, fuel tanks
survive. Things as dense as battery? We've never
found a battery on the ground.
Release conditions. If an object comes
out late in the reentry, after being shielded for a
portion of the reentry, that means a lot of the energy
has been taken out of that trajectory prior to that
object's release; and that object is more likely to
survive. And shielding. Again, objects that have been
shielded for partial reentry can survive; and that's
one reason, by the way, that, for example, you can find
circuit boards on the ground from satellite reentries.
What that means is typically when a satellite is being constructed, circuit boards are built internal to other
boxes which are internal to other structures and so
forth. Again, if you think about this heating process
where you're removing the outer layers as you come in,
every time you do that, you're removing energy and then
finally these things will be released.
Next chart. When these things come down,
there's a typically generated debris footprint. Now,
this is a notional footprint here. I've got several
breakup conditions separated by about 30 seconds in
trajectory time. This shows things like the types of
dispersions that we typically see. This has got
dispersions in winds. So winds will affect things as
they fall, even a big, heavy object, as I'll show you
in a minute. Ballistic coefficient is a measure of, in
a sense, how dense an object is; and that will affect
where things go. Typically on these footprints at
least, for example in the red swatch you see up there,
things that have gone longest downrange, farthest to
your right, would be heavier objects. The lighter
objects would hit towards the up-range portion.
Then atmospheric density. We don't quite
know what density is in most trajectories. So in that
case we have to build a factor in to allow for that.
Then also, as I mentioned, it's possible to get some velocity increment as things come down. So we put in a
delta feed for that.
So basically what you can see here is
these ellipsoids were generated at each of these time
intervals, and you can see how they overlay each other.
If you look carefully at Breakup 4 down there, that's
the one where the trajectory is now healed over a bit
and you can see that even though the same types of
debris are there, the footprint is inside of the one
just prior to that. This indicates that trying to
figure out where debris came from on a reentering
spacecraft and where it happened is a very difficult
process, indeed. These are four specific time steps.
What you have to recognize is this is basically
happening continuously as the spacecraft reenters. So
the footprint is not even as nice as what you see here.
Next chart, please. Noteworthy
reentries. Just to give you a little background, it
was mentioned earlier that someone said this is not a
data-rich area; and I have to agree with that. What
you see here are some of the primary data sources for
doing this type of work. Cosmos 954 came down in 1978.
That was a reactor-powered satellite and there was
radioactive debris that came down in Canada. Since it
was radioactive, you could find it pretty easily and a lot of that debris was recovered and was examined and
documented. That's probably exceptional on these kinds
of things. Typically the effort is simply not put
forward to find debris on the ground, and so we simply
don't have as much.
Skylab occurred in 1979. Some of the
debris fell in Australia. There was some debris found,
but again there was really no detailed analysis of the
footprint itself, as far as I'm aware.
I'll show you some pictures of some
Delta 2nd stages in a minute. We do have large debris
pieces surviving from that. We did reconstruct the
trajectories and try to understand the breakup of
And there were two targeted reentries.
The ones above that were all basically, in a sense,
brought down just by the atmosphere itself. In other
words, the atmosphere drags things out of orbit slowly.
The last two were actually targeted into ocean areas
because of potential hazards they posed. The Compton
Gamma Ray Observatory, that was targeted to an ocean
area. There was no debris found from that one. And
then the Mir space station was also targeted to an
ocean area. The only debris I'm aware that was found
was reported by a guy who was beachcombing down in Fiji, a job I'd like to have. He did have one piece.
It's not been substantiated that it actually came from
Mir, but likely Mir had debris surviving and it may
float up on a beach somewhere.
Next chart. The type of work you can do with a reentry as far as reconstructing what actually happened to it, there are a number of things you need to do. There's maybe tracking data for example, radar data. Video data, for example, the type of thing that people would take. If people have seen it from aircraft, any of that data can be very useful in rebuilding what's happened in a reentry break. Public sightings of witnesses. On most of the reentries we've got here, the public actually has seen some of these things coming down. That information has been very useful in rebuilding what happened in the reentry.
Debris on the ground. Knowing where things are, what they look like, how much they weigh all that information can be critical to rebuilding what happened. That's one of the reasons why the work that's going on now, both from the public and other agencies looking for debris, is really critical to this investigation.
Data on the original vehicle. It's one
thing to have debris on the ground, but you need to know what the original configuration was like.
Sometimes we simply don't have good information on
that. If it's a foreign satellite or something like
that, we may not know exactly what was coming through
the atmosphere. So we don't have a good feel for
taking the debris back up.
The next thing you try to do is fuse all
that information and basically rebuild the reentry
trajectory, try to match the impact locations to
possible release points and take any existing weather
data, any of that sort of thing in, and then finally
conduct metallurgical analyses on the debris to
estimate temperatures, really look at what went on,
those kinds of things.
Next chart. This is an example of a
reentry. This one came down over Canada. This was in
1997. You can see that on that chart we show a breakup
altitude at the magic 42 nautical mile number. And
there are some fragments. We'll talk more about those,
but this is one. This again, the public was out. This
was about 3:00 o'clock in the morning. There were
reports to news stations and so forth, and we actually
used that information.
Next chart. This is some pictures of the
debris recovered from that one; and this is one of the larger debris fields, I guess, that we've actually had
a chance to see. As I say, typically unless it lands
next to a farmer's house as you see in that chart
there, people don't find these things unless they
happen to be out and about. So what you see in the
upper-left corner, this is the original configuration
as it was being loaded onto the launch vehicle up
there. There actually was a satellite on top of that.
This stage was released in orbit and was in orbit for
about nine months and then gradually the atmosphere
dragged it down.
The big brown tank you see over there is
about a 575-pound stainless steel tank. It landed
about 50 yards from a farmer's house here in Texas. He
was not pleased. The woman you see on the top right
actually was brushed on the shoulder by a piece of the
debris. Again, she was very lucky; but it's a very
The sphere you see down here was one of
four on that vehicle. That was the only one found,
although we believe they all survived. So they're
still on the ground somewhere.
The bottom one just shows that these
things can survive in pretty good condition. Those are
screws that you actually could unscrew. They held an aluminum plate onto the tank itself. The aluminum is
gone, but the screws were still there and just fine.
Next chart. This again gives you a
little detail on that one. Again 550-pound tanks.
67-pound titanium sphere. 100-pound thrust chamber.
Footprint length was about 400 nautical miles on this
Next chart. This is a detail of the
trajectory reconstruction. The trajectory comes in
from the top and each of those little black dots is
about two seconds apart. So you can see just by the
spread of those dots that it's moving at a pretty good
clip originally. That's up and around 18 nautical
miles up. When you get down to around 10 nautical
miles, it looks like it does a little dogleg there and
that is due to wind. So basically where an object of
this type comes into the atmosphere, typically all the
orbital and all that motion is gone, the atmosphere has
basically taken that energy out, and it will fall from,
say, 50,000 feet straight down. That's one reason why
when you see debris on the ground, even on the pictures
of the farmer's house with the debris there, you'll
notice there's really no crater. Most people don't
realize these things just fall straight down and they
just land. That's just a characteristic of this. That little dogleg is again caused by winds. It hit the jet
stream, and it blew it over. This, again, was a
575-pound tank. So you can see that even that can be
One of the things that we did was we were able to get a portion of this fragment that brushed Lottie Williams on the shoulder and we actually wanted to find out if, in fact, it did come from the launch vehicle or from that vehicle. We analyzed that and found that if you take the next page that it did. The trajectory time was consistent. She was out walking at around 3:30 in the morning and actually saw the reentry and then this thing came down and brushed her on the shoulder and she recovered that. We did get a piece. We brought it into our labs and did an energy dispersive X-ray analysis of it. There are actually two on this little red chart you see here. There are actually two lines there. One is the original material, and the second is what was recovered. So we are very confident that this material actually came from that vehicle.
Next chart. The second thing we did is take samples from the large tank itself, put it through a metallurgical analysis. We found, for example, that in portions the aluminum actually combined with the stainless steel and that we were able to use that to pin down the maximum temperature on the tank between 1200 and 1280 degrees centigrade. The other interesting thing, and I'll show you another example of this, is that it appears that this aluminum splashing back again, aluminum is there on other parts of the structure that the aluminum splashing back on the tank can actually oxidize or burn and the heat released by that can melt holes. We believe that's why the hole was actually melted in this tank.
Next chart. Just to show you, this is
not all that unusual an event. This is some pictures
of basically the same debris objects. These came down
near Capetown, South Africa, in April of 2000. So
basically the same objects.
Next chart. This is another one we have.
This is a solid rocket motor stage that came down in
Saudi Arabia. This one is made of titanium, which
makes it a little unusual. The ones you saw before
were typically out of steel. This is titanium. It
would be expected to survive very nicely. We have
evidence again that the hole you see here was actually
burned, in a sense, in the casing as the aluminum
oxidized on it.
Next chart. So just learning a little bit from the debris and limitations there, we typically model reentry breakup at the macro level. We simply don't have a good understanding of what happens at the micro level with these kinds of things simply because we don't have a lot of data to base our models on at that level. We do have a few reentries where significant debris is found; but, just by way of information, of the stages that down in Texas and South Africa we have about ten of those that come down a year those are the only two we've found, only two where debris was found. So most of these are laying in water or in places where not discovered. We also see about a hundred reentries of major objects a year. So finding debris on the ground is very unusual, although we do get hits on our website. People E-mail with things they have found and ask us if that potentially is of that type.
Just by rule of thumb, we would estimate
that about anywhere from 10 to 40 percent of an object
will actually survive reentry and that depends on what
it's made of. If it's got some big, heavy, empty fuel
tanks, that can really be a factor there. There has
been relatively little work on reconstructing reentry
breakup events. The ones I've mentioned are about all
there are. Again, one of the most important features is there's really been no systematic retrieval effort
except in a couple of cases. I guess the Cosmos 954
would an exception and, again, the objective there was
to recover the radioactive material.
Next chart. Some observations. As I
mentioned, the heating to an object can really be
exacerbated by burning of other material. For example,
this phenomenon of aluminum melting and splashing back
and the heat of oxidation actually increasing the
temperature and burning holes, we believe that's a real
situation. There are large aerodynamic deceleration
loads, and also you've got an object that's already
been fairly well heated as the reentry progresses. So
that can lead to structural failure and actually can
mask other information about what happened during the
Combing data from multiple sources can be
critical for reconstructing a reentry event. Finally,
the distribution of debris on the footprint may
actually be very useful in providing clues on the
breakup sequence itself. So things like if you find
objects early in a trajectory, that can be really very
critical to seeing how that reentry progressed.
Next chart. So, in summary, reentry
breakup is not well characterized at the micro level. That breakup and subsequent disintegration can and does
destroy clues of critical events. The debris field may
be very useful in helping to track down what ultimately
happened. Data fusing is really a critical part of
this. You really must take everything that you can
learn, all the data you can get, and try to reconstruct
what the event was. Then a final piece of that is
laboratory analysis of the debris pieces themselves to
look for things that can be shown to have occurred
earlier or have been protected by other objects as the
I think that's my briefing.
ADM. GEHMAN: Thank you very much,
Dr. Ailor. All right panel. Let me know if you've got
MR. HUBBARD: Dr. Ailor, thanks for being
here with us. We appreciate someone of your expertise
speaking to us. I have two questions that are
follow-downs on some statements that you made. One is
about the percentage of material that's been recovered
in your previous data base. Where we are today with
the Columbia is something on the order, by weight, of
15 to 20 percent. So I would like your assessment,
based on what you know, of whether you think this is a
low or a high or what we might expect in the future.
DR. AILOR: Well, as I mentioned for
typical reentries we see between, say, 10 and
40 percent. It really can depend on what materials the
object is made of. There may be significant debris
pieces that have yet to be discovered, I don't know,
but I would say that's certainly in the range of the
experience in the past. The other part of this is that
we've never had the detailed look or the energetic
search for debris that we're seeing now. So it's
possible that you may get a higher percentage as time
MR. HUBBARD: Thank you. The other
question was related to your statement about aluminum
rarely being found on the ground. We're finding some
aluminum, although mixed with other debris or attached
to other debris. Can you give us a brief explanation
of why that might be the case?
DR. AILOR: Yes. Our experience has been
that unprotected aluminum will not survive a reentry.
The heating is just too high. It typically comes off
very early in the trajectory. We do find aluminum,
say, bits of aluminum that has been flowed back on to
tanks and been protected, say, by a titanium sphere or
something like. It will flow onto the lee side and be
protected back there. But we typically don't find that. For debris that you're finding now, most likely
aluminum on the ground was protected for a significant
part of the reentry and probably was released late,
when there was sufficient heating to cause it to melt.
ADM. GEHMAN: Thank you.
MR. TETRAULT: Dr. Ailor, one of your
charts talked to the five satellites that had broken up
in the atmosphere. To put this in perspective, could
you tell me how many total pieces of history have we
had compared to the 30,000 pieces that we are now left
with on the shuttle.
DR. AILOR: Well, the history, we
actually have examined probably five or six, just to
give you an example, the several big tanks and so
forth. There was a number of debris pieces that were
picked up from the Cosmos 954. I would say in history
we're probably talking about in the order of maybe 250
or so that have actually been noticed by humans on the
ground and reported.
MR. TETRAULT: One follow-up question.
You talked about the ballistic coefficient. For
everybody's edification, could you kind of distinguish
the difference in the ballistic coefficient between
something like a tile, a tank, and maybe a landing gear
DR. AILOR: Absolutely. Ballistic
coefficient is a measure of how significantly the
atmosphere is going to affect the flight of an object.
The way to think about it is a very low ballistic
coefficient object would be like a feather. Extremely
low ballistic coefficient. A shuttle tile, for
example, released by itself, very light object, would
have a very low ballistic coefficient, as well.
Something with a medium ballistic coefficient would be
something like a tank, an empty fuel tank. That big
tank I just showed you here has a ballistic coefficient
on the order of 15 to 20. Then something like you were
mentioning, a landing gear strut, I probably would
imagine that would be up to 40 or 50, something on that
order. A ball bearing would be something that would
have a high ballistic coefficient. So it would be
something where the aerodynamic properties really would
make it less susceptible to the atmosphere and also its
mass properties would give it a lot of inertia.
MR. TETRAULT: Thank you.
ADM. TURCOTTE: In the examples that you
gave of the different reentries that you had, they were
obviously at different inclinations and they were at
different reentry profiles. Would you kind of explain
the difference in what you know of the shuttle's reentry profile at that inclination and some of the
data that you've had in the past from the other
DR. AILOR: Yes. The other satellites that I spoke of either were deorbited or basically were orbit decayed down, had very shallow path angles typically. They flew what we call ballistic trajectories, which mean there really wasn't much lift involved with them. Of course, the orbiter is a lifting object and lift did play a big role in its trajectory for a good portion of it, anyway. That trajectory will affect the heating rates and so forth and will affect how the object responds to the atmosphere.
MR. WALLACE: This is the first time
we've had a breakup of a vehicle designed for reentry.
Is that a fair statement?
DR. AILOR: Of this type, yes.
MR. WALLACE: This ballpark, your 42-mile
estimate, was pretty close, given the situation of the
Columbia. Does the fact that this was a vehicle
designed for a safe reentry change some of your
estimates about percentage we're likely to find and any
other sort of effect on the breakup sequence?
DR. AILOR: Well, it certainly could. As a matter of fact, the fact that there is a heat
shielding on at least a portion of all the body for a
portion of the time and then some of the body parts
after that will affect what survived. That's certainly
true. I should mention that the shuttle external tank
also is a reentering vehicle after it's released from
the orbiter during launch. That typically breaks up at
a slightly lower altitude, maybe around 40 nautical
miles plus or minus a little bit. What happens there
is there is some amount of heat shielding and it does
protect it for a little bit. So there are objects
where there is a shielding existing and I think the
fact that the breakup sequence that can be shown that
there was a material loss at a very high altitude for
the orbiter may indicate that the heat shield may have
had a problem.
DR. WIDNALL: You mentioned earlier that
aluminum rarely survives, certainly in its bare state.
Could you sort of go over all of the possible things
that you could think of happening to aluminum in
reentry both for, say, an individual panel that
suddenly found itself all by itself in the atmosphere
and also maybe a panel, say, on the leading edge, like
leading edge spar of the shuttle wing, that was
attached to the shuttle but was bare? What are the different range of things that could happen to such
DR. AILOR: I'll give you an example. Some of the work we've done has been looking at a large spacecraft that reentered with solar panels and we believe and have data to indicate that the solar panel came off early in that reentry. Some of data we have makes us believe that that solar panel, even with an aluminum structure, actually survived. So that's a case where again you have a big
DR. WIDNALL: Now, that's ballistic
DR. AILOR: That's exactly right. it's a
big, flat, plate. It spreads out, stops quickly, and
then essentially just falls to the ground. So
something like that could survive. So in that case
aluminum could be expected to survive.
