NASA STS Recordation
Oral History Project
Edited Oral History Transcript
Julie A. Kramer White:
Interviewed by Jennifer Ross-Nazzal
Houston, Texas – 26 April 2011
Today is April 26, 2011. This interview is being conducted with Julie
Kramer White in Houston, Texas, as part of the STS Recordation Oral
History Project. The interviewer is Jennifer Ross-Nazzal, assisted
by Rebecca Wright.
Thanks again for joining us this morning. We certainly appreciate
it. I thought I'd begin by asking you if you could give us a brief
overview of your career at JSC.
Sure. I'm basically JSC Heritage Engineering. I've been matrixed to
different programs over the years, but I've always been a part of
the JSC's engineering institution. I started as a co-op in 1986 and
have spent the last twenty-five years here at JSC. I started out in
the Structures and Mechanics Division [currently Structures Engineering
Division (SED) or ES], and I've worked all the different branches
within that division. Eventually I was assigned in the early nineties
to be a structural subsystem manager for the Orbiter and had a mentor
who was the existing system manager, a man named Stan [Stanley P.]
Weiss. I worked with him for several years until he retired, and then
I became a subsystem manager. I was responsible for the wings and
the tail and the control surfaces of the Orbiter, the elevons and
the body flap, responsible for structural certification, repair modification.
I did that for several years, and then at the point in the late nineties
where we transitioned out of NASA subsystem management into more of
a contractor managed effort, I transitioned off Shuttle, worked X-38
and ISS [International Space Station] for a couple years, in a fire-fighting
capacity, whatever they needed from a structures perspective, and
then went back to Orbiter, and matrixed into vehicle engineering [EA4].
[It was] a similar job to what I had done before, only more of a systems
engineering, more broad application, not just structures, other systems,
but obviously always with a bent towards structural failures and things
like that. That's the [kind of] problems I worked for the Orbiter
I worked there until [Space Shuttle] Columbia [STS-107 accident],
and then I worked the Columbia investigation for several months. Then
after that I went to the NESC [NASA Engineering and Safety Center]
as a mechanical loads expert in the startup of the NESC. Stayed there
for about three years and then came back [to Engineering] into my
current position as the Chief Engineer for Orion capsule.
Quite a broad experience at JSC.
A lot of good opportunities. Right place at the right time.
You know a great deal about Shuttle, and I was interested in learning
more about the pre-flight certifications of the Orbiter. Would you
tell us a little bit about how those operated and if they evolved
over time as you were working in structures?
Sure. A big part of my job in structures as a subsystem manager was
certification of the primary structure. By the time I arrived, most
of the fundamental certification was done. The structural test article
had already been tested and the models all correlated. That was all
done in the late seventies, early eighties, getting prepared for the
first flight of Shuttle.
By the time I arrived, most of the emphasis was on maintaining the
vehicle, so a lot of effort into structural inspections. Initially,
if you go back and look at the program formulation, Shuttle was to
fly many, many times a month, many, many times a year. Original certification
was for ten years, a hundred flights, and by the time I joined the
program, [some of] the vehicles were already ten years old.
So a big part of my job was working with the team to make sure we
were implementing structural inspections and basically reeducating
people about the nature of maintaining an aging vehicle versus a vehicle
that was certified one time and then flew for its life and then was
retired. We clearly weren't going to retire the Shuttles anytime soon,
and so a big part of it was putting in place the measures to keep
the certification sound, even though it was flying much longer than
we originally thought.
We did do some [delta] certification, mostly around increased performance,
so if the program wanted to fly a different trajectory, or the program
wanted to land a heavier payload, or they wanted to land faster, we
would do certification around that. Rockwell [International] would
go and do the analysis. We would work with them; we would review it.
As a team, we would figure out what modifications needed to be made
to the airframe to support those new trajectories, or if they could
do it totally analytically.
Some of the later assessments you may hear some of the people [in
Orbiter that you interview] talk about “Performance Enhancement.”
Some of the later loads sets for the Orbiter, we were able to do totally
analytically without any major modifications, but there were several
modifications to the primary structure throughout the evolution of
Very early on, right after OV-102 [Orbiter Vehicle Columbia]—I
always use Orbiter numbers; it's hard for me to use names. After OV-102
flew the first several flights [STS-1 to 5 in the 1981-83 time frame]
and we realized how high the loads were on the wings, there was a
set of modifications done right away, basically called the double-A
modifications [done in Palmdale, California, between August of 1991
and February of 1992. Columbia underwent approximately 50 modifications,
including the addition of carbon brakes, drag chute, improved nose
wheel steering, removal of development flight instrumentation and
an enhancement of its thermal protection system]. Some of your [NASA
or Rockwell] people may call it the “Death Mod,” because
it was done so quickly over such a short period of time to return
that vehicle to service. It put major doublers into the wings and
straps on the wings and reinforcements to the spars of the wings so
that the wings could take the loads on ascent.
There were mods that were rolled in, particularly as OV-103 and OV-104,
Discovery and Atlantis, as they were being built. They were being
built simultaneously, basically, one just behind the other in the
flow; they rolled a lot of those modifications in line and did them
as they were building the vehicles. Obviously, [OV-] 105 [Endeavour],
because it came so much later, had those mods just built into the
By the time I made it, it was mostly maintaining certification and
doing inspection to make sure the certification was still valid, making
repairs to the airframe as necessary to keep the certification sound,
so that was more the nature of a lot of the work we did.
Can you give an example of making some modifications to the airframes
in terms of the trajectory or the payload?
Sure. If you’re going to land something heavy, there’s
a lot higher loads on the wing, and so it’s pretty simple, really.
It’s not that much different than airframe engineering for an
airplane. You land it harder, and you need to worry about can the
landing gear tied into the airframes take the loads, and then you
just follow that load all through the airframe. Generally, where we
would have to make modifications would be in the wing or the wing
root, where we would have to put in doublers or straps to make sure
the wings basically didn’t yield or fail on a hard landing.
It’s pretty straightforward stuff, actually. It’s actually
real similar to aircraft modifications.
Not being an engineer, it’s kind of interesting. I had no idea
all these changes were going on. Tell us about certifying the vehicle.
You had to sign off?
Absolutely. That was part of a subsystem manager’s responsibilities,
was to sign certification of flight readiness for every flight. We
did that for the project manager. That was part of our matrix responsibilities.
That would generally entail being cognizant of what the mission was;
and therefore what demands were going to be placed from me, in particular,
on the airframe and on the control surfaces, which trajectory we were
going to fly, what landing we were going to fly, make sure it was
all within the certification of the airframe.
We would have to review any kind of MR [material review] activity.
If there was a problem at [NASA] Kennedy [Space Center, Florida (KSC)]
or a problem at Palmdale during its major flow, we would have to review
the paperwork and make sure we felt like the certification was still
sound, based on whatever the problem was and whatever the repair was.
Obviously, we were involved in all that real time. When it came to
the certification, that’s where you rolled it all up and you
sat it down in front of you and you looked at it all together and
made sure holistically it all held together and it made sense. You
hadn’t made a series of smaller repairs that, in the end, compromised
the overall system. So it involved all that.
We did inspections during that interval. If it happened to be at a
depot period, OMDP [Orbiter Maintenance Down Period] or OMM [Orbiter
Major Modification], we’d go back and review all that and make
sure we understood what mods were made and that it all hung together
to keep the certifications sound. And, yes, then we signed the certification
of flight readiness and participated in the Flight Readiness Review
[FRR] process if necessary.
Generally, your structures guys didn’t show up at the FRR or
in the MER [Mission Evaluation Room]. Obviously, a lot of people have
stories about being in Mission Evaluation Room, but generally, if
your structures guys showed up after you were at the pad that was
a bad, bad thing. We didn’t spend a lot of time with the program
managers after the vehicle rolled to the pad. Obviously, our big push
was getting it out the door. But once it rolled to the pad, we generally
didn’t interact with them a lot, didn’t interact, certainly,
during the mission, very seldom.
It was so unusual, at least in the early days in the nineties when
I was there. It’s more common now post-Columbia; you see a lot
of discussion of TPS [Thermal Protection System] damage, which engages
the TPS guys and the structures guys more into the MER. But when I
was there, it was so uncommon. I can actually remember one time on
a Friday night being called to the MER. Brewster [H.] Shaw was the
program manager, and I can remember being called. It was so unusual
to be called over to the MER. There’s a rudder speed brake on
the tail, which was one of my areas, and it’s actually a left
and right panel that opened to slow the vehicle when it lands on the
runway. It was a split rudder, meaning it had an upper panel and a
lower panel on each side, and it had a bulb seal between the two,
upper and lower panel.
The crew member was looking out the back windows, and he sees that
on ascent, the bulb seal has protruded and it’s not bonded in
place anymore. They show me these pictures, and they say, “What
do we do? What do we do?”
I said, "Well, you can reenter. If I tell you you can’t
reenter, are you going to do anything about it?”
They said "Well, no, not really." At that point, pre-Columbia,
really people didn’t think about going EVA [Extravehicular Activity]
to do repairs too much. So they said, "Well, no, no, we’re
not going to do anything about it. We just want you to tell us it’s
I said, “Okay.”
So this would be pretty typical. We go off, we do some analysis. It’s
hard to analyze something like that exactly precisely, but we’d
do the best we could, and we’d put some bounding conditions
on it. We’d say “Well, you know, worst-case scenario,
you’re going to melt part of the rudder speed brake. It’ll
probably still be fine. It’ll be effective enough, it shouldn’t
be a problem.” At that point, we had the drag chute. “A
drag chute should slow [the vehicle] down. Won’t be a problem.”
But that would be the kind of thing you’d do. We’d work
with Rockwell, or later on in the program we’d work with Boeing,
and review with them the analysis and what assumptions they were making
and make sure we felt good, comfortable with the level of conservatism
that was in the analysis. Then we would make a recommendation to the
program manager to say, “Yes, we think it’s really okay.
Just go ahead and reenter. It’ll be fine. Anything that happens,
we can fix it after you get it back. It’s not going to be a
So that was one of the very few times I can actually remember getting
called to the MER, because you just didn’t have structural problems
What mission was that, do you recall?
I don’t. I actually was looking for it in my files, and I couldn’t
find it. I know Brewster was the program manager, because I remember
talking to him about it. I probably could find it for you if I looked
We can take a look too. I was just curious.
There’s some mission where they had a protruding bulb seal on
the rudder speed brake. I’m sure you could probably find it
because it would have been in the MER discussions. But it was very
unusual, as structures, to get called to the MER.
Were you called fairly soon after they had gotten up into orbit, or
was this close to landing?
I think it was fairly close to landing. I think it was one of those
things. This, like I said, was way before Columbia, so they weren’t
looking out the window for damage. It just so happened, whatever they
were doing looking at the payload bay, the guy looked at the tail,
or gal, looked at the tail and went, “Oh, that doesn't look
good. This thing’s sticking out of the tail.” It was long
before we did that kind of inspection normally.
So it was a pretty quick analysis, then, that you did.
Yes, I think probably within twenty-four hours we had gone through
whatever analysis we were going to go through and had established
that we felt like the risk was perfectly reasonable. It might cause
damage to the vehicle, but we could fix it. It wasn’t going
to be that big of a deal to fix it.
Tell us how long it takes to certify a vehicle, on average.
Oh, gosh. That depends a lot. It depends a lot on what they’re
doing. Something very minor [like a mission specific analysis or material
review analysis] is on the order of days. If it’s just I’m
changing a design and it’s similar enough to heritage design
or similar enough to previous applications, it might be a matter of
days or weeks. If it’s a major change, something like “Performance
Enhancement,” which was the last load cycle we went through
on Orbiter, it was a year of analysis that Boeing wound up doing,
or more, before we finally crossed all the t’s and dotted the
i’s and said, “Yes, okay, you’re good. Your certification
is good.” Those cycles were multiple-year cycles sometimes on
the loads. By the time they redid all the stress analysis for the
whole vehicle, because in many cases if it was a different-enough
trajectory or if it stressed the whole airframe versus something that
maybe puts extra stress on the landing gear, puts extra stress on
the tail, those were smaller in scope and you could do different things,
or repairs would be very limited in nature. It’s all very local,
so those could be done quickly.
Something like “Performance Enhancement,” where we were
basically asking to fly the vehicle differently and expose the whole
airframe to different loads, those cycles would take a year or more,
because there are just dozens of volumes of analysis. If you see an
Orbiter stress analysis, it would take up this wall, this wall of
cabinets, just because we generated paper for all that. Now, of course,
it's all done on the computer and kept in the computers; you don't
appreciate how voluminous it is. But when we actually originally certified
the vehicles and did paper analysis and printed the books, it was
Tell us a little bit about the stress analysis. What did that involve?
Was that primarily computer-based or were you also using wind tunnels?
Clearly the loads were established using wind tunnels back in the
day—“back in the day when the crust cooled.” A lot
of it was done by hand-analysis, so they would take the wind tunnels
and they would take their basic analysis and drive the external loads
for the vehicle, and then they would load the airframe. And at that
point, they were using the precursors to NASTRAN [NASA Structural
Analysis software], and we certainly could get you the right people
that could answer your question specifically about what the analytical
tools were. Somebody like Glenn [J.] Miller in ES could tell you more
historically kind of what tools they used.
We eventually migrated everything to NASTRAN. They had some historical
codes that they used at Rockwell. I think in the end we eventually
migrated everything to NASTRAN, because that’s a more commonly
used code now. But, yes, they did some basic finite element to distribute
the load through the airframe, pretty crude by today’s standards,
but good enough, certainly, and then coupled with a lot of hand-analysis,
actually. If you look at the volumes and volumes of stress analysis,
a lot of it was done by hand. It would be using finite element models
to distribute the load through the airframe, to basically get it from
here’s your external air loads while you’re flying, here’s
your external environment on the airframe, and help distribute that
load and then flow all those loads through all the primary structure.
But when you got to a bolted joint or a certain aspect of actual structural
design, all that was done by hand, and you can see that in the stress
reports when you look at them. Nowadays, it’s all automated.
It’s all done in tools like MATLAB [Matrix Laboratory programming
language] and [Microsoft] Excel and basic spreadsheet-type tools or
basic computational tools that they just didn’t have then.
By the point where I was graduating, people were using NASTRAN, so
in the mid-eighties, people were using NASTRAN and people were starting
to use tools like Excel spreadsheets, but things like MATLAB didn’t
really exist, or if they did exist, they existed in formats in labs.
They weren’t really commonly used outside. But nowadays, all
that’s automated. They just plug the equations in, and they
might be analyzing a series of stringers or a series of hat sections,
and it’s all done in computer code now. They just change the
critical parameters and the computer does it all, whereas they used
to do it all by hand.
Yes. If you look at the volumes, a lot of them are works of art. They’re
just a lot of hand sketches. The vast majority of it is hand-analysis
with the exception of how the loads were distributed, which was done
by finite element models.
When you were working with Rockwell and Boeing, were you working with
them on their efforts or were you primarily just seeing the reports
that they generated?
A lot of our work was day in and day out with Rockwell, the vast majority.
I did work with a handful of NASA guys. By the time I got there, the
airframe was divided into three areas, and I had the wings and the
control surfaces. A guy named Trevor [R.] Kott had the midfuselage
and the forward fuselage. It seems to me he and I traded the aft fuselage
back and forth, depending on who drew the unlucky straw. Then we had
another gal named Lynda [R.] Estes, who did the crew module and the
windows, because those were kind of specialized. We divided the airframe
Then we three worked a lot with Rockwell. That was really who we worked
with on a day-to-day basis was Rockwell and Boeing, and so we had
our analytical guys we worked with and our leads, our counterparts
at Rockwell or Boeing, and we would work with them to resolve problems
and figure out what we were going to tell the NASA project manager.
“Oh, we can’t tell him that. We need to work on that one
some more.” So we would spend a lot of our time either on the
phone with Rockwell or at Rockwell. I spent probably half my first
five years, I bet, on the road at either Palmdale or at the Rockwell
facility in Downey [California].
I was curious about that. You must have spent a lot of time on the
When you came, Endeavour was a fairly new vehicle. Did you take part
in certification of that vehicle?
I did not. When I showed up full-time in ’90, of course, I went
through basic intern-type training. The point at which Endeavour was
being delivered, I wasn’t in a subsystem manager role. Matter
of fact, I can remember the first time I traveled with this mentor
of mine, Stan, he took me to Kennedy, and Endeavour had just been
delivered to Kennedy, and I had never been to KSC before. I showed
up and we go into the OPF [Orbiter Processing Facility], and I’m
so excited. I’m probably twenty-three or something, I don’t
know, twenty-four. I walk in and I’m like, “Where is it?
Where is it? Where is it?”
And he says, “Look up,” because we were underneath the
belly of it, and it’s probably as tall as your ceiling. So I
was literally standing in the OPF underneath it and didn’t realize
I was standing under it. He says, “Look up.”
And I look up and OV-105 has never been flown, so it’s black
shiny reflective almost on the bottom with all the new black RCG [Reaction
Cured Glass] tile on it. I still remember. That’s been over
twenty years ago, but such a huge impression to see a brand-new Orbiter
and just look up, and you’re like, “Oh, my god, there
it is,” right there above you. He thought that was pretty funny.
So, yes, quite an impression. But it had been delivered, and so I
worked with him, obviously, cleaning up, getting ready to fly it.
I’m sure there probably was some stuff I did with regard to
cert [certification] for that, but I just don’t recall the specifics.
I was probably just so overwhelmed that I was working on a Shuttle,
I can’t even remember anything I was doing.