If aluminum is being carried along by a heavy object for example, you saw the tanks we have here these were big, solid pieces of material. The aluminum on it is a piece of structure. As it's being carried by that heavier object, it's really governed by the aerodynamic and heating and so forth that's characteristic of that object. That could be much higher than the aluminum itself can stand; and when that aluminum gets weak, it will come apart.
DR. WIDNALL: I'd like to go beyond that.
So you're saying melting?
DR. AILOR: Melting. Absolutely.
DR. WIDNALL: Vaporization?
DR. AILOR: Melting, yes. Turn into
DR. WIDNALL: Well, droplets? How about
individual atoms, vaporization?
DR. AILOR: I would assume. You'd have
to ask somebody more qualified in that area than I am.
DR. WIDNALL: Oxidation?
DR. AILOR: Oxidation for sure. We've
seen evidence of that.
DR. WIDNALL: Of course, another word for
oxidation is burning.
DR. AILOR: Exactly.
DR. WIDNALL: The example you gave was aluminum deposited on another tank which essentially burned and created but I suppose it could also burn all by itself.
DR. AILOR: It could, although aluminum
released by itself probably would stay in a droplet
form and be sorry pretty quickly. So aluminum that
would be carried along by something I think would really be more likely to see that.
MR. TETRAULT: In the hole that was
created that you talked about, was that created by the
aluminum burning or the alloying effect?
DR. AILOR: It was, we believe, by the
oxidation of the aluminum itself; and that raised the
temperature up where you could actually see the
ADM. GEHMAN: I was very interested in your comment about the ball of paper being thrown out the window of the car not just because that's my level of understanding. What you suggested was that in an entry scenario like we're investigating here, there is a heating and an aerodynamic force, one of which is extraordinarily fast, and then when the object then becomes free and floats down to earth, it's still got heat but it's no longer of this extraordinarily short-period high intensity. My question is: When we go looking through debris, should we be able to detect those two phenomena that is, a piece of metal which has been flash heated versus a piece of metal that's been subjected to prolonged by prolonged I means tens of seconds or maybe even more? Can you see the difference, in your experience?
DR. AILOR: For aluminum to actually see, as you say, the flash heating, the way that will work is that when an object is actually kicked off, if it's has got material attached to it for example, it's tile material with some substructure attached to it if it comes out in a way where the tile material is forward and actually protects the material behind it, then that might be likely to survive. The problem is going to be with, No. 1, the breakup process is going to continue on about anything, about any object that's put out into the stream that's going to continue to see heating for a short period of time. If there is much material there and it's a very low ballistic coefficient item like a big, flat plate with some material behind it, structural material, that will heat up very quickly, as you say. The aerodynamic loads will also be quite high as soon as it hits the air stream. That can have a tendency to fracture it further. So this breakup process is going to continue as it comes down. Secondly the dynamics may actually get into the process. So this object is tumbling. Then the different sides will see the air stream. So it will be a difficult process, I think, to try to see a piece on the ground and make sense out of it from that perspective.
ADM. GEHMAN: I take it in one of your viewgraphs, for example, of a sphere that came from one
of the Deltas or something like that in which all of
the burn marks all around the sphere look approximately
the same, would it be, in your experience, safe to
conclude that that sphere had been tumbling and all of
the sides had been subjected to the same amount of
heat, whereas the one that had the hole burned in it
it's safe to analyze that that was another event of
some kind? That's kind of what I was getting at.
DR. AILOR: That certainly can be.
You're right about that. As a matter of fact, on one
of the Delta tanks, one of the spheres, about a 2-foot
diameter sphere, one side actually does have droplets
of aluminum that are clearly visible on it. The other
side is absolutely clean. So you can say that during
the heating phase that one side was facing the oncoming
air stream and saw more heating than the other side
ADM. GEHMAN: Another question. Certainly in the case of those spheres and by the way, in the case of Columbia, I'd ask, Mr. Tetrault, we have found essentially 20 out of 25?
MR. TETRAULT: We found at least 25, not
counting fragments, out of approximately 30. I don't
know what the exact count is from construction.
ADM. GEHMAN: (To Mr. Ailor) As you
predicted, the spheres all survived. But in the debris
field, not discounting the spheres, your suggestion is
that in the terminal velocity, in the terminal vectors,
even when you start off going 10,000 miles an hour, by
the time you reach the thick part of the atmosphere,
you're essentially dropping vertically.
DR. AILOR: Correct.
ADM. GEHMAN: Therefore, how would you
characterize whether or not we should find buried
debris or not? Would you expect most of the debris to
be on or near the surface?
DR. AILOR: I would expect most of the
debris would be on or near the surface. Buried debris
would not be typical for a spacecraft reentry. That
would require a very dense material and would also
require it to have some aerodynamic properties which
you're not going to find on a reentry object.
GEN. DEAL: Dr. Ailor, I've got a few
questions for you. You're probably aware that from the
to the fourth day on orbit this piece of debris that
was separated from the shuttle and that went on to
reenter, we have some extensive analysis going through
testing at Wright Patterson Air Force Base right now,
trying to determine the radar characteristics of it. Are there any type of predictive methods that you know
of that might tell us, by the characteristics of its
reentry, what type of material it was?
DR. AILOR: Certainly if we had
information on the reentry itself, yes. On the rate of
decay, the rate of decay from orbit would be indicative
of the overall aerodynamic properties of the object and
its weight. So that would be some useful information
to have. If there's tracking data, for example, on the
reentry itself, that could be useful.
GEN. DEAL: Then a second question. I looked at your slide that said from a Saudi Arabia reentry back in 2001, analysis is still ongoing, which doesn't bode well for us to get back to our day jobs anytime soon two years later. Can you tell us what we can expect to find through laboratory analysis of the debris in the short term?
DR. AILOR: In the short term, the
critical thing, I think, is going to be to try to
center the analysis on certain debris pieces that
there's some reason to believe have high value. What I
mean by that is if there's debris that can be
determined by analysis to have come from a particular
part of the vehicle itself, that's of interest. Then
you should really focus on that. I think the initiating event is probably what is of interest here.
So a lot of the final debris that is in the debris
field will have happened well after the initiating
event. So the search that's going on for early debris
is really very intelligent and the right thing to do.
The other thing would be to look for the
debris itself and see again if there's characteristics
of the field that would indicate that debris in this
area, for example, came from a portion of the orbiter
of interest. So I would really try to focus on that.
Laboratory analysis? There's too much debris here to
be doing that extensively. So it's going to have to be
DR. WIDNALL: Why do things tumble in the
atmosphere, and is there a possible diagnostic use of
measurements that appear to show something tumbling?
DR. AILOR: Well, even in orbit, things
can tumble. For example, as you come down from orbit,
you know, there's still a little bit of atmosphere up
there and so as you get into the portion where there's
enough to actually affect the dynamics of an object and
have that become a more principled player, it will
gradually overpower the gravity gradient forces which
are there and try to stabilize the spacecraft. That
interaction then will cause an object to tumble.
As you come down through the atmosphere,
the mass properties and aerodynamic properties of an
object will also cause it to tumble. We certainly see
that. As to whether or not things like tumble rate
could be a factor? It certainly could be, but you'd
have to know a fair amount about the aerodynamic
properties, about the geometry and other properties of
the object to be able to determine that, I think.
MR. HUBBARD: I'd like to pursue a little
bit more the question of how we might be able to
determine the initiating event and distinguish that
from the process as it may have happened post breakup.
In your experience, would you say that from
directionality of, let's say, a deposition of molten
materials or the way the surface had been worn away by
heat, we could begin to separate the two? Would that
be fair characterization?
DR. AILOR: Certainly could be. For
example, the orbiter was controlled for a good period
of time and if evidence is found that could have
occurred during that period and it indicates that a
particular flow pattern or something like that, I think
that could be very useful. Absolutely. I think the
early debris would be really critical to an analysis
MR. HUBBARD: Even from debris on the
ground, following the discussion of ballistic entry of
a steel strut, if it's worn away sort of equivalently
versus something that shows that's there's more
deposition or thermal damage on one side or another, it
might be a distinguishing characteristic?
DR. AILOR: It certainly could be.
ADM. GEHMAN: Sir, based on your analysis of previous satellite reentries I don't want to put words in your mouth, but let me make sure I understand it your suggestion there on kind of your first viewgraph was that the typical reentry, the process starts rather slowly and little things come off but then it reaches some catastrophic point where everything flies apart. I have got that right?
DR. AILOR: That's basically correct.
ADM. GEHMAN: And that is not an unusual
scenario, doesn't indicate a design flaw or anything
like, it's just that aerodynamics and heating of the
things reach a point where it can't tolerate it?
DR. AILOR: Exactly. And basically when
the disintegration process starts, it is typified by
not a violent event exactly but you can call it a
catastrophic event where the spacecraft really comes
apart into a number of portions and then from that point on, an observer on the ground would essentially
see a number of objects proceeding through the sky.
MR. TETRAULT: We've struggled, like
everyone, with how do you separate out reentry heating
from the event itself; and our plan is to really look
hard at the differences between the right wing and the
left wing. I would assume that you would agree that
that's probably a good approach in trying to look at
the differences between the two?
DR. AILOR: Yes, indeed, I would.
MR. WALLACE: In the civil aviation field
where I usually work, we often have the challenge of
differentiating damage that may have precipitated a
failure event in the sky or damage that was sort of
part of the failure sequence versus what was impact
damage on the ground, often very critical distinctions
to be made; and, of course, here we add in the thermal
effects. What are your sort of thoughts on the basic
methods you can use to sort those things out?
DR. AILOR: Well, as you say, the
challenge here is going to be that the heating itself
is going to have the potential of masking the heating
and loads during the breakup process; and as an object
comes down and continues to break up as it enters
atmosphere, it's going to have this tendency to mask the initiating event. That's going to be really the
challenge here. That's why I think that the effort
really needs to be focusing on the early debris and on,
as you say, the differences. If there are scenarios
that would cause differences in some of the debris,
that would be very useful to know about. Thirdly, to
focus on surviving objects which can be traced back to
areas of interest by one fashion or another.
MR. WALLACE: Has there been anything
generally in your observation of the Columbia debris
distribution and recovery process that has sort of
DR. AILOR: Well, I've been pleasantly surprised by the efforts that's been made to really recover the debris pieces and get specific information on those things the weights, the latitude and longitudes of those. The amount of effort that's being put into it, I think, is not really characteristic of these kinds of events and may be useful. So I would say I've been very pleasantly surprised by that.
ADM. GEHMAN: Dr. Ailor, the two most western pieces of debris that we've found both have been tiles, either a fragment of a tile or an individual tile, not connected to a metal or any structure. My understanding is you are suggesting, then, that a tile would have a relatively low ballistic coefficient
DR. AILOR: Right.
ADM. GEHMAN: and therefore the flight path is nearly vertical?
DR. AILOR: Well, certainly ultimately
will be vertical, yes.
ADM. GEHMAN: What I mean is compared to
something with a high ballistic coefficient.
DR. AILOR: Yes.
ADM. GEHMAN: Backtracking into space,
then, it would be safe to assume that these things,
these tiles came off relatively close to where they
were found on the ground, compared to a dense object?
DR. AILOR: Yes. That's exactly right.
ADM. GEHMAN: The fact that all the dense
objects that we've found we've found a couple of
hundred miles down range, what I'm trying to do is
rationalize in my find the dichotomy between something
with a low ballistic coefficient that comes off late
versus something with a high ballistic coefficient that
comes off early, because you could have them found in
reverse places on the ground.
DR. AILOR: Well, a lot of that will
depend on the timing of the release, too. If you've got something that's released at a very high altitude
early in the reentry and it has a very low ballistic
coefficient, as you said, that object will, in essence,
stop very quickly and flutter to the ground. It's
complicated by the fact that if it was simply a tile
that came off, that's one thing; but if it was actually
bringing something else with it, then there may be more
going on there. That other piece of material would
have probably increased the ballistic coefficient a
little bit, which would make it blow a little further
As you saw from the footprint chart that
I gave where it had the multiple footprints there, the
altitude and what the trajectory looks like as it
begins to heal over there will really affect how things
fly; but there can be low ballistic coefficient pieces
that are released all through that process. So some
will be carried further because they're attached to
heavier debris. Some will be released and then flutter
to the ground. So as you move forward in time, the
footprint becomes much more complicated.
ADM. GEHMAN: Another question. You
mentioned the inability of aluminum to survive reentry
for one reason or another. It either burns up, melts,
oxidizes, vaporizes. What is your experience with rubber? We have found five of the six tires, and maybe
a fraction of the sixth. We have found five of the six
tires, two or three of which actually look like tires.
DR. AILOR: Well, in the first place,
I've never seen a spacecraft come down with rubber on
ADM. GEHMAN: You've probably never seen
one with wheels either?
DR. AILOR: No, never.
ADM. GEHMAN: You've never seen rubber in
DR. AILOR: I haven't. I'm sure someone
could take a look and basically say if rubber
experienced heating of this type, how would it be
expected to respond. Some materials can be protected
by the fact that they actually shed away external
layers, for example, ablative materials that are used
on the spacecraft reentries typically. So it may have
properties that would enable it to survive of that
ADM. GEHMAN: Very good. This debris field that we have here I think you're familiar with. We're talking about just west of Dallas to just over the Louisiana border, which is about 375 miles or something like that. Are you surprised it's that small or that big, considering that, I guess, the first shedding event that we know about was at about 225,000 feet actually we're going to find that out here in another 20 minutes or so. Right. You had a viewgraph up there that indicated in one of these reentry things it was spread over 400 miles. What do you conclude for this one?
DR. AILOR: That footprint I was talking about was from the little piece that actually brushed the lady on the shoulder. Very low ballistic coefficient piece, probably less than 1 so it was something that, in fact, did flutter down to the fairly large objects which were ballistic coefficients up to around 50, 60, something like that. So those are a reasonable range of ballistic coefficients.
The size of the footprint here is about
what you would expect to see, I think.
MR. WALLACE: You were very complimentary
of the amount of shoe leather that's gone into this
recovery. Do you expect that any further major
breakthroughs or strokes of luck are more a matter of
shoe leather, or are there calculation methods you
think might be further explored, backtracking pieces
you have found?
DR. AILOR: Well, there's a couple of things. First, I think the work that's going on relative to finding the debris is really an important part; and that has to be emphasized. That's going to be key to solving this puzzle, I believe. The second part would be to look at the debris field itself, but you have to have collected debris in that field. So this idea of going out and finding these things, I imagine that pieces will continue to be found over a period of time and they need to be cataloged and brought in and examined just as they are being now. But to really look for anything that's relate to, as I've mentioned before, possible scenarios for example, the right-wing-versus-left-wing scenario and those kinds of things. So I think that's the way it should go.
MR. HUBBARD: One last question for me at
least. Looking at your observations and summary, you
bring up the concept of data fusion here. I wonder if
you could elaborate on that a little bit. What do you
really mean there?
DR. AILOR: Well, basically the data
fusion means that, for example, where we have videos
that have been taken by private citizens, taking those
videos, processing those things, we know the orbiter's
trajectory very well during portions of reentry, in a sense, fusing that data so you take the video data, you
marry it with the trajectory data so you know exactly
what you're looking at. You can use that information
to help derive information about, when objects are
shed, where are these objects, what the timing is, what
are the characteristics of those objects, things like
that. We talked about ballistic coefficient; but you
can estimate, based on how fast something separates
from the orbiter in a video, what the characteristics
of that object are. So that's what I mean by fusion,
just taking all of the existing data and bringing it
all together so that you actually have a complete
picture, as good as you can do with the data you've
got, of what actually happened.
MR. HUBBARD: Would you include
thermodynamic analysis, you know, reentry heating in
addition to these actual empirical observations?
DR. AILOR: Yes, I think that's true; but
the fusing I'm talking about really is more of a
trajectory level, if you see what I mean. There's
certainly other data. The data on the ground, for
example, needs to be brought into this, as well, and
should be. So it's really a question of fusing the
various data. I come out of the trajectory side of the
house. So looking at data from where things happened in the trajectory, tracking them down, trying to derive
information on the ground, and then really developing a
best estimate of what actually happened is what I'm
ADM. GEHMAN: That leads to my last question that is, if you would, make a value judgment for us on the accuracy and efficacy of this reverse trajectory analysis. In other words, if you find something on the ground, how much effort and what value should be placed on trying to predict the point in the sky that this thing became an independent object? If you would, take a shot at that.