I saw that on your résumé and thought, “I wonder
if she actually did any work on Endeavour,” because I thought
that would have been really cool, so it’s neat that that memory
has stuck with you for so long.
Tell us about how you were able to juggle all these different orders.
You’ve got a fleet of four Orbiters, and you’re this subsystem
manager. You’re certifying them, but you’re also getting
them ready to undergo major modifications out at Palmdale, and you’ve
got missions flying. You’ve got about seven missions flying
every year that you’re in the position. How do you maintain
flight readiness, getting them ready? A big challenge.
We have a lot of help. The project guys are there, and we had vehicle
managers on the project side that helped make sure that they kept
cognizant of what work was being done out at OMDP or what work was
being done in any flow, and they helped with the communication. We
[the project vehicle managers] know this is going on at the Cape [Canaveral,
Florida]. “We know you [the SSM (Subsystem Manager)] would want
to be aware.” So they would help keep that communication flowing.
They would help keep decisions queued up and moving forward. “We’ve
got to get back to the project manager with this. We’ve got
to make a decision on this.” They would work with their Rockwell
and KSC counterparts on making sure we didn’t drop the ball,
we didn’t forget we had something in the queue that we were
trying to deal with. So they would help us keep things queued up and
moving forward in a timely fashion.
Then obviously we had KSC engineering counterparts, NASA KSC, and
over the years RSOC [Rockwell Space Operations Company] or Lockheed
or USA [United Space Alliance]. We had those engineering counterparts
at Kennedy, and we had Rockwell or Boeing counterparts on the West
Coast. So depending on where the vehicle was and what was being done,
you had people that were really watching and doing on a day-to-day
basis, and as a subsystems manager, your job was keeping an eye on
a lot of that. These various people helped you by keeping track of
the odds and ends and keeping it queued up when you needed to make
a decision, so that wasn’t so bad.
Actually, over the years, you eventually learn each of the vehicles
was like a child. It had its own things. It had its own idiosyncrasies.
It had its own design. One of the things you asked about was, “Are
they different?” Well, yes, they’re all different. Their
biggest differences were between [OV-]102 and then [OV-]103 and subs
[subsequent, as in subsequent build]. We literally talked about it
that way. That’s the way the engineering was released. You had
the structural test article build, [OV-0]99 [Challenger], that eventually
became an Orbiter, Flight Orbiter, and then you had ALT [Approach
and Landing Test], which was done with Enterprise [OV-101], which
was its own thing. Then you had 102, which was kind of its own thing,
and I’ll give you some more specifics, and then you had 103
and subs. So [OV-10]3, [OV-10]4, [OV-10]5, the engineering being very
similar but quite a bit of difference between 102 and 103 and subs,
102 most visibly. People would have recognized it had a SILTS [Shuttle
Infrared Leeside Temperature Sensor] pod. We never took the SILTS
pod off. It had the bulb on the top of the tail. If you look at it,
it’s got this bulbous kind of thing on the top of the tail.
Then there was the debate do we take it off and put a normal tail
It was one of the early DTO [Detailed Test Objective] flight experiments
that was done. So the pod stayed there. We eventually removed the
equipment, but the pod stayed there because it was too expensive to
remove it, and the certification was done with the pod there, so why
bother removing it? Why pay the money to remove it? So we just always
worked around that uniqueness, and that’s probably the most
visible attribute that people would notice. But there were just a
lot of differences.
[OV-]102 is fundamentally a vehicle that’s built up from structural
pieces. Frames are bolted to panels. You look at the aft fuselage,
and there’s big structurally integral machine panels with these
big frames that are bolted together to build up the aft fuselage.
A lot more of 103 and subs is integrally machined. So you built 102
and you say, “Well, that went well,” or, “That didn’t
go well.” You redlined the engineering and you said, “Next
time I build it, I’m going to do this, and it’s going
to make it easier to build it and it’s going to be lighter weight.”
So there were weight-savings modifications that occurred between the
two, and manufacturing-enhancing changes between 102 and then 3, 4,
and 5 to make more integrally machined structure, to try to make it
lighter weight, and make it easier to build. So those kinds of changes
[OV-]102 has what you’d call honeycombed forward spar, and the
other vehicles have corrugated forward spars. Probably wouldn’t
think about it too much except it came into play when you talked about
Columbia and the disintegration on the wing on Columbia, and it came
into play just on a day-to-day maintenance perspective because it
was quite different to maintain 102 with its honeycombed leading edge
compared to these heavier corrugated wing leading edges that were
on the later vehicles. So it was just a reality of maintenance of
the different vehicles.
[OV-]103 and subs, the later vehicles, had composite spars in the
wings, and that was always interesting when you moved from vehicles
that had composite spars to vehicles that didn’t have composite
spars. The technicians were like, “Well, how come my drill bits
keep getting dull on this black aluminum?”
And you’re like, “No, it’s not black aluminum. It’s
composite.” And that sounds awful, but thirty years ago, it
just wasn’t that common. It’s certainly not state of the
art today, but at the time was state of the art in teaching the technicians,
“No, you can’t do it that way. You’ve got to do
it—[this way].” You’ve got engineers out there working
with the technicians going, “No, no, no, don’t do it that
way.” Of course, the technicians taught us a lot too. A lot
of time on the floor at Palmdale working with the technicians just
on how to maintain it, because nobody ever built it to maintain it.
The access was not in the vehicle. It wasn’t really intended
to be flown for thirty or forty years, have people climbing around
it doing inspections, or have them climbing around it doing modifications.
It just wasn’t built to do those things.
A lot of the work we did was trying to figure out how to work our
way around the limitations inherent in the engineering, that if we
had thought thirty years ago, “We’re still going to be
flying it thirty years from now and trying to inspect it,” we
would have done a lot more. There were things done initially to try
to make sure that you could inspect and you had some inspection doors,
but it really wasn’t designed to be maintained for thirty years.
So from an inspection perspective, it was always a challenge to do
There are definitely physical differences between the vehicles, and
then even where the engineering is the same on 3, 4, and 5, they all
had their own idiosyncrasies based on how they were built. So you
knew when you built 105, we had a harder time getting the wings on,
so this doesn’t quite look like that, and it’s been adjusted.
If you pulled off tile, you’d find this had been done to make
it match up. So you just knew from over the years of working with
the technicians who had built the vehicles and working with Rockwell,
who had built all the vehicles, a lot of continuity was maintained
as we transitioned from our subsystem managers who came from Apollo
into early Shuttle.
Most of the subsystem managers transitioned, so in the early nineties
when I came on board, NASA was making a transition, and so we staggered
those transitions. Rockwell maintained us through our transition and
then Rockwell transitioned, and we helped maintain Rockwell through
their transition. We always staggered those transitions so you didn’t
wind up changing a system manager on the NASA side at the same time
you changed a system manager on the Rockwell side. There was always
somebody there who could remember, “Oh, here’s why that
one’s unique,” or, “Here’s what’s going
on with that one.” It was a good symbiotic relationship with
Rockwell and then Boeing in terms of maintaining that continuity of
knowledge on the vehicles, because they all were like children and
had their own thing.
You’d be surprised—well, if you have kids. I have one,
but if you had four, you remember that kind of stuff. You just knew
that guy’s different because it’s put together this way,
and that guy’s different because it had this problem. Back on
flight, blah, blah, blah, we did this, or back on such-and-such a
flow, we backed 102 into the work stands, and so it’s got a
splice in the elevon, a splice that came from Enterprise.
So now we’re talking about moving Enterprise, and I get calls,
“Can we move Enterprise?”
“Yes, sure, you can. Just don’t forget you’ve got
the splice in the trailing edge.” So even now I periodically
get calls from the guys that are working transition of moving Enterprise.
“Can we move it? Can we ferry it?”
“Yes, probably. Do these things. Check with this guy. Don’t
forget this.” That’s just how it works. Each one’s
kind of different.
Interesting. So you learned a lot from the technicians and the managers.
Did you also have any sort of book that you would all keep on the
Orbiters themselves, or anything that you had about the specific structures
you were handling, the tail?
There were some basic contract deliverables that came with the Orbiters.
Some particularly handy ones in terms of, obviously, the stress analysis
was a deliverable and is a record of a lot of the idiosyncrasies of
the primary structure, including the control surfaces. There were
things called build-flow diagrams, which essentially are books that
show how the Orbiter was built. It’s like your Ikea furniture.
“Take this piece and put it with piece and it makes this piece.
And you take this piece and you put it with this piece, and now you
have this sub assembly.” And it literally goes from piece parts
to full Orbiter assembly, and it will show you by drawing number what
pieces you’re putting together. It’s a fabulous document,
by the way. I think I have some if you would like them.
Yes, it sounds fascinating.