DR. AILOR: That is going to be a real
tough problem, quite frankly.
ADM. GEHMAN: You mean because it's just
not an accurate process?
DR. AILOR: It's not an accurate process.
As I mentioned in my opening remarks, we don't have
internal information from a spacecraft that's breaking
up as to what exactly is happening with it. So
modeling it down and doing computer models of the
reentry and breakup of a spacecraft, we recognize that
there's uncertainty in there. The problem with taking
debris on the ground and transferring it back up is you
don't really know how it got here. There will be debris on the ground that will be surprising, very
lightweight things, things that in a sense could burn
very easily but may have actually survived and impacted
the ground. Those objects we know were shielded,
because they wouldn't have gotten there otherwise; but
where it was originally in the vehicle and then the
scenario that it followed for shedding the various
layers of material and the changes in the aerodynamic
and mass properties of that post object as it came
through the atmosphere is going to be a very tough
thing to derive. That's why I think that really a key
here is to look at the early debris as closely as you
can to really try to determine what really happened
prior to a lot of that breakup process going on.
ADM. GEHMAN: Of course, it's probably a variable once again, I don't want to put words in your mouth. For example, if you were to tell me the ballistic coefficient of a sphere, a fuel sphere, I bet you could pin that ballistic coefficient pretty well; but if it was a piece of debris or a jagged-edged thing that was part tile, part metal, part strut, part bar, the ballistic coefficient might be a pretty big estimate?
DR. AILOR: Yes. In fact, again, the
ballistic coefficient of what you actually find on the ground was different at say, 75,000 feet or
100,000 feet or 120,000 feet. So the higher up you
get, the bigger the changes, if you're talking about
going backwards in time. So what you find on the
ground is one thing, but trying to translate that back
up and say, okay, well, we know it fractured off of
something, what was that? We don't quite know what
DR. WIDNALL: From a forensic point of
view, what are some of the most interesting
observations that you can imagine making on the debris?
The second part of that is does Aerospace Corporation
have any metallurgic capabilities to help us analyze
some of the observations we make on this debris?
DR. AILOR: We do have, and we have
analyzed some of the debris in the past. So we have
some experience of doing this work. The kinds of
things that, again, will be important to look for here
are opportunities, if you want to call them that, for
preserving some of the original events. That could be
where material is found, either heat shield material or
something like that is found from areas where it likely
came off and protected some evidence of the original
events, that would be really critical. So I think it's
going to be a question of looking for objects on the ground where it's likely that some of the original
evidence from the original burning or fragmentation
would be preserved, perhaps behind the wing leading
edge or behind tiles, those kinds of things.
ADM. GEHMAN: Thank you very much. Would
you like to have the last word? Any advice for us on
how to solve this riddle?
DR. AILOR: No. It's certainly a tough
problem, but I think the advantage here is that there's
been so much interest by the public in actually helping
to gather debris pieces. I think that's really to be
complimented. We typically don't see that kind of
interest, and those debris pieces can really be
essential in helping solve this puzzle. So I think
that's really been valuable.
ADM. GEHMAN: Thank you. On behalf of
the board, we thank you for your appearance here today
and for summarizing what I know is a deeper and more
exhaustive study of the reentry physics and
aerodynamics. We appreciate your effort and want you
to know that we've learned from you and we'll see if we
can't solve this riddle with your help. Thank you very
The board will take about a five-minute
ADM. GEHMAN: All right. Board, we're
privileged to have two people who have been studying
this tragedy since the first day and know more about it
than most other people. Paul Hill and Mr. Doug White.
Gentlemen, before we start, we don't
swear witnesses in but we do ask them to affirm that
they're going to tell the truth and the whole truth.
So I will read a statement of affirmation to you and
ask you, if you agree with it, just say you will. So
before we begin, let me ask you to affirm that the
information you provide to this board will be accurate
and complete to the best of your knowledge and belief.
MR. HILL: I will.
MR. WHITE: Yes, I will.
ADM. GEHMAN: Gentlemen, we know you, but
for the record we would like you to introduce yourself
and say a few words about where you work and what your
background is and then we would be delighted to listen
to as much of an opening statement as you want to make.
PAUL HILL and DOUG WHITE
testified as follows:
MR. HILL: My name is Paul Hill, and I'm
a missions operations director here on the space
shuttle. I'm a space station flight director. I've been a flight director for about seven years.
ADM. GEHMAN: And you are currently what are you doing for the MRT?
MR. HILL: For the MRT I run a team
that's called the video sightings assessment team.
After Doug talks about the time line, I'll go into
great detail about what we do and how we do it. The
short answer is we're trying to make some sense out of
the public imagery and any external sensor data that we
can get our hands on to tell us what was happening to
us as early in reentry as possible and maybe shed as
much engineering information as possible on what was
going on with the vehicle before we knew what was
happening on the ground.
MR. WHITE: My name is Doug White. I'm a
director of operations requirements for United Space
Alliance. In my job I'm responsible for turnaround
test requirements at the Cape. I'm also responsible
for anomaly resolution. I'm also responsible for the
engineering support during missions. I do have the
time line to talk about today. As far as what I'm
doing on the mission response team, I am on the team
which we call technical integration team. Basically
our job is, from a management perspective, to try to
pull together all the different efforts of the different teams, the aero, the thermal, the scenario
teams, and try to make sense out of all the data from
all the teams and then try to bring a coherent story
ADM. GEHMAN: Thank you very much. Which
one of you is going to go first?
MR. WHITE: I think I'll go first. I
plan to walk everyone through the time line. If you go
to page 3 of my briefing, please.
On page 3, this is a graphic showing the
sensors that we're most interested in in the left wing.
This particular chart shows the sensors in the left
wing. There are a number of sensors in the wheel well
that we are interested in that we got data from that
behaved in an off-nominal way. There are also
temperature sensors out in the wing, some of which went
off line, which was off-nominal, and some of which did
stay on line, which also tells us things that were not
The different colored wires that you see
represent the wiring runs for those particular sensors.
The pink one is for sensors that were aft in the wing
and runs forward past the wheel well and then
ultimately into the mid body where some sidewall
temperature sensors, one of which has a yellow line coming from it, that indicates the wire run for that
sensor which was inside the mid body. There's also a
green and a gray wire run you see in the back there
that goes through a connector box and into the aft.
The green wire run is for sensor data from those
particular sensors indicated by green dots. Then the
gray wire run is for a power cable. It's a little bit
different than the sensor wires. This provided power
to the actuators and came from a box there which is
labeled ASSA4. That stands for air surface servo
amplifier. That basically provides electrical power
and commanding to the actuators for the elevons on the
back of the wing.
ADM. GEHMAN: Doug, before we leave that,
pardon me for interrupting. To what degree is that a
cartoon and to what degree is that a fairly accurate
representation of where the cables actually run?
MR. WHITE: It's kind of in between a
cartoon and fairly accurate. For example, the pink
wire does run exactly alongside the wheel well and does
turn and go in front of the wheel well and does run to
a connector right forward of the wheel well, as is
indicated there. So those are approximate locations of
where those wire runs. Now, in the back there we see
the green and gray and pink all together. Those wires may actually be separated in space by 1 or 2 or 3 feet.
This is looking down on the wing, and so you can't see
the actual vertical separation between these wire runs.
Just because they happen to show up on top of each
other in the picture doesn't necessarily mean that
they're bundled together within the vehicle.
ADM. GEHMAN: What's the little insert
MR. WHITE: I'm sorry, I forgot to
mention that. That little insert is for some sensors
that were forward on the orbiter. These are
temperature sensors on a supply water dump nozzle,
which is a nozzle used to dump excess water overboard.
Right below that is a temperature sensors for the waste
water dump nozzle, again used to dump waster water
overboard. Then there's another one forward which is
called the vacuum vent dump nozzle. Those sensors also
gave us some off-nominal readings. Since they were too
far forward to show in this scale, we just put them in
as a little inset.
MR. WALLACE: Just to follow on Admiral
Gehman's first question, are the orbiters different?
Are there variances in the actual location of the wires
in the orbiters?
DR. AILOR: There maybe slight differences between 102 since it was the first one built. 102 had a lot of wiring which was called development flight instrumentation, a lot of wiring for that. During its most major modification period, we removed a lot of that wiring. Some of it we just left in place. So the wiring on 102 was substantially different in the DFI aspect. But for the sensor wiring, it was pretty much the same
ADM. GEHMAN: DFI? Developmental flight
MR. WHITE: Yes. DFI, developmental
ADM. GEHMAN: I'm the acronym authority
MR. TETRAULT: Let me continue with the
wire questioning. We do know that there were actually
four cable runs running back aft that went around the
wheel well compartment, one on top of the other. Are
all of those sensors that you show going off in one
those runs or in all of those runs or some portion in
each of those runs?
DR. AILOR: All of the ones in the pink are all within one particular cable. We don't have the specifics about whether or not, for a particular part of run, any one of the wires was like at the back of that bundle or on the top of that bundle. There are also more
MR. TETRAULT: The question is: As I
look inside the shuttle wheel well door and look up,
there were four wire bundles that run aft?
MR. WHITE: Right. All of the ones in
the pink wire are in a single bundle.
MR. TETRAULT: Okay. Are the red ones in
that same bundle, the ones that went off in the aft
MR. WHITE: Yes, all of the ones that
went off in the aft.
MR. TETRAULT: So everything that went
off are in one single bundle?
MR. WHITE: Yes. There are also many
other wires, though, in that bundle for which we do not
MR. TETRAULT: Understood. Do we know if
that's the top bundle or the middle bundle or the lower
MR. WHITE: If I remember the picture
right, it's the upper one.
ADM. GEHMAN: But we'll find that out.
MR. WHITE: Yeah. And I can give you the
more exact answer. I'm just trying to remember it off the top of my head now.
ADM. GEHMAN: We'll go back to the
blueprints. Okay. Please continue. Thank you.
DR. AILOR: All right. Next slide,
please. This particular time is about 7 1/2 minutes
before loss of signal, at a GMT of 13:52, and all of
our sensors appeared nominal.
Next slide, please. Now, this slide we
didn't show any sensors going off line but we put this
in the time line. This particular time 13:52:05 is the
first indication that we had some off nominal from an
aerodynamic standpoint. We were able to derive
aerodynamic coefficients in yaw and roll which showed
us that we were flying differently than we expected to.
You're going to hear a lot more about that tomorrow,
but basically the way we have done that is to look at
the way we should have been flying, look at the way we
actually were flying, and take the difference between
the two and come out with some moments on the vehicle
both in the yaw and the roll. This particular
off-nominal event, it started first in the yaw
component. We are seeing a different yaw at this point
in time than we expected to see.
Next slide, please. This is our first sensor that we saw with a small rise, and I want to stress that this was a very small
ADM. GEHMAN: Excuse me for interrupting
again. If it's okay with you, we'll talk about these
things while you have them up.
MR. WHITE: All right. That's fine.
ADM. GEHMAN: This off-nominal
measurement we will talk about tomorrow when we talk
about aerodynamics. I want to get to the level of
detail that and your team have been going through. You
didn't realize this until about Rev 12 or Rev 10. Can
you tell me when this became apparent?
DR. AILOR: Well, fairly early on, the
aerodynamic guys knew that we had differences in the
flight control from what we would have normally seen.
They looked at the aileron, and the aileron was
behaving differently and continued to behave
differently throughout the entry. It took a while
before we could back out that particular moment in time
that we just went through there was the very first
indication that this derived yaw delta was first
affecting us at that point in time, but fairly early on
we were able to see some of the larger flight control
responses that were off nominal to us.
ADM. GEHMAN: I could look it up here,
but you may be able to tell me. We are approximately what altitude and what speed here?
MR. WHITE: I don't have those numbers.
There are versions of this that do have all those
numbers on there. I guess I could look it up, too. I
have some notes here.
ADM. GEHMAN: But we're approximately
MR. WHITE: That's about right.
ADM. GEHMAN: Okay. Please go forward.
MR. WHITE: All right. This is the first
sensor that went off line. This is a left main gear
brake line, Temperature D. It began a very slow rise.
We call it a bit flip, which is essentially one bit in
the data stream showed that it was rising. And we
believe this is the first indication that there was an
off-nominal event and something was going on with the
orbiter inside that was causing that measurement to
Going on to the next page, these are the supply water dump nozzles A and B that I talked about. There are three nozzles to the forward there the supply water dump; the vacuum vent dump, which is the very forward one; and the waste water dump, which is actually below the supply water dump. These nozzle temperatures A and B both began an off-nominal rise rate. If you look at the graphs, you'll see a very small knee in the graph where the two sensors are rising at a particular rate and then there's a bend where they start rising at a faster rate. This continues for about 15 seconds or so and then it bends back over and starts rising at the same rate that it had been before, at the nominal rate.
MR. WALLACE: This picture doesn't tell
you where that is, does it?
MR. WHITE: Well, again, that's why it
was an inset. They're very far forward on the orbiter,
just right at the beginning of the wing. That little
diagonal you see there is the very beginning of the
wing chine, and they're just aft of the crew module
portion of the vehicle. They're on the side wall.
We're just showing them on the top for visibility.
They're actually both on the side wall, just above the
MR. HUBBARD: Now, this anomaly is in a completely different place as you say, well forward. Is there anything that would lead you to believe that this is, in fact, a sensor malfunction, you know, something wrong with the box, the electronics box?
MR. WHITE: It does not appear to be. We
don't know of failure scenario that would explain this as a sensor malfunction. We think it is real data
showing us there was a change. Now, whether or not the
change that caused these temperatures to rise is
related to what ultimately caused our tragedy, we don't
know. They may be detected. So we're including this
in our data, and we'll continue to look at it until we
can explain it.
MR. HUBBARD: So you're including that
this is really data, from everything that you know?
MR. WHITE: Yes.
DR. WIDNALL: How anomalous was this
anomaly? Have you looked at early shuttle flights to
see if you had similar events?
MR. WHITE: For this particular
measurement, we did look at every single mission; and
every single mission, these vent nozzle temperatures
rise at a very straight, steady rate. So this is an
anomaly in that the rate changed; but it was a very
short duration, about 15 seconds or so. They were
rising at a higher rate; and after that, they went back
to their same nominal rate. So whatever caused them to
rise at this higher rate was a transient, at least
locally transient event.
ADM. GEHMAN: I'm just stating the
obvious here. Obviously this is pre-video here. We're out over the ocean?
MR. WHITE: Right. There is out over the
ocean. If you notice in the lower left, there's a
ground track trying to show approximately where we were
with regards to the ground tracking. We're still well
off the coast.
ADM. GEHMAN: So if something was going
on, we have no video of it.
MR. WHITE: Right.
MR. HILL: We are within a few minutes of
having our first video when we see this.
MR. WHITE: All right. If you go on to
the next slide. This is the vacuum vent, just a few
seconds later. It began its rise as well.
Next slide. Now we're back into the
wheel well. This is the left main gear brake line
temperature A. This is down on the strut for the
landing gear, and it began a very slow rise. Again,
all of temperatures in the wheel well first exhibit a
very slow rise rate. It wasn't until about two minutes
from now in the time line that they began a much more
rapid rise rate.
ADM. GEHMAN: We're both trying to do the
same thing here. We're trying to characterize the heat
in the wheel well.
MR. WHITE: Yes.
ADM. GEHMAN: Can you describe to me
exactly where the sensor is? Is it inside a block
that's measuring the hydraulic fluid temperature, or is
it up against the block where the sensor is out?
MR. WHITE: This particular one is on the
hydraulic line that's on the strut. So it does have
some exposure, fairly good exposure to the atmosphere
in the wheel well.
ADM. GEHMAN: So it's not buried inside a
great big block or something?
MR. WHITE: That particular one is not;
but, you know, there is a heat sink of the actual strut
itself. That provides some heat sink. Some of the
temp sensors down in the wheel, you have the heat sink
of the wheel itself. Many of the temp sensors that you
see lined up four in a row that are on the side wall,
some of those are actually under epoxy covers and so
would not have a good exposure to radiation or
ADM. GEHMAN: But this particular one?
MR. WHITE: This particular one would
have a fairly good exposure.
ADM. GEHMAN: Thank you.
MR. WHITE: Next slide, please. This is back on the side wall. Again, this is the left main
gear brake line temp C. Again, beginning a very slow
Next slide, please. All right. Now we
start to see things going on in the wing and we believe
this is directly related to some sort of burning or
disintegration of that pink wire run that's affecting
these sensors. The reason we believe that is because
some of the other sensors nearby them don't show any
effects and these sensor do start to show effects. So
we think it's happening away from where those sensors
It's showing that completely colored in.
It's off line. These sensors, we've done some testing
that when you burn through the wire, you end up with a
variable shorting, a variable resistance in the wire
and you start to see the same kind of trail-off in
time. It doesn't immediately just go off to its
off-scale low reading. So this particular sensor at
this time began to trend down.
Next slide, please. Then a few seconds
later that sensor was completely off line.
Next slide, please. All right. Here's
another indication that we put in the time line of
another off-nominal aero event. This is the first clear indication. We mentioned before that we had the
derived yaw moment showing us we're off nominal. At
this point we began to have an off-nominal roll
component to the aerodynamics.
Next slide, please. Again, this is
another sensor in the wing which began to trend down.
This is the hydraulic System 1 left inboard elevon
actuator return line temperature, and it began its
Next slide, please. Hydraulic System 3 for the left outboard elevon
MR. HUBBARD: Just clarification as we go
here. The ones that you feel fairly certain are
showing the actual wire damage, have you been able to
back up and reconstruct in the wire bundle what was
MR. WHITE: No, that's one of the things
that we don't know. The drawings are not specific
enough to allow you to reconstruct which wire might
have been on the outside of the bundle, if you will,
and which wire might have been farther back in the
bundle, which wire might have been right in the center.
We don't have that level of detail to know what the
placement of each single wire was within its larger
MR. HUBBARD: Is there a hope of reconstructing that from closeout photos or as-built drawings or anything or is that pretty much
MR. WHITE: No, we will not be able to
ADM. GEHMAN: Are the wire bundles
themselves encapsulated or covered other than the
individual wires being covered?
MR. WHITE: Individual wires, sometimes
you have like twisted shielded pairs and you have
shielding around those; but then once you make a larger
wire bundle, no, the wires themselves are not covered
with any kind of insulation. We do, for a lot of our
wire runs, put convoluted tubing around, that black
crenelated tubing that provides some impact resistance
for people working around the wire. That's made out of
a Teflon-like material and provides some impact
resistance, but it wasn't designed to provide any kind
of a thermal barrier or anything like that.
ADM. TURCOTTE: As you're talking about
all the wire here, all of this wire that you are
talking about is all Kapton wire. Is that correct?
DR. AILOR: Yes. This is all
Kapton-covered wire. Yes.
All right. We'll go to the next slide. This is the hydraulic System 3 left outboard elevon actuator and return line temp that actually finally went off line. As I said, it had begun its little it takes a few seconds for these things to go down. Some of the ones that I'll show you a little bit later actually took quite a while to go off line, which indicates to us that maybe they were shorting or that part of the wire was burning through more slowly at that point.
Next slide, please. This is back to the
system 1 on the inboard. That one has now gone off
Next one. This is hydraulic System 1 on
the left outboard. That particular sensor is now gone
off line. Again, as I said before, the reason we
believe that the damage is occurring away from the
actually location of the sensor is because you see that
green dot right next to it and that particular sensor
was not reading anything off nominal at that particular
time. So whatever was causing the damage was happening
Next slide, please. This is back to
Hydraulic System 2 left inboard elevon actuator.
Return line temperature again started its slow change
to going off line.
Next slide, please. Now we'll go back forward, and you notice that our supply water dump nozzles have now come back to their nominal rise rates. So whatever effect was going up front is now not there anymore and the supply water dump temperatures are back to their they're still increasing. That's nominal, the way they've been for every other flight.
Next slide, please. Then also the vacuum
vent nozzle also at the same time went back to nominal.
You can see at this point we're just now crossing the
California coast and just about to pick up video, which
Paul will talk to you about in a moment.
ADM. GEHMAN: Doug, the sensors back by the elevons, all of them back there I've got the same thing in front of me that you have. For the people in the audience, there's a time line, this little sliding scale across the top of the viewgraph.
MR. WHITE: Right.
ADM. GEHMAN: The first sensor. I'm talking about just the sensors that dropped off scale. The first one is 52:56, and then the one just before this you've said was 53:35. So essentially that scenario that you just went through with these five sensors, that happened in 40 seconds. By my arithmetic it took about 40 seconds, that little scenario you just went through. If we assume that you're right that the insulation of the wires were melted and they shorted to each other or shorted to ground or opened and by the way you should be able to tell us that, right?
MR. WHITE: Well, again, we haven't done testing so far to where we took we're planning on doing more tests to get a more representative case, but we took a wire bundle, we attached sensors to the end of it, we put a torch on it, and we looked at the characteristics of the sensors going off line, and they do look similar to what we saw in the vehicle. We do see them begin to do a slow decline, and then they eventually go off scale low.
ADM. GEHMAN: So just for my mental
picture, then this little scenario of whatever happened
in that wire bundle took about 40 seconds, according to
MR. WHITE: Yes.
ADM. TURCOTTE: Before we continue, could you explain the physical I guess the void that is the wing, is it possible, for example, for air to flow freely in there? Is it a sealed compartment? Could you explain that as you're looking at the sensors in particular, the relationship?
MR. WHITE: Let me see if I can explain a little bit. If you see the panels all along the edge
there of the wing, those are the reinforced
carbon-carbon panels or RCC panels. Behind them is an
aluminum spar that runs all the way down the length of
the wing. You see the vertical lines. Those are solid
aluminum spars with some cutouts through them that
would allow a vent passage, if you will. There's one
main vent passage pretty much where the pink wire runs,
which is how you get through those spars. The
horizontal lines are representative of rows of boron
aluminum rib struts which are basically tubes that are
there for reinforcing the structure of the wing. So
that area from up and down on the slide here would be
all open; but in each one of the spars, which are those
vertical lines, you're closed out except for some small
ADM. GEHMAN: And the wheel well?
DR. AILOR: The wheel well is completely
enclosed from the rest of the wing. There is a hole in
the very front of the wheel well that's about 5 inches
in diameter which would allow some flow into there.
There are some other drain holes and some small
openings around some of the hinge covers which would
allow a very small amount of flow out. The square area
of the hole into the wheel well in the front is about 19 square inches. The remaining holes altogether total
less than 1 square inch.
ADM. GEHMAN: So the forward bulkhead of the wheel well, there's a hole with a screen
MR. WHITE: Yes, it does have a screen on
ADM. GEHMAN: which allows kind of free communication into this what we call the glove area.
MR. WHITE: Yes.
ADM. TURCOTTE: So it's safe to say that
an air molecule, once inside the wing, is pretty much
free to flow around the inside of the wing?
MR. WHITE: Through the vent passages.
Right. Also there's another hole between the wing
glove area and the mid body that's forward, about where
that yellow arrow is. There's another hole in the mid
body there which is rather large. That particular hole
is about 146 square inches.
DR. WIDNALL: What is the material that
the wheel well structure is made out of?
DR. AILOR: It's made out of aluminum
DR. WIDNALL: How thick is it?
DR. AILOR: I do not know that thickness. We can get that for you.
DR. WIDNALL: Okay. But it's basically a
thin piece of the honeycomb and another piece?
MR. WHITE: Right. A thin face sheet,
some honeycomb material, and another face sheet.
Next slide, please. All right. We've
annotated the debris events. We are over California
now and we'll see in the videos from the public that we
were starting to see debris being shed from the
orbiter. This is the first one that we've seen in any
of the videos that have been provided to us. So we
call it Debris No. 1. The timing on that is plus or
minus 2 seconds, which is about the best we can resolve
from the video.
Next slide, please. Debris No. 2.
Next slide. Debris No. 3. Coming off
Next slide, please. You notice with the
little time hack up at the top there, we're starting to
put triangles below the line for the debris events.
The diamonds along the line there are for the
off-nominal sensor readings, and then the two triangles
on the top of the line are for the aerodynamic
readings. That's how you read that little graph up at
Next slide. This is the fifth debris.
Next slide. Okay. Now, we start to see
another temperature rise in the wheel well. This is
again also on the strut. Also should have some fairly
good communication with the flow of whatever is
happening in there. This is left main gear brake line
MR. TETRAULT: Can I ask a question about that? This one is probably the most confusing sensor for me personally. Line Temperature A went off and I notice that you appear to have changed the timing on this a little bit went off at about a minute to two minutes. Line Temperature A and B are about the sensors are about 2 inches apart.
MR. WHITE: That's correct.
MR. TETRAULT: At the same time, you have
D and C gone, which have significantly gone off already
early, significantly separated both in the X, Y, and Z
dimensions, which would tend to suggest that the entire
wheel well compartment is warm. Why do you see this
big, huge time lapse between A and B, which are
separated by 2 inches? Is there any explanation that
you all have come up with, or at least theory on why
there is this big separation in time?
MR. WHITE: Right now we do not know of a good theory that holds together that says why one would
show the rise and not the other. At about this time
now, the rises are starting to become significant. So
we don't have a good theory. It may be the amount of
heat sink, the way it was attached to the strut itself
that provided some more resistance to temperature rise.
We really don't have a good theory right now for why
one 2 inches away would rise earlier than another one.
MR. TETRAULT: It's significant in terms
of the time. A minute in this entire frame is a
MR. WHITE: Yes. One possible
explanation that we've been kicking around is the fact
that whatever the event is that is causing heating in
the wheel well might not be constant in the sense that
it's continuing to direct flow into the wheel well.
Perhaps we were directing flow in at one point in time
and through the dynamics of the vehicle through the
evolving change in the damage to the vehicle that the
flow was redirected to some other part of the wing for
a time and then came back.
MR. TETRAULT: You're talking about the
equivalent of a run-away fire hose kind of thing.
MR. WHITE: Something like that. I
wouldn't describe it quite that way; but, yeah, something like that where if you had some sort of a
plume heating into the wing that maybe it was pointing
one direction first and then another and then back
DR. WIDNALL: Given the extensive damage
that has already occurred to the vehicle at this early
time, I guess I'd question the use of the word "early
debris." I guess from my point of view I would call
them mid debris. I mean it's clear to me from the time
line that things must have fallen off in the ocean well
before California. And we don't know obviously.
MR. WHITE: Right. We don't have any
evidence of that. These are the first debris events
that we see. So we just started at 1.
DR. WIDNALL: But at this point you've
already got some kind of hole in the vehicle, you've
got a wire bundle that's either completely burned
through or burning through, you've started to pick up
what I call flow inside the wing. So clearly some
structural damage has already taken place, by whatever
MR. WHITE: Right. We do believe that we
had structural damage somehow at this point in time
that was allowing flow into the wing. Whether or not
we had shed any debris out over the ocean earlier, we can't say one way or the other. It would be
MR. HILL: We call them early debris to
distinguish them from the actual spacecraft breakup
DR. WIDNALL: I understand that.
ADM. GEHMAN: Doug, in your machine here,
you don't have the sister viewgraph?
DR. AILOR: I do, but they told me they could only project one at once. If you want to see the other one you're talking about for the vertical elevations between these?
ADM. GEHMAN: Right. If you could do one
of them. I don't know if you could do the companion to
this one or not.
MR. WHITE: Well, if they want to go
ahead and bring it up, it's called Part 2.
ADM. GEHMAN: Well, okay. Let's not do
MR. WHITE: Okay. We could do that. I
think they only have the capability to show one at
All right. Let's go on to the next slide. All right. You asked about how early we were able to see things. The start of the slow aileron trim change again, I put the triangle up on top of the line there this was one of the early aerodynamic things that we noticed. The two events that we talked about earlier took some time for us to back out or reconstruct. From examining the data shortly after the accident, this was one of the things that we noted pretty early in the data. So this is another aerodynamic event that's off nominal. We started to see a slow trim change in the aileron.
In the orbiter there is no real physical
aileron like you might have in an airplane. The
aileron is a theoretical difference between the elevon
position on one side of the vehicle and the elevon
position on the other side of the vehicle. So by
adjusting the relative different positions between
those, you can create the aileron effect. So that
aileron effect was keeping the vehicle flying the way
we wanted it to. So as the forces began to change on
the vehicle, the trim changed; and we saw that in the
MR. HUBBARD: Doug, I just want to check
and see that we're working from the same time line
here. What I've got is what's called Rev 15.
MR. WHITE: Yes. This should be Rev 15.
MR. HUBBARD: Now, you skipped past what are labeled "Unexpected Com Dropouts." Is that because
they are not part of the temperature sensor story?
MR. WHITE: When I was coming here today
and preparing for this, it was a question to myself
whether I should brief from the time line that has
every single event in it or I should brief from this
more graphical presentation which did leave some of the
events out. This particular graphical presentation
does not have every single event like some of the com
dropouts. To this point we've already had numerous com
dropouts that we consider anomalous. We just did not
model those in this particular graphical presentation.
MR. HUBBARD: So I guess the follow-up
question to that is: Where are the avionics boxes or
the antennas or whatever associated with those and can
you make any connection between this set of anomalies
and the com dropouts?
MR. WHITE: Well, we are trying to do
that. We are trying to create an entire picture where
we can explain events that would affect everything that
we see. So com dropouts would be one of the things
that we would try to explain. As for the location of
the actual avionics boxes, they're in the avionics bays
which are forward in the crew module; and the antennas
are in the crew module region, on the top and the bottom of the vehicle both.
MR. HUBBARD: So this is work in process.
MR. WHITE: So they're well forward of
this area where we're seeing the heating, but that's
not to say whether or not some disturbance in the hot
gas flow around the vehicle may or may not create a
situation that would cause the com to drop out. We
were at fairly good look angles between us and the
satellite. So we should have had good communication in
this region. We have looked at past flights. So we
did have good communication in these regions. So
again, that's why we consider some of these com
dropouts as anomalous events.
MR. TETRAULT: Have you seen any
relationship to the com dropout and the debris event?
MR. WHITE: I'd have to look at the
timing that says how close one was to the other, but I
don't believe we have been able to link any of those
MR. HILL: There are debris events that
are within seconds of some of the com dropouts. That
doesn't necessarily tell you they're related, but there
are debris shedding events in this same time frame.
MR. HUBBARD: Okay. So the set of charts
here, Rev 15, just looking quickly through those since you're not going to cover these, I see up through
Com Event 14. How many of those are there?
MR. WHITE: Well, let's see here. Let me get my other version of the time line. We had at 13:52:09 through 13:52 well, let's back up. 13:50:00 through 13:50:43, we had five periods of com dropout from one to six seconds each. 13:52:09 through 13:52:55, there were four periods again from one to six seconds each. That would cover Events 6 through 9. Then again, 13:53:32 through 13:54:22, which would be right in this period here, there were two more periods. One was two seconds. One was 8 seconds. Those would be Com Events 10 and 11. There are some more events, 12 and 13, that are down in the 55, 56 time frame; and Com Event 14 was down at 13:56:55.
MR. HUBBARD: Okay. So can we expect to
see some point in the near future a composite plot or a
plot like this that shows the antenna wire, the
antenna, where the avionics is and so forth and kind of
be able to put it together?
MR. WHITE: Well, the scale we could probably do on a separate page just because of the scale. Yes, we could go ahead and do some kind of a graphical representation of that. Again, we don't see anything anomalous in the behavior of the com system other than com wasn't getting through to the ground. So there may not have been anything physical going on within the orbiter itself at that location on the vehicle itself.
MR. HUBBARD: It could have been some
interference between the orbiter and receiving
MR. WHITE: Yes, it could have been,
again, as I said, some kind of disturbance in the hot
gas around the vehicle at that time possibly.
MR. HUBBARD: Okay. Thank you. We'll,
I'm sure, be pursuing this further.
GEN. DEAL: I'd like to bring up a
question about Dr. Widnall's statement about perhaps
earlier debris that was not witnessed. Can you kind of
put it in context, where we saw heat onset and also the
beginning of peak heating?
MR. WHITE: Let's see here. Let me look at my really detailed time line and the event times for that. The beginning of entry interface, which is about 400,000 feet, is 13:44:09. The start of peak heating is at
DR. WIDNALL: 50.
MR. WHITE: 50. Okay. Thank you.
GEN. DEAL: The reason I ask that is to underscore her statement. There could have been things that weren't witnessed because you aren't starting to experience heat before
MR. WHITE: Right. There could have
DR. WIDNALL: About the com. I'm very
interested in the com. Is that anomalous for the whole
range of shuttle missions, this loss of com?
MR. WHITE: Yes. For this particular period, we have called these losses of com "anomalous events." We've compared them to other flights of Columbia on similar trajectories and we believe we should again, because of the look angles and where we were, we believe we should have had good com.
DR. WIDNALL: So it wasn't just a simple
matter of shielding by the vehicle of some antenna?
You've already dismissed that possibility?
MR. WHITE: Yes. We've looked at that,
and we truly believe there is something anomalous going
on here. Now, what it was and how to describe the
effect, we're not sure how to do that yet. We're still
working on it; but, yes, we do believe that the com
dropouts in this period were anomalous.