If you’d like them for historical purposes, I think I might
have one or two of them. But it’s a fabulous document, like
if you’re trying to understand how the vehicle’s put together
and what engineering you need to reference. We obviously had engineering
logs, drawing trees. One of the other unique things about the Orbiter
is it was essentially farmed out to different subcontracts. The wings
were built by Grumman [Aerospace Corporation], tail built by Fairchild
[Industries], forward fuselage, crew module built by Rockwell, payload
bay doors built by Rockwell-Tulsa [Division], GD [General Dynamics
Corporation, Convair Aerospace Division] for the midfuselage, aft
fuselage was built Rockwell also. So everybody had their own drawings,
and we all sort of specialized. I knew how to read Grumman drawings
because I did the wings, and my buddy Trevor knew how to read GD drawings
because he did the midfuselage, and we all knew how to read Rockwell
drawings. And they were all different. They all used different designations.
They all used different coordinate systems. So you really had to learn
to translate between how Grumman did their engineering designations
and engineer drawings and how Rockwell did. So we were all translators
of our own whatever subcontractor we happened to be responsible for
covering. It was interesting in that regard too.
I had Grumman drawing trees and things that would help me navigate
the drawing system, because it wasn’t the same as the Rockwell
system. We had maintenance documents. Maintenance documents were never
delivered with the vehicle, because it was never intended to be maintained
over that long a period of time. So a big part of what I did in the
years I worked with them was help develop maintenance documents, not
only as archival records but for working documents for Kennedy and
Palmdale to be doing inspection. A big part of it was just what inspections
are you supposed to be doing, what are the work instructions, how
do you get the inspections and the work instructions tailored to each
unique Orbiter, because each one’s different. So you’d
tell them to do an inspection and they’d go do it on 102, and
they’d go, “Oh, yes, that worked great.” Then they’d
go to do it on 103, and they’d go, “I can’t do an
inspection because this bracket’s in the way,” or “That
door’s not here,” or whatever.
So you’d be out at Palmdale, and you’d be rewriting the
inspection criteria or you’d be out there with the technician
while he’s doing the inspection, and he’d say, “Is
that what you want to see?”
“Well, no, that’s not really quite what I want. Let’s
try this. Let’s try that. Let’s take this door off. Let’s
do this. Let’s get a boroscope. Let’s go do an ultrasonic.”
So you’d do whatever you had to do to figure out how to get
the inspection, and we became sort of the only people that knew the
So we had OMRSD [Operations Maintenance Requirements Specifications
Document], which is the inspection document, which includes the structural
inspections. They’re called V30s. Each subsystem had its own
designator, V09 for TPS, or V whatever for ECLSS [Environmental Control
and Life Support System], so V30, 31 for primary structure, secondary
structure. The V30s defined all the inspections they had to do and
when they did it. In the early eighties we figured out, “Maybe
we’re not going to fly them as often as we thought we were.
Maybe we’re not going to want to retire them when they’re
ten years old.” So they already had started to realize, “This
maybe isn’t going to work the way we thought. We better put
some sustaining engineering structural inspection-type stuff into
We already had the subsystem stuff down pretty good. I had to make
sure all my ECLSS subsystems were working between flights, check those
out, normal checkout between flights. Are the batteries going to work?
Is the toilet going to work? Is the power, all that? They had all
that, but nobody really thought about the primary structure, built
for ten years, hundred flights, don’t have to do anything to
it. Well, by the early eighties, mid eighties, they’d already
figured out, “Mm, that’s not going to work. We’re
going to fly them way longer than ten years.” So about ’83
or ’84, they went off to Pan Am. Remember the big airline?
Way back in the day, the airline Pan Am. So they went off and talked
to them about how they developed their structural inspection for their
aircraft, what the critical parameters were, how would you do this
for a Shuttle. “Let’s talk about how it was designed,
how it was built,” because those things are important in airplanes.
Even though it may look like a big kind of chubby airplane, it’s
different. It’s how it’s designed. The factors of safety
that are used are different than they are for aircraft. Aircraft has
very stringent failsafe requirements, like it has to be able to fly
with a tear in the skin. It has to be able to fly with cracks. Yes,
don’t look under the floorboard of your airplane. Don’t
look at me like that. Don’t look under the floorboard in the
toilet of your airplane, or you would not ever fly again.
So it has a lot more failsafe requirements than a Shuttle does, mostly
because they can afford to carry the weight. Being able to fly with
a big rip in your fuselage requires that you carry bigger frames to
carry the load. Well, we couldn’t carry all that extra weight
and meet our mass margins getting off the ground, so we just are built
different. They’re fatigue-critical. They do these ground-air-ground
cycles, pressurize, unpressurize, pressurize, unpressurize, multiple
times every day. So they’re driven by fatigue. We don’t
tend to be driven that way. We tend to be driven by static load requirements.
So what drove our inspections were different. They worked with us
on that, and they talked to our system managers at the time and said,
“We’ll tell you what we know about doing aircraft maintenance
and inspection, and you tell us what you know about how your airframe
is designed. We’ll figure out an inspection routine.”
They went and they figured out an inspection scheme. It was never
intended to be the forever inspection scheme. There was no reference.
Shuttle was the first reusable, obviously, airplane to space. So there
was no precedent.
They took what was available as an aircraft inspection program and
morphed it to the design for Shuttle and laid in a set of inspections
that wound up starting in the down period after [Space Shuttle] Challenger
[STS 51-L accident] So it was in progress, Challenger happened, and
really started implementing it in all earnest after Challenger. In
the down period after Challenger, all the vehicles were inspected.
Then the idea was that it would be a living document, that as we learned,
we would change it. We set a set of intervals. While we’d go
do those inspections, if we found anything, we would adjust the inspection
intervals. If we inspected and we inspected several times and we never
found anything, then we would lengthen the intervals [for the subsequent
vehicles]. Unfortunately, when you’re only flying a few times
a year, it takes a long time to build that history on the airframe,
but that’s what we did over time. We built that history. We
built the work instructions to implement the inspections. We modified
the work instructions as required to meet the needs of each of the
vehicles, and we would modify the inspection methods. We’d either
use visual inspection or we’d use ultrasonic inspection or we
would use x-ray.
As inspection tools matured over that twenty-year period, obviously
ultrasonics improved a lot, eddy current has improved a lot. There’s
all kinds of new x-ray techniques. We would go out, and we would solicit
from outside NASA, at Southwest Research [Institute] or Sandia [National
Laboratories] or out in the airlines or at [NASA] Langley [Research
Center (LaRC), Hampton, Virginia], wherever they were doing more of
this aging aircraft work. What could we use? What tools could we use?
We would bring people in, because we had some unique challenges trying
to inspect vast acreage under tile. We weren’t going to remove
tile. We weren’t going to remove all the blankets on the outside
of the vehicle. Airplane doesn’t have that problem. Airplane
goes through its major depot. It’s stripped of its paint, it’s
inspected, and then you repaint delivery on it and you send it back
out to fly. We can’t do that. We couldn’t do that with
We had our own unique inspection challenges, so we kind of brought
in a lot of these people from outside and said, “Well, what
would you do, or what tool would you apply, or what technique would
you apply?” We would have to certify these techniques to try
to use them on Orbiter, involving trying it and trying samples with
known defects and developing what you’d call a probability detection
curve. Are you going to see it? Are you really going to see it? How
small could it be? How would you set the machine to do the inspection,
and if you set it this way, will it find it? And if you set it that
way, it doesn’t. So you’d have to develop all these work
instructions and show the technicians how to use the tools. We did
a lot of that over the ten years that I worked in inspection.
I worked a lot with Palmdale on that kind of stuff. Palmdale was really
in many regards the proving ground, where we proved a lot of those
techniques. Once they were more mature, we would tend to farm them
out to KSC, and they would use them between flights if we needed them.
But a lot of times that was where that kind of development occurred
because they had time in major mod to do that, whereas we didn’t
have time in the standard flow to go do that kind of development work.
That evolved over the years, and I’m sure it has evolved since
I’ve left and probably until the day they all roll to a stop,
it’s all evolving. A big part of our job was to study the results
of those inspections. So we’d do the inspections, and pretty
soon we’d have a history of having done that inspection on all
four vehicles. Then pretty soon we had a history of two or three sequences
of doing that inspection on the same vehicle, and how did the results
vary and was the inspection effective and did I find anything? Would
what I found lead me to do more inspection in that area or different
Corrosion’s a good example. The vehicle, like I said earlier,
is not really fatigue-driven. We haven’t had a lot of history
with cracking problems in the primary structure of the Orbiter. There
are a few isolated cases where either we didn’t anticipate the
load properly or we had a manufacturing issue where we introduced
a lot of residual stress by essentially bolting it together and putting
a bunch of residual stress in it that wasn’t intended to be
there, so we’ve had a few issues. I probably could count on
one hand in the primary airframe number of cracking issues we’ve
Get into the subsystems and you have a lot of failure issues just
with aging and things. It’s a different story. But in the primary
airframe, just not very many instances. But as the vehicle aged, we
did have a lot of corrosion problems. Trying to be proactive and maintain
that airframe over time and not let that corrosion problem get out
of hand was a constant, constant challenge. And getting the programs
to be thinking about the airframe as an aging airframe. You can’t
just blow off the inspections. You need to go do them. We need to
clean it up. Showing them was a big part. Showing them and educating
them as to what the ramifications of it would be if we didn’t
correct it and why we needed to keep doing it on the other vehicles,
and what kinds of more aggressive things we need to put into place
to protect the airframes over the long haul.