ADM. TURCOTTE: This is one of the first
aerodynamic events that you've indicated here and I'm guessing you're interpolating here roughly we're in the
220s, probably lower Mach 20s. What kind of
aerodynamic pressure is the air foil experiencing at
MR. WHITE: Again, I don't have those numbers in front of me. There are versions of this that have
DR. WIDNALL: Fifty.
MR. WHITE: Thank you. I was going to go
look that up in my notes.
ADM. TURCOTTE: If you were to put that in layman's terms, we're looking at, say, around 120 knots or something like that
ADM. GEHMAN: Less. The QBAR was 29 PSF.
MR. WHITE: Okay. That's pounds per
ADM. TURCOTTE: Probably rough 80 knots,
something like that.
ADM. GEHMAN: And the Mach is 22.7. So
you used PSF?
MR. WHITE: Yeah. QBAR is in pounds per
ADM. GEHMAN: Yeah, I know that. When
you're doing conversion to knots, you use PSF? So
something like 75 or 80 knots air speed, something like that.
MR. WHITE: Okay.
ADM. GEHMAN: And we are in a stagnation
temperature now of 2850.
MR. WHITE: Yes.
ADM. GEHMAN: So we're peak heating.
MR. WHITE: Yes. Very high heating at
ADM. GEHMAN: I think the point is that
there is not 10,000 knots of air flowing past this
MR. WHITE: Right. We were at a very low
dynamic pressure at this region. Right. Lots of heat
but very low dynamic pressure.
ADM. GEHMAN: But things are falling off.
MR. WHITE: That is correct.
Next one, then. This is another
temperature. This is on a left main gear strut
Next slide, please. This is a side wall
temperature. This is the left aft fuselage side wall
temperature. Now, this particular temperature is about
where it's indicated there on the left aft side wall,
almost at the end of the wing. This is another
indication that something going on externally in the flow above the wing is causing this heating up on the
side wall that far aft.
ADM. GEHMAN: Now, would you attribute
this more to external heating rather than internal
MR. WHITE: Yes, I would. We have done
some calculations, though, that say you could
theoretically get enough flow or heating internally to
cause this to rise. We have shown, though, that
externally, if you were just missing the blankets, you
wouldn't have enough heat to cause the temperature to
rise. But theoretically it would be possible. We've
done some numbers that said you could have had heating
from internal. That's also possible.
ADM. GEHMAN: Is this sensor right underneath the blanket
MR. WHITE: Yes. This is on the skin
ADM. GEHMAN: On the skin right
MR. WHITE: Underneath the blanket. Yes,
ADM. GEHMAN: Thank you.
MR. WHITE: Next slide. Now, we're back
to the left main gear strut actuator temperature. This
particular temperature is on a strut when the gear goes down that supports and braces the gear, and again this
one saw a rise. Again, you also notice, as you
mentioned earlier, there are other sensors in the
neighborhood that are still showing nominal at this
Next slide. Flash 1. The triangles
below line there, this is another debris event. We saw
a brightening of the orbiter image on the video, which
occurred where the orbiter was; and then as the orbiter
moved away, the splash tended to persist in the trail
that was showing behind the orbiter.
Debris No. 6. Next slide, please.
Debris No. 6 is the sixth piece of debris that we've
been able to observe in the video. This one I used a
larger triangle, to indicate that this was a relatively
significant piece of debris compared to the other ones.
Debris No. 6 and Debris No. 14, from the video that we
have, appear to be the largest and brightest debris.
DR. WIDNALL: Could you back up one?
MR. WHITE: Yes.
DR. WIDNALL: Do you have an explanation
for Flash No. 1?
MR. HILL: We think Flash No. 1 is
attributed to Debris 6 actually separating from the
vehicle. We just don't see Debris 6 as a separate object until a few seconds later, but we really do
think this is the initial event as that object came off
the vehicle, crossed through the plasma wake and shock
ADM. GEHMAN: But we're going to get a
chance to talk about that.
DR. WIDNALL: Yes. Tomorrow.
MR. WHITE: Debris No. 6 was right after
that. And next slide, please.
Now we start to see some temperatures on
the wheels themselves. These temperature measurements
are down on the body of the wheel. This is the first
one of these. So we're starting to see a little bit of
a rise. Again, we noted there was two bits. There was
a very small increase in the temperature of the wheel.
Next slide. Debris No. 7. Again, we are
Next slide. All right. Another
temperature measurement on the side wall of the wheel
well. This is System 3 left-hand forward brake
switching valve return line temperature.
Next slide. Debris No. 8. Approaching
the Utah border.
Next slide. Debris No. 9.
Next slide. Debris No. 10. These all come off relatively close to each other.
Next slide. Debris No. 11.
ADM. GEHMAN: And you're going Mach 22 at
this time with a QBAR of about 35 PSF.
MR. WHITE: Thank you. Next slide,
please. This is another temperature on the side wall.
This particular one is on the sill, which is actually
the top of the wall. It would be underneath the
payload bay as the payload bay door comes up and over.
This particular temperature would be sitting about
right here, just under the door, on the top of the side
wall. So we're getting some more heating up there.
Again, this leads us to believe that we had something
going on with the external flow that was causing
higher-than-normal heating above the wing in this
ADM. GEHMAN: At this point, the orbiter
is flying with its right wing down, left wing up.
MR. WHITE: Yes.
ADM. GEHMAN: Yes, it is. Hasn't done
MR. WHITE: Hasn't done the roll
ADM. GEHMAN: So these are left fuselage
MR. WHITE: Yes, they are on the left
ADM. GEHMAN: Left side of the body. Is
there a hotter side or a cooler side? I know the
bottom heating is all true to form, but is there any
reason aerodynamically or thermally to account for the
left side being warmer? In other words, should I read
anything into it? Would you expect the left side to be
cooler, this particular side, since it's up and away?
MR. WHITE: Well, I think you really need
to ask the thermal guys tomorrow.
ADM. GEHMAN: You're right.
MR. WHITE: Generally, from what they've
told us, it should be about the same and we believe
these rises here were from some off-nominal event
causing more heating on the left-hand side. As
compared on a normal entry, one roll reversal compared
to another roll reversal, I really can't comment on the
relative slight differences you might see in
ADM. GEHMAN: We'll pursue that tomorrow.
MR. WHITE: Next slide, please. This is
Debris No. 12; and we're just crossing the Arizona
Next slide. Debris No. 13.
Debris No. 14. Next slide. This again
is a very large debris relative to the other debris
events. So we show the triangle a little bit larger at
ADM. GEHMAN: So it's Debris No. 6 and 14
we want to pay attention to.
MR. WHITE: Right. Paul's going to talk
to you about that, about our efforts to track Debris
No. 6 and 14 and see if we can figure out a footprint
and perhaps recover those debris.
All right. Next slide, please. Now, we lost these five wing temperature measurements early on; and now we are starting to lose some more. This particular one is the left lower wing skin temperature. This measurement is on the lower wing skin itself, right on the bottom side of the vehicle. This one is starting to this decline. And as you'll notice, these took quite a bit more time to go off line than the previous five that did go off line.
ADM. GEHMAN: Now, these five that went off earlier, I can't tell from the color code whether or not they are in the same
MR. WHITE: Yes, they are in the same
wire bundle as the five that went off.
ADM. GEHMAN: They're in the same wire bundle, but they're not on the same circuit. It kind
of shows up here all tangled.
MR. WHITE: Well, yes. Each one of these
sensors would have its own wire within the wire bundle,
ADM. GEHMAN: So we should not read
anything into the fact that there's a difference
between these five going and these two here. I mean
they're just different wires.
MR. WHITE: Different wires within the
same bundle, yes, sir. And, you know, I was talking
about twisted shielded pairs earlier. These wires for
each one of these sensors is actually, if I remember
right, a triplet of wires which is then encased in
Kapton and then that particular wire that's formed from
the triplet is one wire of many in the larger bundle.
Next slide, please. This is Debris
Next slide, please. Now, we have another
wheel well temperature. This is a left main gear
uplock actuator temperature. This is the actuator that
holds the gear in the lock for the gear, locked in the
up position; and we're seeing an off-nominal
temperature rise there. Also notice that there's
another temperature on the side wall. We've colored it orange, which means its temperature rise now has
exceeded 15 degrees from what we would consider
nominal. So the temperature on the side wall continues
Next slide, please. Now, there's another
skin temperature. This one happens to be the upper
wing skin temperature. It's approximately above the
one in the lower but on the upper surface of the wing,
and this one is starting to go off line. You also
notice that the lower one hasn't quite failed all the
way completely yet by this point in time.
Next slide, please.
ADM. GEHMAN: Excuse me. Now, what
should we read into the fact now that on your cartoon
here every sensor on this line here has now failed?
Are there other wires in that bundle?
MR. WHITE: There are many other wires in
ADM. GEHMAN: In the same bundle?
MR. WHITE: In the same bundle. Yes, sir. These are the only on that particular bundle, that pink that we indicated in pink there, those are the only ones that we have data for. The other wires in the bundle are either not used anymore because they were development flight instrumentation which we are no longer using or they're a series of instruments that are recorded on what we call our orbiter experiment recorder, which records measurements and then we dump the tape when we get it to the ground and look at the values for that; but they're not available to us in realtime. One of the things we've been hoping to find in the debris is that recorder to see whether or not any of the tape survived that may give us some of the data to tell us how other measurements in this area were faring at this time and so we can learn more about the event.
ADM. GEHMAN: Would you estimate how many
of those sensors there are in there?
MR. WHITE: I went and got the number
once for somebody. I do not remember the exact number
off the top of my head.
ADM. GEHMAN: Dozens more?
MR. WHITE: It's on the order of a dozen
ADM. GEHMAN: Thank you very much.
MR. WHITE: Next slide, please. Okay.
This is Debris No. 16. This is a debris event that was
picked up in the Kirtland video, which I'm sure
everybody's heard about a video shot by some of the
folks at Kirtland Air Force Base; and we were able to see a debris event in that particular video.
Next slide. All right. This is the main
landing gear. Back on the tires again and on the
wheel. The main landing gear left-hand outboard tire
pressure No. 2. It's starting to show a little bit of
an increase, only one bit.
ADM. GEHMAN: Could we back up just a
second here? I think for the time line we need to
determine when the roll reversal was. I think it
happens right about 56:55. About 30 seconds ago we did
the roll reversal.
MR. WHITE: That's correct.
MR. HILL: We start at 56:30 and finish at
MR. WHITE: Right. 56:30.
ADM. GEHMAN: So the roll reversal is now
MR. WHITE: Yes. That's the complete of
the first roll reversal.
ADM. GEHMAN: Now the left wing is down.
MR. WHITE: Right.
ADM. GEHMAN: People keep telling me that
that doesn't make any difference in coordinated flight,
but I think it helps to understand.
MR. WHITE: All right. Next slide, please. All right. This is the lower wing skin temperature finally completes its descent down to off-scale low. It did take a little longer than the first five. Again, to us that just indicates that the rate of burning or the rate of shorting of that particular wire was different than the first five again, possibly indicative that whatever was causing the burning was changing direction or heat rates or something like that.
Next slide, please. And then the upper
wing skin temperature follows that shortly.
Next slide, please. Now, we start to see finally the last of the hydraulic measurements in the wheel well start to go up. You can notice some of the other measurements have now turned orange again, indicating that they are continuing to rise and have gone more than 15 degrees above what we could consider nominal for this particular point in the flight.
Next slide, please. This is what we're
call Flare 1. This is another event that we observed
out of the video taken at Kirtland Air Force Base. We
see an asymmetrical brightening of the shape. In the
video you can see one side of the orbiter image get
brighter than the other side.
DR. WIDNALL: Which side?
MR. WHITE: It appears to us to be the
Next slide, please. Then Flare 2. Again
you see another little bit of a flare, again apparently
from the left side.
Next slide, please. This is another
aerodynamic event that we put in here graphically.
This is the start of the sharp aileron trim increase.
Remember we've been doing a slow aileron trim increase,
trying to keep vehicle flying the way we want it to
fly, trying to make it respond. At this point there is
some event that happens that causes the aerodynamic
forces to require a much greater trim on the aileron
and so the trim begins increasing very rapidly here.
Again, you'll have some charts tomorrow, when the
aerodynamics guys talk, to show you how rapidly that
aerodynamic set of forces was increasing.
Next slide, please. We're also seeing an
increase now in the derived rolling and yawing moments,
those moments I told you that we were able to back out
way up early that showed something off nominal. Again,
the slopes of these moments are starting to change
substantially at this point.
Next slide, please. This is on the tire itself. This is main landing gear left-hand tire
pressure No. 1. Again, it's starting to show this
damage trend as it's going down. Again, as you
mentioned earlier, one of things that's a mystery to us
is why the measurements on the tire seem to hang in
there for so long whereas other measurements farther
back in the wheel well seem to be significantly off
nominal by this point in time. Again, it may have
something to do with how well those measurements are
protected by the tires themselves and the heat sink and
the mass of the wheels themselves.
Next slide, please. This is on the other
tire. This is main landing gear left-hand inboard tire
pressure No. 1. It's showing some damage trends.
Something else I might say at this point
too is you watch all these temperature measurements and
pressure measurements for the wheels go off line. We
saw these in a staggered kind of a fashion, which
indicates to us that the tires themselves did not
rupture or blow up, at least not at this point in time.
That may have happened after our loss of signal, but at
this point in time these measurements are going off in
a staggered fashion. That says that the tires were
still intact at this time.
Next slide, please. Back to the left outboard wire damage trend showing on one of the
sensors there. Wheel temp.
Next slide, please. Back to the inboard
one. Damage trend there.
Next slide, please. We finally get the
landing gear left-hand outboard tire pressure No. 1 to
go completely off line.
Next slide, please. Now the left
outboard wheel temp goes off line.
Next slide. Now, the landing gear
left-hand outboard tire pressure No. 2 starts to go off
ADM. GEHMAN: Doug, once again, the
people in the audience can't see the companion
viewgraph that goes with this that shows the actual
MR. WHITE: Right.
ADM. GEHMAN: But I'll describe. I'll
hold it up, for example. Which one are we on? The
left-hand outboard tire pressure. The temperature is
normal. There's no rise in temperature, and then the
thing drops off.
MR. WHITE: The thing just goes off.
Right. The temperature is constant, and then it just
drops off. Right.
ADM. GEHMAN: And that's true of all of
DR. AILOR: Right. That indicates to us
that the tire was intact, that we weren't seeing some
sort of a pressure increase in the tire that it was
about to rupture and that there was damage to the wire
for that measurement that caused it to drop off line.
ADM. GEHMAN: And whatever heat was causing all these temperature sensors to rise, that heat was not present up here and
MR. WHITE: Well, it was present to some different degree. It was having different effects. Again, since it's difficult to model the propagation of how the heat was getting in there and we're working on that and it's a difficult thing but it was obviously having different effects in there than it was farther back in the wheel well.
ADM. GEHMAN: Let me rephrase the
question. These temperature sensors here are all
MR. WHITE: Yes.
ADM. GEHMAN: These temperature sensors
here, there's no temperature rise in any of those
sensors. They just drop off.
MR. WHITE: They just drop off, right, which says the wires were getting damaged.
ADM. GEHMAN: I understand neither you
nor I can figure out why that happened, but these
temperatures are rising and some of them have now gone
orange, indicating that the rate of the rise is now
alarming, whereas these don't show any rise whatsoever.
MR. WHITE: That's correct.
DR. WIDNALL: Where is the cable located
for those wires, the blue ones?
MR. WHITE: The ones on the wheels
themselves, the lines run on the back of the gear, on
the back of the strut and they run up the strut.
DR. WIDNALL: Can you show it?
MR. WHITE: They run along the strut
here. They come up to the back of the wheel well.
They come to actually a kind of a junction box here and
they run across the ceiling to the front of the wheel
well and then they run out through a connector into the
mid body about there.
DR. WIDNALL: So they're inside the wheel
MR. WHITE: Yes, they are inside the
wheel well structure.
DR. WIDNALL: And at least over part of
the area, they're mounted on the front bulkhead.
MR. WHITE: Yes.
ADM. GEHMAN: But I think Sheila's point
is very pertinent because even though these sensors did
not show any temperature rises, the wire that feeds
these temperature goes all the way back into this
MR. WHITE: Yes.
ADM. GEHMAN: And then comes back out of
that region again because of the way the landing gear
was folded back over on itself.
MR. WHITE: Yes. And if you want to
surmise that maybe we're just today burning through
wires here, you would want to think that it was down
closer to the sensors themselves on the strut because
there are other temperature measurements again that are
coming in this bundle across the top of the wheel well
and then out through that connector that are still
reading and acting just fine. So some kind of burning
was going on there. It was most likely down on the
strut next to the wheels themselves rather than up on
the ceiling of the wheel well.
ADM. GEHMAN: Thank you.
MR. WHITE: Next slide. This is main
landing gear left-hand inboard tire pressure No. 1 has
gone off line.
Next slide. This is main landing gear
left-hand inboard tire pressure. Again it's showing a
very slight increase in tire pressure. A
3 1/2 pressure rise in two seconds. That didn't last
very long because that sensor went off line shortly
Next slide. You see right there in the
next slide it started to go off line and that
measurement started to trend down.
Next slide, please. Another main wheel
well temperature that went off line.
Next slide. Then the next-to-the-last
one went off line.
Next slide. Then finally the last one.
So all of our sensors, both temperature and pressure on
the wheels, have gone off; but again since it was a
staggered fashion, we don't believe that one or the
other of the wheels let loose, which would have lost
all of them simultaneously.
Next slide, please. This particular
measurement, the change here, this is called the left
main gear downlocked. This is a sensor which tells us
that the gear would be down and locked. This
particular sensor changed to a 1 state, which is an
off-nominal reading for this state. We did do some wire testing to see how this particular sensor would
fail if its wire was burned through. It would fail to
a 1 state. So this could be either real, that said
that maybe the gear did come down at this point and it
was a 1 because it was supposed to, or it could be just
that the wire had burned through. The other sensors in
the wheel well, you can see the other three red squares
there, they were still all reading their nominal
values, which tells us that the door was up and locked.
We have three other sensors. We have the door up, a
gear up, and a no weight on wheels; and all of those
were reading their nominal values. However, from
testing that we did from wire burning to see how those
would fail, those could fail in their nominal state if
their wires were burned through. So it is possible
that those wires were already failed but we didn't know
it. It's also possible they were reading exactly the
way they should have because the door was still up and
locked at this time.
ADM. TURCOTTE: Is this the same location
of the previous tire pressure wire bundle that you
described before and that is located along the center
line of the gear?
MR. WHITE: Right. This particular one
is along the strut. Now, the one that you see very forward there, that particular wire bundle runs all by
itself across the front of the wheel well and up to
that connector. It's not in the same bundle until very
late with this particular one that's failed here. So
that's a separate bundle, but the three on the gear
there are all in the same bundle.
ADM. TURCOTTE: So that's the one that's
located on the assembly by the dust cover where it goes
through into the wing?
MR. WHITE: This particular one is on the
strut itself, but the wires then run as you described
back into the mid body there across the top.
Next slide, please. Right. This is
sensors starting to go off line, one of the ones that
had been reading temperatures, system 2 left-hand aft
brake switching valve return temperature, starting to
go off line.
Next slide, please. Now, this other wire
that goes to the ASSA that was the gray wire that
actually looks kind of purplish here, this is starting
to show that it was burning through somewhere and
shorting. We have evidence that our air surface servo
amplifier was shorting out and was not providing power
the way it should have to Channel No. 4 for the elevon
actuators, but the inboard and the outboard we begin to see off-nominal events and in the detailed time line
there are quite a few off-nominal events. This is
right before LOS or one second before we lost signal
here, but this does indicate to us a sequence of events
that I just labeled with this one event here, that we
were burning through this power wire, causing shorting
to go on there that air surface servo amplifier. What
we also see from the data here at this point is that
the other three channels were taking over and the
redundancy management that's built into the system was
working the way it was supposed to be working. The
other three channels took over and were in control even
though this system was failing.
Next slide, please. This is just prior
to loss of signal. You can see all the things off
MR. HUBBARD: Doug, before you get to
that loss of signal. If you were to come up with some
kind of a metric of event as a function of time and you
plot that from the beginning to this point, do you
imagine that that's linear or is there some knee in the
curve? Is there some point in this nine minutes or so
here where things pick up?
MR. WHITE: Yes. I would call the knee
in the curve the place where we showed the start of the sharp aileron trim increase, which is back up with one
of those triangles there on the top. The vehicle was
in control and was responding to commands up to that
point, and after that point something changed
apparently and it still continued to be in control and
still continued to respond to commands but the rates
and the amount of muscle it needed to continue flying
the vehicle the way it should be flown was continuing
to increase. Something definitely happened at that
point. Again, we don't know what; but something
definitely happened at that point to cause the flight
control system to need more muscle and start to have to
fight harder to control the vehicle.
MR. HUBBARD: And that was at about?
DR. WIDNALL: I think that's about 57.
MR. WHITE: Yeah. That would be about
DR. WIDNALL: I guess the comment I would make because I have looked at that particular instance of time that really coincides with a rather sharp increase in the rate of rise of dynamic pressure.
MR. WHITE: Yes, it does.
ADM. GEHMAN: Okay. Thank you.
MR. WHITE: Right. That's as far as I
planned to brief in these charts. As you know, there is some data that we recovered from the satellites
post-LOS. If you want to talk about that, I can answer
questions about that; but I don't have any more charts.
ADM. GEHMAN: Okay. Let's let Paul have
the floor for a few minutes and then questions.
MR. HILL: Okay. Now, as I mentioned
before, what my team has been doing is evaluating
various public imagery, various external sensors and
trying to make some sense out of the data and see if we
can get smarter about what's coming off the vehicle
earlier on as far west as we can, as well as get some
engineering data to tell us specifically what those
objects are and where they're going.
I don't really have prepared presentation
charts. I'm going to wander through some discussion on
this map. I have a few other pictures I'm going to
show you, and I did bring a composite video that shows
examples of continuous video from the California coast
through about mid New Mexico. Since this video was put
together, we have added one that takes us about
50 miles offshore California and we have some video
from Kirtland Air Force Base that takes us through just
about the New Mexico, Texas border. Those aren't going
to be on this tape that we're going to see here in a
Let me start with the process, then we'll
play the tape. To give you an idea, when we first
starting getting these videos, our first job really was
to put them in chronological order. That's still
photographs, video, et cetera. We very quickly focused
on just the video and saved a lot of the still
photography analysis for later.
Our first goal is to establish some
absolute reference for time in each one of the videos.
Once we have that, we can put them in chronological
order. As we were going through that process, probably
three or four days after the accident, we first saw in
these videos individual debris shedding events; and
that was our first indication that something, in fact,
was coming off the vehicle early on, that we didn't
just start having structural damage, say, over west or
east Texas. You'll see, as we play the tape, some of
the things that we use for cues in establishing time
and establishing relative geometry. There are a couple
of celestial references in a couple of the tapes.
You'll see a star. You'll see Venus crossing, which
will be very clear. At least half the photographers
snapped their GPS location so we know exactly where
they were standing. In the case of the Venus crossing,
because we know where that photographer was standing and we see the orbiter actually flying in front of
Venus, we can calculate when in time that had to have
happened. So now we can put that tape exactly where it
was in time and we know exactly where the orbiter was
in space and then we can sync the videos that preceded
that one and the ones that followed to that tape. We
had a few other cues like that in other tapes, and I'll
try to describe those as we go when we play the tape.
As we started seeing these debris shedding events and you'll see these in the tape, although some of them you do have to look closely because they only last in the order of a second or second and a half in cases, we then set about calculating the exact times that the debris was coming off the vehicle. As we established those exact times, we went to work, trying to do relative motion and ballistic analysis. I'll come back and talk about that here in a few minutes.
Interestingly, not only was NASA not
aware that debris was coming off that early before we
looked at this video but most of these photographers
did not see any debris shedding in their own
photography until they heard about the accident on the
radio or on TV and went back and played back their
video. Then they could see them. Like I said, in most cases debris flash or the speck that you see in the
video lasted for a second and a half or so, in most
cases less than a second.
The types of things to look for in the
video. In some cases there's flashes, like Doug talked
about. In other cases you can see a bright dot which
is orbiter and plasma wake behind the orbiter, and then
you'll see another dot come from a dot. And you'll see
when we play the video we are not seeing images of an
orbiter against a dark sky where we can clearly make
out the planform and shape of the spacecraft where we
can clearly resolve down and see where some object is
coming off the vehicle. We see a dot, we see another
dot appear from that dot, and one of the dot goes away.
And we will talk about that some more as the video
The other thing to think about as we
watch the video is we are making some speculations
about what we are seeing. We think that the brighter
objects are more massive, are more significant,
potentially higher ballistics numbers. Certainly the
things that the individual light for the individual
pieces of debris persists longer, we expect that those
objects are more massive, higher ballistic number
because we think that the reason they persist longer is they are moving faster. So they stay lit. They have
their own plasma wake, longer than, say, some lighter
thing, say, an individual tile comes off versus maybe
some other heavier object. But I'll also say we cannot
just look at these videos and just determine what is it
that's coming off the vehicle. Are we losing a tile
here? Are we losing some section of the thermal
blanket that's on part of the external surface of the
vehicle? We can't tell that, and to this day with the
good data that we have on the ballistic motion and the
ballistic analysis and the footprints, we still cannot
say exactly what it is we see coming off. We are
making some judgments on which of them are more
significant or more massive than the others. And we
talked about Debris 6 and Debris 14. When we play the
video, you'll see why we're focusing on those.
So why don't we go ahead and play the
video and then we'll come back and I'll talk some more
about what we've done on trajectory analysis.
ADM. GEHMAN: You can feel free to stand
up and narrate or point. However, you feel comfortable
showing us what happened.
MR. HILL: This is just after the
California coast. As I mentioned, you see a dot.
That's the orbiter. And the view looks more or less like this as we change the vantage point. We'll start
picking up the con trail.
Now, if you missed that, that was
Debris 1 and that was Debris 2. Those little dots that
came off, that was debris. As I mentioned, you can't
make out the planform, you really can't see the
orbiter, and you have no idea what's coming off. Also,
as I mentioned, on some of these or most of these, the
debris itself doesn't last very long at all.
ADM. GEHMAN: Now, this is a significant
MR. HILL: Yes. Now this bright dot you
see here, this is Venus. When our flight dynamics
folks saw this, they were very excited because this
allowed us to put this video within plus or minus a
Now, you can see the flash persist in the
wake and then you see Debris 6 come off. Even though
they're separated by a few seconds there, our
speculation is the flash was some burning event
associated with Debris 6 and then that object coming
ADM. GEHMAN: If I understand it, Debris
No. 6 is the one you tracked to the vicinity of
Caliente, Nevada, and we are valiantly trying to find.
MR. HILL: We do think that is Debris 6,
and I'm going to show the footprints for that and
explain that a little bit more.
There you saw Debris 7 come off. Now,
again, also just for a reference, all of these are
taken with camcorders. These are commercial
camcorders. This is somebody in the public, standing
outside with a camcorder, generally zoomed way in,
trying to track the orbiter flying overhead at
12,000 miles per hour.
ADM. GEHMAN: You recommend people pay
attention to Debris No. 14. That's the other one.
MR. HILL: Now, as we come up on
Debris 14, the thing to think how is bright that flash
was before Debris 6. Compare that to what Debris 14
looks like. Also, for comparison, Debris 6 was lit
from between 6 and 12 seconds.
Now, there you saw how bright that was and also you saw that you have this cloud where around the orbiter, the video itself became saturated. That is the most bright the brightest object that we saw in any of the video. And I'm going to come back and talk about its relative motion and Debris 6's relative motion here in a few minutes.
You can see here we're getting further east. We're getting out over New Mexico. The sky is
lightning up, which makes it more and more difficult in
the videos that we have out there to track the orbiter
and specifically to pick out individual debris shedding
ADM. GEHMAN: But in your experience and
the experience of the experts, that hot gas envelope
right there looks just like any other entry that you
MR. HILL: That's right. Except for any
of the flaring or flashes or anything else, the bright
spot you see there looks like just all the other videos
that we have. As a matter of fact, one of the
photographers that sent us this video sent us six
previous entry videos that he took, most of which with
the same cameras, and looked just like this except
absolutely no flares, no dots coming off.
ADM. GEHMAN: The number down in the
right-hand corner is what's on the camcorder, but
that's not calibrated time. Your times are in the
bottom left-hand corner.
MR. HILL: That's right. Now, we have
done a fair amount of work. Again, about half of these
photographers were amateur astronomers and they had
synced their clocks themselves to atomic clocks. Some of them went back and taped the atomic clock so that we
could do our own calibration, and some of them did some
of that afterwards.
Now, the things you're seeing here are
just prior to or including the main breakup.
ADM. GEHMAN: But this is post loss of
MR. HILL: Correct. We left this in here
for completeness. We're going to talk a little bit
about post-breakup and pre-breakup. I thought we would
go ahead and run the tape through this to give us a
place to start from. These videos were all taken from
Texas, of course.
This was taken from an Apache helicopter, looking through its forward-looking IR targeting sensor. Now, the thing to think about here we'll come back and talk about this in a while is the significant number of secondary and tertiary breakups that you see in these videos. That will be important when we talk trajectory analysis.
DR. WIDNALL: Can I ask a question? Are
there any gaps in time missing, where you don't have
video? Is there a continuous time line between the
first sighting and these later pictures? Are you
MR. HILL: There is a small gap in the
East Texas or the East New Mexico, West Texas area. It
is not as big as represented on this tape.
DR. WIDNALL: How long is it? A minute?
MR. HILL: I would say it's on the order
of a minute or two minutes. Everything else west of
Albuquerque, we have near-continuous video for. Now,
it shifts around from vantage point to vantage point
and there are dropouts in individual video. As a
matter of fact, if you segue into the map here for a
few minutes, the blue dots that you can see on the map,
those represent where the individual photographers were
standing. If you take this one, for example, here,
this is in Flagstaff. This blue line extending out
this way, there's another that extends out this way on
the map, that wedge represents the full part of the
trajectory that that photographer filmed in his camera.
It doesn't necessarily mean that that photographer has
continuous coverage of the orbiter for that full swath
because many of them dropped track, lost the orbiter.
They'd look away from the view finder. The camera came
down, and they had to go find it again. But for the
most part, with all of the overlapping video we have
from California all the way through New Mexico, we've
been able to piece together essentially continuous views of the orbiter.
Now, the other important thing is on some
of these objects when we see them coming off the
orbiter in one view, we may not see that same object
coming off for another second or so in another view.
In some cases we don't see it from a different vantage
point of the same incident. Some of that is because
one observer, say, may be looking from the north side
of the trajectory and the folks down here are videoing
from the south and one of them may have the orbiter
itself maybe obscuring the view of, say, the flash or
the individual debris coming off. Since that debris
only persisted for maybe a second in most videos, it
wouldn't take much obscuration at all for one video not
to see it. The short answer is we have near-continuous
video until right about here, and that's east of
Albuquerque, New Mexico, and there's there gap and we
pick up with that Texas video of the main breakup.
South of Dallas.
DR. WIDNALL: You have a gap between Albuquerque and
MR. HILL: Albuquerque and about the
Dallas area, which I guess you would expect because of
the relative population. Most of the video we have,
even out here in Arizona and New Mexico, which is relatively thinly populated, most of that we have from
Albuquerque and Flagstaff and from Las Vegas. And the
one from Flagstaff in particular, they tracked for a
significant period of time, from horizon to horizon.
So that's our explanation for the gap there.
Now, going back to the video a little
bit, you see the type of relative motion or the type of
relative distances you see in the objects that come off
the orbiter. We're able to zoom in on those objects.
We're able to zoom in on the orbiter. The imagery
folks here at JSC are able to take all that jitter out
so that there's no motion except for the relative
motion between the object and the orbiter. We can then
measure how that objects moves away from the orbiter;
and since we know exactly where the orbiter is in space
relative to the photographer and we know exactly what
the timing is, we can calculate ballistic number of
that object, based on how it moves relative to the
orbiter, because we know the orbiter's ballistic
number, of course. We then take that ballistic number
for the object and we propagate that down and build a
vector so that we can propagate the object forward all
the way down to the ground. Then we generate a series
of footprints at 80,000 and 35,000 feet and ground
If we can put up page 2 of my charts.
We've done a couple of things. What you see here is a
very generic footprint. We started with this. Before
you could calculate relative motion and ballistics off
the video, we made some simple assumptions like we were
shedding a tile every two seconds from California all
the way to Texas. Based on the known ballistic
properties of the tile, that gives us a debris swath
that looks like this, which is still enormous; and it's
about 30 miles above, 30 miles below the ground track
for that full distance. That's what we knew very
quickly, within a day or so of the accident.
If we move on to the next page, a similar
footprint based on the main body breakup, also based on
various simplified assumptions on ballistic numbers,
both the light and heavy objects. This footprint is
for the debris field in East Texas; and it, in fact, is
centered right over the debris in East Texas. On the
far right side down in the lower corner, that's near
Fort Polk, Louisiana, which, in fact, is where main
engine components have been found. Now, again, these
are both very generic and they're based on relatively
wide simplified assumptions.