No reason we couldn’t use the airframes, essentially, indefinitely
if we took care of them, but we really needed to stay on top of the
corrosion problems. Particularly, 102 probably had the worst, for
various reasons, but they all had corrosion problems as they aged.
The vehicles essentially are airplane airframes. They’re an
aluminum alloy. They sit on the beach a significant portion of their
time, particularly cumulatively when you talk about over their lifetime.
We started the program with no proactive corrosion control measures.
We built them. We designed them not to corrode. I can actually remember
when I first started, the guys at Rockwell M&P [Materials &
Processes], god bless them, telling me, “Well, we built it not
to corrode, so it’s not going to corrode.” And the fact
that it did so well is, in fact, really a testament to the great work
they did, in terms of being very proactive in choosing alloys that
wouldn’t corrode as easily. Choosing alloys that were corrosion-resistant,
choosing alloys that, when they did corrode, they corroded gracefully.
You’ve seen your car when it starts to corrode, it starts to
exfoliate and basically the layers of the metal separate, and it looks
like big sheets of sloughing metal off of an old car.
You don’t want your airframe to do that. There are certain alloys
that are very susceptible to that, and there are certain alloys that,
when they corrode, they tend to pit, which is less detrimental from
an airframe-integrity perspective. You don’t want alloys that
are going to stress corrosion crack, which is essentially a corrosion
phenomena that propagates very rapidly and can result in catastrophic
NASA had done a lot of work on characterizing metals, particularly
[NASA] Marshall [Space Flight Center, Huntsville, Alabama], had done
a lot of work on characterizing what kinds of metals we should be
using. Rockwell leveraged all that research. “Let’s use
good alloys that don’t corrode, that don’t fail catastrophically
when they do corrode, so that we’ll have a chance to see it.
We’ll use state-of-the-art corrosion protectants, primers, paints.”
And that was great, and it did a great job for probably the first
five to ten years with the vehicles. But after that, the primers and
paints start to break down just like they do in your house, and no
proactive measures really in place to deal with that. Over the years
that I was with the Orbiter we did a lot of repainting, resurfacing
the rudder speed brakes, the wing leading edges. We stripped them,
just like you would strip an airplane, stripped them, repainted them,
re-primed them, tried to separate galvanic couples.
Rockwell had rules. “Here’s your rules. You don’t
ever put this metal with this metal, because the chemistry between
them is so strong that if you put them in a saltwater environment,
they will cause corrosion." So they had very strict rules about
how they use the metals, and that helped, but, unfortunately, when
you’re building an aircraft that goes to space and you have
to thermally isolate the exterior of the vehicle from the interior
of the vehicle, you use alloys like inconel. Inconel is very aggressive
against aluminum. You use things like gold in your multi-layer insulation
to protect it thermally from heat being transferred, managing the
heat being transferred in and out of the crew module. Gold is extremely
aggressive to aluminum and will actually go right through the primer.
It will pit right through the primer.
So no matter how careful they were, there were still cases where we
didn’t quite get it right, and after five years, ten years,
it started to show in the airframe areas. Particularly it started
basically in the wing leading areas and the rudder areas that were
exposed externally to the vehicle, not purged, basically directly
exposed to saltwater environment or humid air. That’s where
it first started to show. The body flap was another area where it
first started to show. Then you started to see it inside the vehicle.
You’d see it inside the payload bay. You’d see it on the
582 frame, the big frame that separates the forward fuselage from
the midfuselage. We had a lot of corrosion problems on that, lots
of corrosion problems on the forward fuselage of particularly 102
because of the gold multi-layer insulation blankets. Luckily, the
later vehicles, they used aluminum, which actually has a weight and
cost savings. It worked well enough. You didn’t need the gold
from a performance perspective, so they switched to aluminum, which
is actually great. It’s good for the airframe.
We dealt with a lot of that kind of stuff, and so for me, it worked
out. It worked out really well for me. I came out of my undergraduate
with an aerospace degree, emphasis in structures and dynamics and
went into the Structures and Mechanics Division and did all kinds
of stuff, thermal and failure analysis and things of that nature,
but settled in structures.
I went back in ’94 and got my graduate degree. It was kind of
a crossover between materials and mechanical. Basically went on NASA
fellowship on the justification that we have aging Orbiters. We don’t
have anybody here that knows how to deal with an aging Orbiter. We
need to get more cognizant of what’s going on out in industry.
We need to get more cognizant of what’s going on out in academia.
We need the connections, we need the knowledge, and we need to bring
it from the outside back in. And they said, “Yes, that sounds
like a great idea.” So they sent me to graduate school, and
I was able to bring a lot of that back with me, both academically
and then contacts both in academia, and ironically, my time in academia
connected me to Langley, that I brought back to JSC.
And they [LaRC would] say, “Well, we talked to JSC,” and
I say, “Well, no, you don’t, not as far as I’m concerned
you don’t. I’m the person that’s in charge of structural
integrity of this airframe. I don’t know any of you people,
and I need your help.” I was able to forge relationships with
some of the engineers at Langley that I have to this day that when
I have materials problems, I call them. And that wound up putting
me in really great stead through Columbia and through my time at the
NESC and on to Orion. Had I stayed parochial to JSC, I [would have]
just had no exposure to them, none. It was through going out into
academia that I actually made these contacts at Langley that in the
end I plowed back into my team at JSC to get their expertise to apply
to this aging aircraft thing that JSC had no clue, but these guys
did every day. Langley works every day with aircraft, people doing
aging aircraft stuff. We were able to really leverage that expertise
out of Langley to help us. So that was pretty cool.
That’s funny, going to Utah to make that connection in Virginia.
Yes, it was really interesting.
While you were talking, it just jarred another question I was curious
about. You mentioned having to convince folks of the real need to
maintain the airframe. While you were in that position, we had an
Administrator whose mantra was “faster, better, cheaper.”
Yes, we did, as a matter of fact.
Did that have any sort of impact on what you were trying to achieve
or the improvements you were trying to make to the Orbiter?
Let me tell you a funny story. I started in ’90 full-time. I
started in ’90, maybe early ’91, and sometime in that
year or possibly in the next year, it couldn’t have been too
much later than that, [Daniel S.] Goldin was down, and he wanted to
have a meeting, a lunch meeting, with all the new interns.
By this point, I’d already acquired a little bit of a reputation
because I worked with this guy Stan, and he was pretty outspoken.
He and I got along really well, and so I was pretty outspoken. So
I already had gotten a reputation a little bit. It’s great.
I was raised by a great group of guys in ES, these old-school Apollo
guys that were like, “You can say anything you want, you just
better be right, and we will back you up, but you’ve got to
be right or you’ve got to use your authority responsibly.”
Okay, great. So I had a little bit of this reputation, they were kind
of raising me this way.
They said, “Well, we want these interns.” Somehow or other,
I got selected for this list of interns, and my division chief just
went, “Oh, my god.”
Okay. So they send me off to this luncheon, and they said, “Please
“Okay, I’ll be good.”
I go, and I mean I literally go into the luncheon, I pick up my lunch,
and I sit at the table and I’m eating my lunch. I’m being
very good. I’m not saying a word. I’m just minding my
own business, and I kid you not, out of this room of thirty interns,
he comes up to me and he says, “Who are you?” Dan Goldin.
“Who are you?”
“Oh, I’m Julie Kramer. I’m a new intern. I work
in Structures, and I’m a structural subsystem manager in training
for the Orbiter.”
“Oh, so you’re the one who tells us if we fly.”
At this point, he’d already developed this sensitivity. “So
you’re the one who tells me if we fly ten flights or a hundred.”
“Well, what do you think?”
“Well, I think you’re not going to fly a hundred if you
don’t get your corrosion problems under control.” Oh,
He [Goldin] goes down[stairs] after lunch. This [lunch] is on the
ninth floor, [the Center Director’s suite is located on the
9th floor of Building 1 at JSC], and he goes down to the sixth floor
[the Orbiter Project Office is located on 6th floor of Building 1
at JSC]. Dan [M.] Germany is the project manager at the time. I’d
already gotten in trouble once, so I pick up the phone and I call
my counterparts down on the sixth floor, and I said, “Oh, my
god. Tell Dan he’s coming.”