If we go to the next page, this is based
on Debris 6. This is that object that we see coming off somewhere near the Nevada, California border. In
fact, this footprint, this blue line here, that's the
New Mexico, Nevada, Utah state line. This small box
you see here, if we exactly nailed the debris shedding
time, if we exactly nailed our ballistic analysis,
that's where you would expect that object to be laying,
if it also didn't generate any lift.
We've done a bunch of other detailed
analysis. If you go to the next page, just for
comparison sake, depending on the errors that we had,
it is just as likely that the object, instead of
landing in that no-lifting box here in the middle,
could have drifted off track to the north, off track to
the south, just by generating lift. If we had some
error in the time that we calculate in that object
coming off or in our ballistic analysis, then it could
also fall short up here in this part of the footprint
or along down here.
Could we back up a page, please. Now,
this is Debris 6. This is the first one we had
analysis on. We were able to get analysis completed on
this one earlier because we had that Venus crossing and
we really knew the relative motion of this one much
better than we knew everything else.
After we build the footprint, then the process would be going through the FAA radar data which
we have saved off and recorded; and we're working with
the NTSB for them to search that radar data to find
patterns that would not normally be noticed by air
traffic controllers. In that process we have found a
thread up here in this area which is just inside Nevada
before crossing into Utah and another one down here
just south and then another one over here in Utah near
Mount Zion National Park. These are the first three
radar threads that we found; and, in fact, these are
the three areas that we have been trying to search here
for about a month now.
The one in Utah is very mountainous
terrain and is most likely only going to be searched by
air. It has been searched already by air. We're
talking about doing some more air search. This one up
here in Nevada which is near a place called Caliente,
Nevada, we have had folks on the ground there,
searching. It's also snowed out there about five
times, to the tune of 4 to 5 inches of snow each, since
February 1st, which certainly our problem of searching
and finding things.
We also say again we don't know what this
object was. We know that, based on its relative
motion, it has a ballistic number on the order of 3.75 to 4.75, which compared to the orbiter ballistic
number, which is on the order of 100 to 110, makes it
something that's relatively small and light.
Like you said, Admiral, we expect this
object to be Debris 6. I mean, the objects that we're
finding the radar threads for, we expect it to be
Debris 6 because it lies right in this Debris 6
footprint and so close to the no-lift in the box. We
don't know for a fact that it is because, as Dr. Ailor
said, as these things come off the vehicle, they could
continue to fail, break into smaller pieces, which then
could completely changes their ballistic properties.
Our general process is the same, though. We calculate
relative motion, calculate ballistics, propagate out
this footprint, and then we search the footprint for
If we go to the next page again, this is
Debris 14. That is second object that was so bright
compared to Debris 6. Let me correct something that I
told you on Thursday. Debris 6, you can see, persists,
depending on the video you look at, for between 6 and
12 seconds. Debris 14, we see, persists for 4 1/2 to
7 1/2 seconds, depending on the video you look at; but
Debris 14 is also much, much brighter than any other
object, including Debris 6.
How do you interpret that? We're not
sure. We do think that relative brightness is an
indicator of something that's larger and more massive.
We think that the amount of time that individual flares
or the light around that debris persists is also
indicative of the larger ballistic number, which tells
you you're dealing with something that's probably
larger and heavier. That's as much as we know. We
know how these things behave ballistically way up high
when there's not a lot of air.
In addition to just searching these
footprints for its FAA radar, we've also moved all the
way out west to the west coast of California and we are
searching all air traffic control radar anywhere it
intersects our ground track or that wide generic swath
around the ground track to again see if we see any
patterns of Columbia debris falling through that radar
that would have been ignored by air traffic control.
To date, we still have not found any threads out there;
and as you know, we have not found any of the Columbia
debris laying on the ground out west, based on these
Now, searching the radar data bases is
relatively labor intensive. Clearly, putting people on
the ground out there to search even 5 square miles is labor intensive. We have since started testing various
shuttle components up at Wright Patterson Air Force
Base at the Air Force research lab.
Our initial focus was on that Flight Day 2 object and to try to determine what we could do to identify what may have fallen off the orbiter or fallen out of the orbiter if, in fact, that's what that object is attributable to. So for those radars, we specified a list of thermal protection system, predominantly a couple of different of types of tiles, a couple of different types of blanket type of insulation that's on the outside of orbiter. Were also going to send up an RCC panel, a carrier plate, and a horse collar, that frontal seal that goes around the carrier plate. Those are all in work right now. And we sent up some different types of thermal insulation that go in the payload bay.
Once we had that in work, it occurred to
us we could do similar type radar testing also at
Wright Pat that is tuned towards the radars, these air
traffic control radars, that we are looking for our
debris falling down through. And that also is in work.
For many of those materials, that testing, too, has
already been completed and we are expecting detailed
results sometime this week.
By the same token, we are looking to
identify a set of SRB components and ET components and
we'll have the full set tested for the C band radars
who track their ascent, UHF radars who track while
they're in orbit, and then the L band air traffic
control radars that would drop debris down through the
air. All of that is supposed tell us is it reasonable
to expect that we could track the materials that are
most likely to come off the orbiter or, to look at it
another way, how big would those materials have to be.
So would we have to have a tile the size of a car to be
able to track out here, or is it reasonable to think we
could track a single tile or piece of tile? I expect
that we'll have information on that here within the
That gives you an idea how we think we're
going to find any of this debris. Also, as Dr. Ailor
said, the key to finding or looking for this debris is
we know what happened more or less in East Texas, at
least at the gross level. It will be difficult for us
to do trajectory work with the debris we find in East
Texas and back it up to the vehicle and try to
determine what was happening over Texas. This debris
could tell us where the breach started; and if we can
locate some of this and use it to isolate where the breach on the outside of the vehicle started, that's
going to make us immensely smarter on exactly how the
Now, at the same time there are some
folks out at Ames Research Center in California that
are capable of analyzing the spectral data, the
luminosity in the video and still photography, and it's
possible they'll be able to get us some engineering
data on exactly what's burning, exactly what they see
coming off in the plasma wake. Probably the easier of
the two analyses will be looking at the relative
luminosity, and it is possible that by looking at and
measuring the luminosity of the debris in video,
comparing that to the orbiter's luminosity where the
orbiter is not saturating the video, we know what the
orbiter's instantaneous drag is, we can use a ratio of
that drag and the luminosity, compare that to the
debris, and it's possible we'll be able to estimate the
actual drag on the debris, which then makes us smarter
about what's coming off.
Our initial hope was to also get good
enough spectral data to resolve down the actual
material. Unfortunately, we expect that the three
colors we can get from commercial camcorders will not
be good enough. In combination with the distances they were shot through, the fact that a lot of this light
was having to go through both the orbiter plasma wake
as well as some plasma wake around the debris, our
hopes are much lower that we'll get good spectral data,
but we're setting up feasibility tests for both of
those out at Ames Research Center and we expect to have
those tests set up and in the works sometime during the
very near future.
The last thing I'll tell you about is the
miscellaneous sensor analysis we have in the works.
Again, the first one is something that we were
originally very hopeful about and we are much less
hopeful now. And that will be the infrasonic analysis
or infrasonic data. There are various type of
microphones that are set up across the continental
United States and out in Hawaii. They did measure
sound data on the orbiter during this entry. They have
similar data on previous entries. We thought that it
would be possible potentially to bring some of that
sensor data to the orbiter ground track and essentially
give us a calibration or a signature of these debris
shedding events as they occurred across the ground
track. We have since found that the various variables
associated with bringing that data back to our ground
track and back to our place and time are probably going to be large enough that we're not going to be able to
do that. So we expect that not to pan out.
We have various other DOD sensor data
like radar, and then there are other types of data like
that that we also have evaluated and we have put on the
time line. You have seen some of those. Most of that
data also, regardless of the type of sensor, is not
good enough to specify, say, engineering properties or
specify any kind of properties on any individual
tracked object, unfortunately. We had originally hoped
that we would be able to track individual pieces of
debris coming off the orbiter, specify the vectors on
those things, and use those to be smarter to get them
all the way to the ground. And across the board, the
types of sensor data, the external sensor data that we
have is not going to be good enough to do that and,
interestingly, the public video we have is probably the
best data we have to try to find some of this debris
The last thing I guess I could tell you.
On the ground track here, without going into a lot of
detail, I mentioned these blue dots are the
photographers. The white dots you see on the ground
track, each one of those is an individual debris
shedding event. If you stand back and just kind of look at the view from 10,000 feet, you can see that
from California pretty much all the way to Texas you
see a relatively steady stream of objects coming off
the orbiter. Now, there's a few places where you don't
see as much. That doesn't really necessarily mean we
don't have small pieces of debris continuing to come
off the vehicle. It could just be the perspective and
point of view during that phase of flight and the
photographers just couldn't see it. Likewise, you
don't see any of these white dots out here where we
don't have video because we don't have any way of
seeing it, but I think it would be valid assumption
that we are continuing to drop debris all the way
through. And it is likely that if we had video during
this time frame, because we had a lit sky, we wouldn't
have seen individual objects coming off unless they
were relatively large and we saw some bright flare.
MR. WALLACE: Paul, why don't you show
with your pointer there where this west piece of debris
MR. HILL: Let's see. The westernmost
piece of debris was found just south of Lubbock, which
I would say is right around in here. Let me also say
for that westernmost piece of debris, that Littlefield
tile, which is generally how we refer to it, we have done some top-level trajectory analysis on that. We
expect that piece of tile came off somewhere in this
time frame here, potentially while we had video from
Kirtland Air Force Base, but that also is based on the
mass properties and size of that tile in its state on
the ground. Of course, it is part of a larger piece
higher up in the air and it probably also came off much
earlier than that.
ADM. GEHMAN: That trajectory analysis
you just spoke of, that does include true winds of the
MR. HILL: Yes, sir, it does. Now, what
doesn't include true winds of the day is that generic
swath you saw from California all the way to Texas,
although we are in process of putting real winds of the
day in that.
Let me go back up a page in my slides,
please. Now, I don't show the radar threads here but,
again, I mentioned here around this band there is a
radar thread, probably the radar thread we were most
interested in that we followed, where radar thread is
just the long string of radar hits that we followed in
this pattern on air traffic control radar that we think
is a attributable to Debris 6 or some piece of
Debris 6. Now, that radar started right around here. Again, right on the ground track, right where you would
expected a non-lifting object to be, and then it
tracked to the north and east, which also was with the
prevailing winds of the day. So our interpretation of
that is, as that object dropped down into the heavier
air where you would acquire it on air traffic control
radar, which is about 80,000 feet, then it fell
ballistically above that, got down into heavier the
air, started becoming more lifty, started wafting with
the winds, and again then started tracking here in the
north and east as it came down lower. If you look at
the topographical map of that radar site and where that
object lost track, our speculation there is that we
tracked that object to within about a thousand feet of
the ground, which is why we think we have about a
5-square-mile search area for that object out west.
That was everything I was going to tell
you in a big picture and how we're doing what we're
doing. In general, we're continuing with the relative
motion analysis on all these objects. I expect here in
the next couple of weeks we'll have ballistic
footprints rolling in at a relatively regular rate,
starting with Debris 1 and 2 out west and then working
its way east. We also expect that we're going to see
those footprints start to stack up and overlap significantly with Debris 6 and Debris 14, and then
we're working on figuring out from those overlaps how
to come up with concentrated search areas based on
where we think it's most likely we'll find any and all
of this debris out west.
MR. WALLACE: So this piece that you
tracked to a thousand feet above the ground, there's no
question that that arrived at the ground; but is there
a question about a lot of this other debris is likely
to have just been burned up?
MR. HILL: Let's see. The debris we have radar threads for, any one we have radar threads for, if you assume that those are our debris which is still somewhat of an assumption then we are relatively confident that those are on the ground somewhere near where we lose track of those objects. Now, the other things we see in video very likely could have either burned or completely disintegrated from G loads or aerodynamic forces before they got to the ground. We don't know.
MR. WALLACE: When we say 1 through 14,
can you say how many of those came up on radar.
MR. HILL: Well, I can answer a different
way. We have about four key radar threads that we are
searching out west. There's these three here that are in the Debris 6 footprint, and there's another one in
Albuquerque that did not come from this analysis. It
was started based on some folks in Albuquerque who
thought they heard something fall through the sky and
impact the ground, and NTSB found those radar threads.
Now, if you assume those are ours, we are reasonably
confident that those things are on the ground
somewhere. All the rest of these that we don't have
any radar threads for yet may or may not have made it
to the ground. We have just now started searching the
Debris 14 footprint for radar threads. So we could go
another one to two weeks before we get finished
searching all of that radar to determine whether or not
we see these.
MR. WALLACE: In how much of the area
that you're searching are you dealing with snow-covered
MR. HILL: All of these areas out west,
certainly in the Nevada and Utah area, have been snow
covered off and on at least four or five times since
February 1st. As a matter of fact, the primary search
box out there in Caliente, Nevada, was on hold and we
had about 15 percent of that area to finish searching
and it's been like that for two weeks, maybe going on
three weeks now, all because it was snow covered. If you're looking for something small like a piece of a
tile, it's reasonable to assume they're not going to
find that on snow-covered ground.
ADM. GEHMAN: What can you say about
still photography? Has that been evaluated?
MR. HILL: We're doing some work with
still photography. There is photography that was taken
from California, in particular, time-lapse photography
that may yield us the best spectral data. It did give
us a few more cues when we were trying to narrow down
maybe one or two seconds on debris shedding timing. I
don't expect we're going to get a whole lot smarter
from the still photography than that, however; but we
are buying still cameras from many of the
photographers, just like we are on the camcorders so we
can try to calibrate what we're seeing in the film and
get a better idea of what kind of spectral data we can
pull out. In the ideal case, we'll be able to take
some of that still photography and clearly show that we
have aluminum burning in the plasma or maybe silica or
maybe RCC. We'll see.
ADM. GEHMAN: Did you want to talk about
the Kirtland photographs?
MR. WHITE: I can talk about that a
little bit. I've been working on a tiger team to try to understand the images there. There were a number of
images acquired at Kirtland by the folks there who were
doing that on their own time, not using the Starfire
Optical Range equipment. They did have some pretty
sophisticated home-built stuff, but it wasn't the
Kirtland Starfire equipment.
They did manage to get four videos and
three stills. I think some of those have been in the
media already. We are trying a number of ways to
deconvolute those photos to try to make them as precise
as possible to see what sort of images we can get off
of them at that time. There do appear to be some
irregularities in the shape that we see from the still.
We have to still run that down and find out, you know,
what exactly the shock wave field should have looked
like from that point of view around the orbiter at that
time, whether or not we would have expected to see it
look like that, whether we would have expected to see
it be different. As you know, we've already shed quite
a few pieces of debris by the time we got there. We
were also able to pull one more piece of debris out of
the Kirtland video in the two flares that I talked
MR. HILL: Let me put that another way, Admiral. We are capable using the same techniques we used for measuring relative motion from the video, we are capable of drawing pictures of exactly what the orbiter should have looked like to its Kirtland photographers, whether it's for their still photo or for their video. We're then capable of using computational fluid dynamics and projecting what the flow field should look like around those fissures and then we're also capable of taking that and handing it off to plasma physicists like at the Ames Research Center and generating what the plasma wake should have looked like around those still images. Then we can compare those against what's in video and what's in that still photo.
I would caution anybody in reading
anything into either the video or the still photograph
until we've gone through that process. The vast
majority of people that have studied though images are
imagery experts. They're not experts at what the
orbiter looks like during entry, the flow field around
the orbiter, the plasma dynamics or anything like that;
and we're definitely premature trying to read
engineering conclusions into any of those images before
we've gone through that process.
ADM. GEHMAN: Thank you very much.
Members of the board here.
DR. WIDNALL: I want to make sure I understood something that you said. I asked you about whether there was a time gap in the coverage. You said there was basically you don't have any video pretty much between Kirtland and the more spectacular big events.
MR. HILL: That's correct.
DR. WIDNALL: You said that you thought
there you expected that during this time gap that there
probably was continual debris shedding but that we just
didn't have pictures of.
MR. HILL: I think it's reasonable to
DR. WIDNALL: But it also might be
possible that there was, in fact, a catastrophic event
such as losing a wing or something like that.
MR. HILL: While I can't say that technically
DR. WIDNALL: But it can't be ruled out
on the basis of the data that you have.
MR. HILL: It would definitely surprise
me personally that we would have something significant
like loss of a wing that is not covered in the later
video that we have of the main body breakup, based on
what we have in telemetry and we know how the vehicle was flying and we know the sensor data that we have.
My personal expectation is we capture that in the
video, just based on what we see in the time line.
DR. WIDNALL: So where is loss of signal
relative to the gap that you have in the video?
MR. WHITE: Loss of signal is over Texas.
So we have data from the vehicle.
DR. WIDNALL: You're saying you have data from the vehicle that covers this region in time
MR. WHITE: Yes, we do.
DR. WIDNALL: where you don't have video.
MR. HILL: That's right. These red dots
you see here, all of these represent actual GPS vector
DR. WIDNALL: So you do have data during
MR. WHITE: Right. We do have data
through that video gap period. So, yeah, it's highly
unlikely that any large piece of the orbiter like a
wing would have come off, because we still have data
from all of our systems that show that, even though
they were failing, they were still there.
MR. HILL: Another way of saying that, if
you look at the map, is these blue lines show you everywhere we had video. Everywhere where a line is
red on the ground track, we had data coming down from
the orbiter. Then where it's yellow is the LOS time
DR. WIDNALL: Okay. Fine. Then the
second question really concerns this Debris 6 and the
flash. As I understand the observations that were made
in California of the flash, the flash was unusually
persistent and it also was stationary in the
atmosphere. So the question is: What is it? What do
you think it is? Do you think it is aluminum burning
in the atmosphere?
MR. HILL: It is possible that it is
something that burned and came off the vehicle. It is
what you would expect to see if we were to, say, vent a
fluid or if we were to burn something and as we gave
off combustion products, significant combustion
products, not something on the order of, say, one of
our reaction control system jets, but if we were
actually burning something substantial and as we put
that out in the plasma wake, you expect because that
would have relatively no mass, certainly compared to an
object, that those combustion products would
immediately go essentially static compared to the
orbiter or compared to what we consider normal ballistic behavior for an object that has significant
mass. So it is reasonable to assume that something
came off that was very light or that that was some kind
of combustion product like potentially aluminum slag
that also was burning as it came off the orbiter and
then went stationary during the wake.
DR. WIDNALL: So are you ruling out the
possibility that there could be a chemical reaction
that was stationary? In other words, are you assuming
that as soon as it was all by itself in the atmosphere,
it was not reacting.
MR. HILL: I'm not assuming that at all.
DR. WIDNALL: That's what your words seem
MR. HILL: All I'm trying to say is it is
difficult, if not impossible, for us to get much more
specific about what we're seeing technically other than
we see this bright thing come off the orbiter and there
are a handful of things that that could lead you to
believe as to what those objects are or what the
phenomena are, like a flash or the persistence of that
flash. I agree with you that the persistence of that
flash certainly indicates that you either have
continued plasma wake around something or some
continued reaction. The fact that it becomes more or less stationary would also suggest that it is something
that is extremely light, probably more like a cloud or
a combustion product.
DR. WIDNALL: Okay. I just want to make
sure that we're talking the same language.
ADM. GEHMAN: But the best that you've
been able to analyze so far is that flash that precedes
Debris Shedding No. 6 is not merely a disturbance in
the hot air. It's not just a wave or of the hot air or
hot gases around the orbiter.
MR. HILL: Probably not. Just due to its
persistence, it is telling you that it is more than
just something crossing through the wing. Something
else is happening there.
MR. WALLACE: A question on the far end
of the time line. The SSMEs. I've heard some opinions
that those three bright objects you see in the last
daylight video might be the SSMEs. I would like your
opinion on that, and I haven't heard that we've
recovered much of the SSMEs. Do you expect to? What
are your thoughts on that?
MR. HILL: First of all, we do expect that those three bright dots in that Apache FLIR video are main engines or large components of the main engines. If you look at how they're behaving ballistically, they are certainly objects that are very heavy, relatively high ballistic numbers; and because they're bright, they're continuing to move really fast. We also know from radar data and from, in fact, the SSME components we have found in Fort Polk, Louisiana, that, in fact, the engines or the large components that are up did stay intact for a long period of time and did go further east than any of the rest of the video. I don't know personally maybe Doug does how much of each of the main engines we've found. I know that we do have main engine components that have been found and shipped to KSC.
MR. WHITE: Yeah. That's true. I don't
have a reading of how much of each engine we've found.
I can get you that number.
MR. HILL: That does beg a question on what we can learn from post-breakup trajectory analysis. Everything that I have talked about is pre-breakup. My entire team's focus has been pre-breakup. Everything that we have been trying to do is figure out what's coming off as early as possible and where is it so that we have some idea of where did the breach start, what caused the waterfall of events. It is certainly the opinion of the trajectory experts here at JSC that, taking the debris field as we find it in East Texas and trying to reverse-propagate it back to the vehicle is not something we are capable of doing. Again, going back to the FLIR video from the Apache helicopter, you saw all the secondary and tertiary breakups. As soon as you have additional breakups and those objects then become free fliers, they each have their own individual ballistic behavior. They're all now going somewhere else in the sky. We take the GPC we find laying on the ground in East Texas, we can back it up into the sky to some altitude but at some point we lose all truth, we lose all accuracy because that GPC at some point was in an avionics bay which at some point was surrounded by a compartment and at some point
ADM. GEHMAN: What's the GPC?
MR. WHITE: General purpose computers.
MR. HILL: The fact that we know it
behaves ballistically doesn't mean we can take it all
the way back up to the orbiter. At some point it was
surround by other structure. If we could take the
initial main breakup and assume that all the components
we found in East Texas became free fliers at that
point, we could do a pretty good job
backward-propagating those things all the way up to the
orbiter; but we know, in fact, that it didn't happen that way. As Dr. Ailor said, even the individual
components, say, individual pieces of tile that we find
on the ground, whether we find them out west or the,
say, the Littlefield tile, we don't know that that tile
or object came off the vehicle looking like that. We
have a full expectation for something fragile like a
tile that, in fact, it did come apart.
Using some of the video, we know in
several cases that the object, when you go frame by
frame in the video, anyway, as you're looking at the
object, you see a white dot come off the orbiter and
then you see that white dot shower into a lot more dots
and then you see all the light go away. Probably
indicative of something breaking. Now, is that several
tiles coming off together and then flying apart? Is it
a tile coming off and shattering into a lot of pieces?
We have no idea.
MR. TETRAULT: As you're probably aware,
we have found both of the forward corners of left wheel
well structure; and that's where the wheel well door
interfaces with the structure itself. So you have the
inboard and you have the outboard corners, each of
which demonstrates some venting coming out from the
wheel well itself. My question is: If that's, in
fact, going on, wouldn't you have an interruption to the plasma and wouldn't that show itself, to some
degree perhaps, as a flare?
MR. HILL: Maybe. I hate to not be more
specific; but, again it depends how did that hot gas
get into the wheel well, was it flowing in or was it
MR. TETRAULT: We're talking here about
an outflow from the wheel well at the corners, flowing
MR. HILL: Probably.
MR. TETRAULT: And it has an effect on
the tiles at least, it's a guesstimate, 12 inches to
18 inches outboard from that venting. So it's quite a
vent, if you will.
MR. HILL: Possibly. I mean, if you
assume that that occurred pre-breakup and while the
orbiter was intact and still flying through the sky,
it's possible that a jet like that coming out of the
wheel well might change the plasma wake, might change
what the orbiter looks like to video taken from the
ground; but we don't know. It depends on what
direction was the shock, what direction was the plasma
wake flowing in that is normally around the orbiter,
and did that jet actually make it all the way to the
normal plasma wake and cause a disturbance or was it hidden or shielded behind the plasma wake that already
existed around the orbiter. We don't know the answer
MR. HUBBARD: Two kinds of questions.
First type has to do where all the material, raw
material came from. Obviously we owe the public a
great debt of gratitude for such cooperation. Can you
tell us how many different submissions or contributions
there have been and how many you sorted it into and a
little bit about how you determined what was useful and
MR. HILL: Sure. Within three days of
the accident, we had almost a thousand reports.
Probably within a day or so of the accident actually,
we were approaching a thousand different reports that
varied from people calling in or sending E-mails and
saying, "Hey, I looked up in the sky and saw this
bright dot overhead," to, "I saw something happen and I
want to talk to somebody about it," or videos where
somebody called and said, "I have a video and I think I
see something coming off the orbiter," or," I have
still photography and I think I see something coming
off the orbiter. Do you want it?" For the first day
we spent most of our efforts sorting through a stack of
close to 1,000 reports and, within about two weeks, about 3,000 reports that were all across the map. Just
like that. We very quickly figured out if we were
going to learn anything technically or anything of
engineering value, it probably was not going to be in a
report where people say, "Hey, I looked up and saw
something in the sky," unless they said, "I looked up
and I clearly saw something fall through the sky and
smoke was coming off and that thing hit the ground
close to my house." And there aren't very many of
So we very quickly narrowed it down to
let's look for videos as far west as we can, let's look
for still photography of the orbiter in the sky as far
west as we can, particularly time-lapse photography,
and let's look for people that are amateur astronomers
because those people are going to have a lot better
secondary data like GPS coordinates on exactly where
they were standing, exact zoom settings on their
cameras, things like that, or exact time references,
say, in the case of the video.
Within a week we had it narrowed down to
about 15 videos that form the core of what we now have
on this map, with the videos that actually show debris
shedding that we were able to time to within plus or
minus a second. Then we spent some time after that first week or so prioritizing which of those we have
the best celestial cues in, which of those that we
think we are most likely to be able to calculate
relative motion, and then which of those, like Debris 6
and 14, did we think would be so substantial that we
might have a chance of getting them all the way to the
ground and finding them in radar and putting boots on
the ground and go and collect the hard way.
So I would say it took us about a week to sort out the initial round, maybe a week after that before we knew very well which of the videos, which of the stills were going to give us any meaningful data. From there on, it was a continuous process of analyzing the video, measuring the relative motion, generating these footprints, and then searching through radar. And without the public having taken these pictures on their own because, to our great surprise, people are still very interested, apparently, in the space program and these folks got up before sunrise and went out on their own and stuck their cameras up in the sky and most of them also knew exactly where to look in the sky because we knew they were amateur astronomers without those folks, we wouldn't know any of this. I mean, these people are definitely our heroes. And there are about 15 to 20 of these people or these videos that are probably the most key to us having been able to do any of this analysis.
ADM. GEHMAN: We join you in being
thankful for that. We're also thankful for a
crystal-clear morning across the entire southwest part
of the United States.
MR. HUBBARD: Just a follow-up. There
was a lot of debate early on about whether or not we
were seeing some type of just bright gas or whatever.
How confident are you, when you label the event in the
time line of this debris, that it actually is debris?
MR. HILL: I'm not sure how to answer.
We are reasonably confident. Again, I would say I am
confident, if not sure, that many, if not all, of the
things that we labeled as debris shedding events are,
in fact, some object coming off the orbiter. Can I
tell you is it golf ball size or is it the size of this
sheet of paper? I can't. It could very well be
something as small as a marble in most of those videos
and the ones that we think are so significant and that
have gotten us so excited, those things could be golf
ball size. We really don't know. We know relative
sizes, relative motion, but we don't know specifically
what they are. But we are very confident, based on the
way they behave after they separate from the orbiter, that they are, in fact separate ballistic objects that
have mass, in almost all cases. In the case of some of
these flares, they could be something different.
MR. HUBBARD: Just a final follow-up to
this line of thinking here. When somebody sends
something in, how do you determine that it's the real
deal and not cooked up by a photo shop somewhere?
MR. HILL: For one thing, for most of these videos, we have had them for we got them probably within a week. First week to ten days. Well, we got them within a week to ten days of the accident. In some cases we had them before that. It is possible, I guess, that some people could go and doctor them up. My expectation is we got most of these so quickly they didn't have the time to do that.
The other thing is in most cases we have
overlapping videos, so we have redundant cues. In
fact, we are taking advantage of that. We measure
relative motion from one and we go back and measure
relative motion on the other and we compare them. I
would say they would have to be really darn smart to
have doctor two opposite videos and give us the same
relative motion in the two.
MR. WHITE: Our image analysts have also discovered some hoaxes that have been out there in the public and know they're hoaxes. They've also identified some things that have been anomalies or quirks of the way the photograph was taken a jiggle of the camera, for example, that produced an effect in the photo that looked real but was not real, was an artifact of the way the photo was taken. They've also dispelled some things. Some of you may have seen what looked like a triangular shape when we were zoomed in close on the orbiter that appeared to actually be showing the orbiter in some detail. That wasn't it at all. So they have been able to sort out the hoaxes and the false images and the artifacts from the things that are real.
MR. HILL: Actually, most of our early hoaxes and we did get some early on were cars driving down the road with their headlights on. It was relatively clear to us that it wasn't something in space.
ADM. TURCOTTE: One last question from
me. With your analysis of the radar and your being
able to integrate the time line and the photographs
together, are you surprised at the amount of wreckage
that we have, i.e, do we have more than you expected
from that analysis or do you think that you're
surprised at that, at the amount that we do have?
MR. HILL: Pre-breakup, I would say we
continue to be shocked that we had debris coming off
the orbiter as we crossed the California coast and were
dropping debris, clearly had an external breach in the
vehicle and had hot gas somewhere in the left wing for
that significant period of time and the vehicle flew
perfectly, no indication of what was going on at flight
control and virtually no indication of what was going
on telemetry on the ground other than we saw a few
temperature pressure indications that didn't make sense
to us and we had a few sensors that dropped off line.
Aside from that, the vehicle flew like a champ until
right up to the breakup. So that did surprise us.
Now, from things we are finding in East
Texas, are we surprised that we only have 15 to
20 percent by weight of the orbiter? I don't think so.
I think when you first see the debris count and you see
how many individual pieces of debris, our first
reaction was one of surprise, how could we have gotten
that much of the orbiter down from 200,000 feet intact.
Of course, I think you've also seen at KSC what they
have is a whole lot of little, tiny pieces of what used
to be an orbiter. If you go look at it laying on the
ground at KSC, you don't have a spacecraft lying there;
you got a whole lot of nothing. I think that does fit in with what our conventional wisdom was prior to this
GEN. DEAL: Follow-up to Scott Hubbard's
question. Are you still expecting any more imagery, or
do you think the well has run dry?
MR. HILL: No, sir, I think for the most
part the well has run dry. Again, most people
contacted us right away. We had most of the video in
hand within a week. Overall, the support from the
public has just flat been overwhelming. So I would
expect not to get any more in.
Now, there have been two isolated cases
out west of two individuals who strung us along for
several weeks before it finally became apparent to us
that they must have been under the impression they were
going to collect on the Columbia gravy train. And it
did take us a while to figure out while they trickled
an individual image to us or an individual video to us
that is, in fact, what was going on. They must have
discovered this was their 15 minutes, but they are huge
exceptions to the rule. The overwhelming support has
just been fantastic, and I think we have it all.
GEN. DEAL: In the early days when the
Admiral took us to Nacogdoches, there was talk about
everywhere from offering a bounty money incentive for people turning in parts, you know, going out in fields
and looking for parts, to certificates from NASA to
thank them. Are any of those still under
consideration, or are we just in a debris collection
MR. HILL: To my knowledge, we are not
planning on offering any rewards to people to
incentivize them to come forward if they have not
already. I can tell you the folks here that are doing
work have every intention, when the dust settles, to
come up with some formal recognition. We have various
folks we want to recognize. In my team's case, we
definitely want to recognize the people that took these
images for us and made all this possible; and there are
various things that, at the working level, we are
kicking around that we would like to do. Now, I'm sure
the program will do something when this is all over.
GEN. DEAL: Great. Thank you.
One more for Doug. You gave us an
excellent tracing of all the sensory. You've had
plenty of time now to do some reflections on it and
some lessons learned. Anything that you've already
considered that we ought to be thinking about as far as
sensor wiring, sensor location or junction boxes and
how they're constructed?
MR. WHITE: I'd have to say no. It's
probably too soon to speculate on any type of redesign
that we might want to do with our instrumentation. As
you know, the instrumentation wasn't designed to have
flow inside of the wing; and so it probably failed in
the way we would have expected it to. So as of yet, we
have not considered any sort of internal redesign to
better protect that instrumentation or even make more
ADM. GEHMAN: Gentlemen, on behalf of the board, we want to thank you for appearing today; but I hope you will also take back to your working groups, of which I know you are the tip of an iceberg of literally hundreds of people that are working with extreme zeal and professionalism to try and solve this riddle because many of us have visited your working groups and we know how many people are working on this please pass on to all of them our deepest gratitude and our deepest respect for the work that you all have done and will continue to do. We appreciate it very much. We haven't solved this thing yet, but someplace in your work we'll find the answer and we appreciate it very much. Thank you for appearing here today. We appreciate it.
(Hearing concluded at 4:24 p.m.)
Back to Event