Well, they didn’t get to him soon enough, so he shows up in
Dan’s office and he says, “Tell me about this corrosion
problem you have on the Orbiter.” Oh, it was bad. It was just
ugly. Luckily, Henry [O.] Pohl was the director of Engineering at
the time, and he came to my rescue and basically said, “Well,
she’s right. We could have talked about a more tactful way we
could have had this conversation, but she’s right, and we really
need to deal with this problem.”
Actually, once we produced the objective evidence and showed them
and could talk to them about what we could do, it’s not to say
we got to do everything we ever wanted to do. There were a lot of
controversies about do we strip off the TPS and look for corrosion,
because we were seeing signs that we might be having corrosion underneath
the TPS. It had to be stripped. That’s just hours and hours
and millions of dollars worth of work. But we were able to work with
the program. They had some objectives they wanted to achieve in terms
of weight savings and changing out TPS systems, and so we said, “Great.
While you’re doing that, we want to do inspections. We want
to be careful about how we’re stripping the TPS off. We want
to look at the surfaces, make sure the primers are holding up okay.
We want to do sampling.”
We tried to work with them to be cost and schedule conscious, but
yet still get those objectives, and, of course, that’s always
a big part of subsystem management is trying to be responsive as an
engineering community to the project’s cost and schedule constraints
because those are very real. I can’t just say, “Hey, take
the Orbiter down for a couple years and strip all the TPS off of it
so I can go look and see if it’s corroding.” Even if I
really believed that’s in their best interest, that’s
not practical. So, trying to find a compromise through there where
you can get the technical data you need to protect them from something
that could happen and yet not cripple them from a cost and schedule
perspective is always going to be a challenge.
And we didn’t always win our arguments. There were plenty of
things I wanted to do that we didn’t do, but in the end, we
either found ways to sample or found other ways to gather data to
try to work that, and it worked pretty well, I think. From that perspective,
we were able to find a lot, deal with it proactively, give them some
changes they could make to help mitigate damage in the future and
able to refinish specific areas and do things to help maintain the
airframes. I think, overall, the airframes, at least when I left,
were in pretty good shape, and I’m sure that, yes, the guys
would say they’re still in pretty good shape.
I would say, for the most part, I always felt like the project was
pretty supportive, although they expected me to be reasonable and
responsible in exercising that authority to try to say, “Look,
hey, we need to do something.” Usually by the time we found
corrosion in the airframe, there was never a question of resources.
I always got the resources I needed from Rockwell or Boeing, always
got the support I needed. It was never an issue. “Hey, you’ve
got corrosion in your airframe,” and they would say, “Okay,
fine, go fix it. I don’t care what it costs. I don’t care.”
They might say, “Hurry.” They might say, “Hurry,
hurry up, because I really want to get out of OMDP. I need to get
out of OMDP. I need to keep my manifest.” But they recognized
that for me to do that, I needed the resources, and they were always
very good about providing the resources we needed to fix the airframe.
They understood that was in their best interest.
Was corrosion considered a Criticality 1 of the vehicle?
Well, the airframe is all considered, by default, 1:1, Crit 1:1, which
is a little bit of an untrue-ism. Back in the day, nobody wanted to
pay for a FMEACIL [Failure Mode Effects Analysis and Critical Item
List] of the airframe, which would have required that you go through
and look at each component and say, “Well, if this happened
to it, would it fail? If that happened to it, would it fail?”
So basically, the whole airframe, primary structure, was classified
as 1:1, which generated its own set of problems. Basically, any defect
in the primary structure was considered a Crit 1:1 defect, even though
it might just be a scratch or an elongated hole. Over time we had
to learn to deal with that and classify certain types of defects as
what we used to call fair wear and tear. “It’s okay if
you scratch it. Just go touch it up. But if the scratch is deeper
than this, go call a stress analyst. Or if the corrosion is like this,
here’s a standard repair. Just go do this.”
Those are all very standard things in the aircraft industry, but we
had no clue how to do it. We had no clue how to build a standard repair
manual. We had no clue how to give the Cape or Palmdale damage limits
that they could operate under with delegated authority so they didn’t
even have to call us. Kind of like the stuff you’d have in an
aircraft depot. Send my aircraft off to maintenance, and they just
clean it up. They don’t call Boeing, as long as it’s within
the limits of the standard repair manual.
We had to learn how to do all of that and document it, and it was
always a trade. How close are we to the end of the Shuttle Program,
such that how much money do I want to expend into these standard repair
manuals? It was always balance, and so we built them over time. We
said, well, if we found we were doing a repair frequently enough,
we would spend the money to have Rockwell go develop, or Boeing at
the time, go develop a standard repair. And the Cape gradually over
time built up a standard repair manual that they could use and Palmdale
would use it. They gave them some delegated authority because of this
Crit 1:1 aspect of the airframe, because all the airframe was Crit
You’ve talked a lot about the different inspections that were
used on the Orbiter. Do you want to talk about your time at Palmdale
during the OMDPs on some of the vehicles themselves, and the major
Sure. I’m pretty proud of the fact I’ve supported all
the OMDPs at Palmdale, every vehicle that was there for OMDP, and
then eventually assisted in transition of OMDP to Kennedy when they
shut Palmdale down. So, yes, I supported all the OMDPs out there,
initially starting as a structures guy, obviously working with them
on all the structural inspections. And like I said, I used to spend
quite a bit of time out there. I’d basically go out on three-week
rotations, and I’d spend three weeks out at Palmdale and then
I’d spend three weeks at home. Then I’d spend three weeks
out at Palmdale and three weeks at home. The other systems managers
would go as well, and we would stagger it based on what critical milestones
were going on when, but we always had a structures person out at Palmdale
working with the contractor.
Then later on when I was resident for 105, obviously my specialization
was structures, so I tended to get drawn into structures issues, but
the job was broader than that. It was more of a systems engineering
role, working multiple systems, so I did get more exposure to other
systems and to modifications that were going into the vehicles. Obviously,
you tend to gravitate to what you know, and the other part of it is
people tend to come to you because they know that’s your area
of specialization. So we tended to divide the issues up that way.
For 105 I was there for the full year during its OMDP, and just spent
a lot of time doing everything from writing paper with the technicians
and with Rockwell [Engineering]. We were pretty badgeless in that
regard. Nobody really cared whether the guy that wrote the paperwork
was an engineer at NASA or an engineer at Rockwell. All the technician
cared about was did he get his paperwork with the signatures on it
so he could do his work. So we pretty frequently did that work interchangeably.
I can remember being there second shift writing paperwork to pull
bolts out of the tail, with the guy from Rockwell sitting there saying,
“Well, shall we tell them to do this?”
“Yes, I think we should tell them to.” It’s all
stuff we’d never done before, so we didn’t know how to
write work instructions, other than basically we knew what we wanted
them to do. So we would write them, and the technicians would do it,
and then they’d go, “Oh, that’s not working right,”
and then you’d get to go back and rewrite them and redline them
and work with Rockwell on getting everything sorted out and cleaned
up and signed.
But it was good. It was a lot of fun, spent a lot of time on the floor,
and, of course, the technicians taught you a ton. The technicians
and the other Rockwell engineers just taught you so much about how
the airframe was put together and why, which for somebody like me,
who wasn’t here during the design phase, was invaluable. You
can only learn so much from looking at books and being told. You really
need to get out there with the guys that built it and designed it,
and that’s how we learned, was getting out there, and in the
process of modifying and inspecting, we learned a lot about how and
why it was designed the way it was designed.
Were you also working on the wings and the aft fuselage and the other
Were those items taken off, and then the Orbiter inspected?
No. During OMDP they would remove the hypergol systems. They would
remove the OMS [Orbital Maneuvering System] pods in the FRCS [Forward
Reaction Control System] and they’d leave those at KSC because
they were uniquely qualified to handle the hot systems. Obviously,
Rockwell built them, but once they were contaminated, if you will,
with the hypergols, we handled them only at KSC, for the most part.
Obviously we had to deal with it if we landed it at [NASA] Dryden
[Flight Research Center, Edwards, California], but we didn’t
take those into Palmdale. So they were removed and farings were put
on, fake OMS pods and a fake FRCS to keep the OML [Orbiter Mold Line],
to keep the mold line appropriate for flying it on the back of the
We’d fly it out to Palmdale and they’d remove those elements,
and they would either open the payload bay doors or remove the payload
bay doors. They would open all the access doors. There’s a lot
of access doors in the side of the fuselage to get you into the wings.
They’d open up all the cove seals, the body flap, anything that’s
hinged. The body flap, the elevons, the rudder speed brake, they all
have special seals around those hinges so that the flaps can articulate
but they don’t let hot air in. It’s not like an airplane
where on entry, when you look out the window, you see they put the
flap down and you have a huge exposed cove. You can see the hinges
and the rods that drive the elevons. You can’t do that, obviously,
on a Shuttle reentry because it’s hot.
All that stuff is closed coves with sliding seals. They had to open
all that stuff up, take it all apart to expose all the actuators and
rods and primary structure. The vehicle is just chock-a-block full
of multilayer insulation [MLI] blankets in the payload bay. They’d
remove all that. They’d remove all those blankets off the doors.
When the doors open on orbit, you can see that the radiators are there.
They’d remove the radiators. They’d remove all the MLI
blankets, everything that covered the primary structure. As much as
they could reasonably remove, they would remove.
Then the guys would come in and start doing structural inspection.
They’d go into the wings. They would use boroscopes and go into
the more confined spaces into the control surfaces themselves. They
all have lightning holes on the front spars. They’re just basically
a box construction. It’s got an upper surface, a lower surface,
and a trailing edge, and it’s got a panel that closes off the
front of it. But it’s got lightning holes in it, and so they’d
go in. If it was big enough in the body flap, you could almost stick
your head in there and look, or you’d use a boroscope, a long
snake-like visual inspection thing, and you’d go in with a boroscope
and you’d look. You’d go literally fastener by fastener.
If the inspection was I want to make sure that this line of fasteners,
that there’s no corrosion, there’s no cracking, the inspector
would go, literally, “I have fifty-two rivets on this line I
need to inspect,” and they’d count them as they went.
“Okay, I inspected all fifty-two and they’re all good,”
or, “I inspected all fifty-two and rivet number twenty-seven
has a suspect finding on it.” An engineer would come in, they’d
review the tape. “Oh, no, that’s good. That’s not
corrosion. That’s gypsum we ingested when we landed at White
Sands [New Mexico, on STS-3].” [OV-]102 was full of it, full
of it, and it was a big problem because the technicians would see
it, it would be adhered to, or fused to the paint from having sat
in there for so long, or water. It would sit out and get rained on,
and water would run down. It would take all that gypsum and run it
into the lower part of the vehicle, which is where all the corrosion
would be and is. So they’d see the white gypsum and they’d
go, “Oh, my god, it’s corrosion.”
You’d have to go out and you’d have to try to clean it
out or try to take a sample. You could use a boroscope and get a sample,
and you went, “Nope, it’s gypsum. It’s gypsum. It’s
all right.” So we would do that kind of stuff.
We would do that in the lower forward fuselages because you can’t
get in there. You go in through antenna holes. They’d remove
antennas from the lower part of the vehicle. There’s several
access doors on the lower fuselage of the vehicle for antennas. We’d
open the doors, they’d remove the antennas. You could get in
there with boroscopes. You could get in there. Some of the antenna
holes were big enough I could fit in there up to my waist. So you’d
stick your head in, into that area, the volume between the actual
pressurized crew module and the forward fuselage.
The crew module is basically hung inside the fuselage to separate
it so it doesn’t have [body bending ] loading. All the body
loading, the landing loads, all the bending [portion of that load]
all that goes through the airframe through the forward fuselage. It’s
not passed on to the crew module. The crew module is hung on a series
of swing links, so you think about it like a basket that’s hung
inside the forward fuselage. As the fuselage bends and warps on orbit
due to thermal, or bends due to landing loads, it’s hanging
in there and it doesn’t pass any of that load onto the crew
module. It’s sized basically for pressurization loads. But the
volume in between the two, at the back part by the 582 frame, you
could actually stick your head up in there. A person could fit in
there, and so you would go up in there and you’d inspect from
there. Or you’d get in the wing and you’d go through the
There were holes in each spar, mostly for manufacturing purposes,
so there were big doorways in each of the spars that would be closed
off. On 102 they were closed off. I think on the subsequent Orbiters
they were strengthened so that they didn’t have to have the
doors there, if I recall correctly. [Or maybe that was 101 that had
the doors] You’d go through these doors, you’d remove
the door, go through the cutout, and you’d go all the way to
the back of the wing. Pretty soon, you’re on your knees. Well,
you could stand. In the mid part of the wing when you went in, you
could stand, almost. You’d be crouched over, but you could stand.
Then you’d go to the next bay back and you’d be on your
knees, and you’d go to the next bay back and you’d been
on your stomach, so pretty soon you’re kind of inching your
way all the way back to the back of the wing on your belly and hope
that the oxygen detector didn’t go off because it was heck to
get out of that thing once you were all the way in the back, because
the inside of the wing is filled with these tubes. It’s what
gives it structural integrity.
It’s got skins on the outside, and it’s got these frames
that are made with tubes to keep it lightweight. They were so fragile.
They were great in tension and compression, which is what they’re
designed for, but you’d be in there banging around doing maintenance,
because it was never designed for maintenance. So you’d be in
there banging around doing maintenance, and god forbid, you’d
bump a tube because you could actually literally almost squeeze it
like a Coke can. It’s not quite as fragile as a Coke can, but
you could have. If you grabbed it, you could dent it, and then it
had to be replaced. Then, your boss would say, “Don’t
ever come home with your name on the problem report saying that you
tripped on something or you broke it or you whatever.” You did
not want that. You did not want to be issuing the problem report that
said you were the one that caused the problem by stepping on something
or busting it.
It just wasn’t designed right for people tromping in and out
of the wing like that, but that’s how we did the inspections.
We just would go in and go as far as you could, because a visual inspection
was always better than any kind of nondestructive evaluation or boroscope
inspection. If you could see it with your own two eyes, that was the
best. So we’d send the guys in as far as they could get into
the airframe, and then we’d go to boroscopes, and then if we
couldn’t do it any other way, we’d go to ultrasonic or
eddy current inspections where you were relying on nondestructive
evaluation of structure that was hidden or sandwiched in, you couldn’t
We never took the wings off. Putting the wings on, boy, we actually
talked about that and various options of relocating Orbiters and thought,
“Oh god, that would be a disaster.” Getting them off and
then getting them back on again would be just a nightmare. So, no
taking the wings off. The wings were basically through permanent installation,
as was the tail, although I think push came to shove, we could have
gotten the tail off. It wouldn’t have been that big of a deal.
But the wings were never designed really to come off, and all the
GSE [Ground Servicing Equipment] now is long since gone stored for
years in Bell [Aerosystems Company] warehouse and at Downey, so really
not feasible to get the wings off.
But we’d take the payload bay doors off for various modifications
and/or they were composite, one of the few composite pieces of primary
structure on the Orbiter. I talked earlier about the TPS weight savers,
where they went from heavier blankets to lighter blankets, a lot of
cases like a AFRSI [advanced flexible reusable surface insulation]
to FRSI [flexible flexible reusable surface insulation], because they
didn’t need it from a temperature perspective, and the FRSI
was lighter. They’d have to strip all that stuff off like peeling
paint at your house. We eventually found some non-aggressive, environmentally-friendly
stripping agents that would help the guys, but literally it was pull
it back, strip it with a nonmetallic putty knife, pull it back, strip
it with a nonmetallic putty knife. We had technicians that were getting
carpal tunnel syndrome and all kinds of problems from stripping this
stuff off of the payload bay doors and wings.
You had to be careful, because if they pushed in or they tried to
use a metallic scraper, the skins were so thin, they would just go
right through the skin, and then we’d have to go do a structural
repair. It was a constant battle with the technicians. They wanted
to get it off fast and not hurt themselves in the process, and we
wanted them to take it off slow enough that they weren’t damaging
the substructure. So it was always a challenge to do that.
You mentioned that you worked the second shift. Were they working
In a lot of cases they were, particularly when we were there for structural
inspection. Inspections would occur. What they would typically do
is they would run visual inspection and boroscope operators on first
shift, and then they’d run x-ray inspections on second and third,
or anything that needed an area clear, which would be when we were
taking bolts out of the tail, they didn’t want people working
around the vehicle in case something happened. They would have a clear
in the back of the vehicles so we’d do stuff like that on second
and third shift, because if there’s any danger to the other
technicians or engineers, they wanted to keep them away, x-rays obviously
being a big one. You didn’t want to be irradiating all the engineers.
We would just do that on second and third, and you had to be authorized
to be in there and stay in certain parts of the high bay. That’s
the way they kept us engineers from running around getting irradiated
was they did it on third shift.
How many people do you think were working on these OMDPs, do you recall?
Well, from a NASA perspective or everybody? I couldn’t even
tell you how Rockwell was staffing those kinds of things. Hundreds
of people supporting OMDP at any given time when it was up there,
between the inspections and the modifications and all the systems
work that was being done, hundreds of people from Rockwell and the
Dedicated NASA people there, less than a dozen on any given OMDP.
They had a resident office out there. Orbiter Project Office had MV
[mail code for Orbiter Vehicle]-8, I believe, MV-8 or MV-6, I can’t
remember which. I think it was MV-8 that was there in the Palmdale
facility. When we weren’t in a major mod, it was pretty much
a skeleton crew. It was an office manager and a couple guys that watched
ET [External Tank] umbilical production, and some of the other manufacturing
that went on in support of the Shuttle Program in that facility.
When the Orbiter rolled in, we generally would bring some of our NASA
guys up from the resident office at Downey, and we would bring in
subject-matter experts from JSC or KSC to support whatever the needs
were of the NASA resident office at Palmdale. So that’s frequently
how I would get there. I’d be there just in support of that
Then you spent some time out at KSC, but you were the chief engineer,
is that correct?
Chief engineer for OMDP as it moved from Palmdale to KSC, so that
was just broader scope. Similar scope but broader. It wasn’t
just inspections; it was modifications and whatever the project wanted
done at that mod period. Then particularly in the case of transition
from Palmdale to KSC, dealing with any issues with skills and/or making
sure KSC would be able to do the same work logistically that we could
do at Palmdale, because the Palmdale facility was built specifically
to build Orbiters. It’s just a very different footprint, very
different capability than what’s in the OPF at KSC. We made
sure we had all the capability we needed in the OPF to do all the
same kind of work, as well as all the critical skills from an inspection
perspective to run all those special inspections, x-rays, and the
ultrasonic eddy current. You have to have certified operators of the
Kennedy had those, but they also were running around between multiple
OPFs trying to support multiple Orbiters in the flow, and so you bring
a vehicle there on OMDP and the demand on that service is a lot higher.
So, trying to make sure we had the right support to do all that until
we transitioned it over.
Those were done in the OPF at KSC?
Did you ever have any concerns about the payload bay and future payloads
that were going to go in, like the Hubble Space Telescope?
I’m sure there were people that worried about the cleanliness,
but the point that we were doing structural inspections, it was a
shirt-sleeve environment, so that was done by design. You’d
come in and you’d strip everything out of the vehicle. Guys
would come in in shirt sleeves and basically do all the inspections.
Then as you backed out of areas, you would clean them. As you finished
everything in the wing, the inspectors would go in and make sure it
was visually clean, and then you’d seal it. It would be like,
“Okay, this is done. We’re not going back in the wing.”
And we’d put the door on. You can’t take the door off
without getting into a whole series of back-out inspections and photographs
again. They would just basically go in, do all the structural inspections
and repairs that needed to be done, and as the paperwork was done,
they would close areas out and just back their way out of the vehicle.
It was funny that you asked that question, because I can remember
the first time I went to Palmdale and there’s owls in the building.
And the payload bay doors are open, and there’s a dead rat carcass
in the payload bay. That’s probably not very nice for the historical
Well, you can edit.
But it’s true. It’s true. It’s just people think
that everything is this bunny-suit environment, and, sure, inside
the crew module it is, because it’s really hard to clean and
you can get skin and hair and dirt and things inside the crew module
that are really hard to get out. So the crew module was maintained
as a clean-room environment and obviously the whole vehicle, they
kept very strict FOD [Foreign Object Debris] and tool control. I don’t
want you leaving a wrench. I don’t want you leaving a badge.
I don’t want you dropping cigarettes, which we have done in
the early days of Orbiter before they did a lot of those kinds of
controls. Some guy who leans over the payload bay and dumps a pack
of cigarettes, and he doesn’t know how many cigarettes he had
in the pack because it was open. And they’d be like, “Oh,
you know what? We really need to be better about FOD control.”
A lot of those things are evolutionary in nature. We learned about
keeping control of those kinds of things.
Palmdale was a manufacturing facility. It was built in there, and
it wasn’t clean room. When we did mods and we did inspections,
it was not a clean room. As we would back out of areas, we would eventually
tent the payload bay. When they were done and they were backing out
and they were reinstalling all the MLI and they were reinstalling,
closing out wiring trays and things like that, then they would go
to a clean room. They would specifically tent it and vent it positive
pressure to keep junk out of it and clean it really good and then
Most of the time when I dealt with it, it was a shirt-sleeve environment,
which actually was a treat. Then you go to KSC and they tell you you
have to put a bunny suit on, you’re like, “Eh, I’ve
been in the wings. I don’t need a bunny suit.” So it’s
much stricter, much stricter at KSC.
Now, when they went into the OMDP mode, they went more like Palmdale
did when they were in there doing inspections, still very strict FOD
control from a tool perspective, but basically would clean it as they
would go, and shut areas out once they were sure they were clean and
had been inspected. So, no, we weren’t very clean, surprisingly.
One of the questions that I did want to ask you, you came in the early
1990s as an engineer. Were you one of the few female engineers working
in the Shuttle Program, or were there a lot of other female engineers
that were working with you at that time? Was it unusual for you to
be on the floor and working in management?
I was one of the few at Palmdale. Actually, ES, too, which was the
organization I went into, was pretty integrated, if you want to say
that. I had several contemporary women that I worked with, so in an
org [organization] of probably a couple dozen people, there were probably
half a dozen of us at a similar grade. Now, there weren’t really
any women above us. I’m trying to think if there were any women
above us. I don’t really think there were. In division management,
all my supervisors were always men, and all the way up into the project
always men, probably until, I think, late in Orbiter, probably right
around Columbia, I think, Ralph [R. Roe] had a gal deputy. Trish [Patricia]
Petete, I think, was his deputy at that time, and probably was the
first female customer or supervisory-type person I had. I hope I haven’t
Now, Rockwell was a little better. You interviewed Frances [A.] Ferris.
Certainly she was somebody I dealt with as an Orbiter project manager,
customer and on the contractor side. So I think Rockwell was probably
a little more integrated than we were.
I will honestly say from my perspective I never felt like there was
any issue. I was treated just the same as everybody else and given
every opportunity, so I always felt really lucky. As a co-op, a lot
of my female schoolmates at Purdue [University, West Lafayette, Indiana],
which is fairly integrated as an engineering school goes, would talk
about their experiences out in industry, either in petroleum or in
industrial engineering. I had a girlfriend that worked at Kodak and
a girlfriend that worked at IBM. They would come back and go, “Oh,
And I’d say, “Well, I got to do this.”
They’d be like, “Oh, my god, I got called Honey.”
And if I got called Honey, it was because the guy was sixty, and I
was twenty, and it wasn’t one of those things. It was he looked
at me and he thought of his daughter, and that was fine. There was
never an issue with, “Well, you can’t do that because
you’re a girl. We’re not going to let you.” You
never ever felt that. I had absolutely awesome bosses and absolutely
awesome mentors who gave me just a ton of opportunity, just as much
as you could handle, which was great. Even the project managers, I
never felt like there was ever any issue in that regard.
I didn’t have any real women mentors, certainly. All my mentors
were all these all old crusty Apollo guys, and so maybe that probably
warped me a little bit, but they were great. I never really felt like
there was an issue. I do still feel a little uncomfortable when I’m
in a room with just women, because it’s odd still, even still.
Orion, now, as the next-generation aerospace program, is extremely
integrated. Lots of women. Lots of women in critical engineering leadership
positions, and, of course, now Engineering is very integrated, women
division chiefs and women office managers, women on division staff
and Center staff, so you see a lot more of that now, but certainly
not when I started.
I was just curious about that. I wanted to shift gears, but I wondered
if you wanted to look at your notes. I think you pretty much covered
all the questions that I had.
I tried to.
I just thought we’d talk a little bit about Columbia and your
work with the NESC.
Oh, okay, yes.
But you were very thorough.
Yes, I think I got most of what I refreshed my memory on.
So tell us about your work following the Columbia accident. You were
on the Hardware Forensic Analysis Team.
That was one of those right place at the right time, unfortunately,
lest you mistake my enthusiasm. Obviously, Columbia was a terrible,
terrible thing, but from my career perspective, you couldn’t
have picked a more perfect storm of an incident that occurred. I was
working in vehicle engineering at the time, so I was matrixed to the
project. I was right there in their direct engineering staff. My expertise
was in the wing, which was what they wanted, people that had expertise
in the wing. My academic background was in aging aircraft problems,
fatigue, corrosion, metallurgy.
When the accident happened, I was actually TDY [temporary duty]. I
flew back and went straight to Nacogdoches [Texas], because they wanted
people in the field that could tell the difference between a tail
and a wing, because it’s not an obvious thing to somebody that
[doesn’t] know the airframe. It’s actually hard to tell
the difference. There are subtle differences between the way Grumman
built and Fairchild built, and so they were looking for people that
could actually look at a piece of hardware and from very just subtle
characteristics could tell is that a wing, is that a tail, is that
a body flap, is it a different part of the primary structure.
I went with a gentleman from KSC NASA and a gentleman from Boeing,
and we literally, the three of us, rode in a car from debris-collection
site to debris-collection site. That’s what we did all day long.
We would go on this circuit between the debris-collection sites at
Nacogdoches and Hemphill [Texas] and that area, and we would say,
“That’s wing. That’s tail. That’s something
different. That’s right wing. That’s left wing.”
So that’s an item of interest, and it would be red-tagged and
it would be triaged and sent to Barksdale [Air Force Base, Louisiana]
and on to KSC. So that’s what we did.
Did that for a couple weeks, until the majority of the big debris
was picked up, and then I traveled to Kennedy. At that point I went,
really, merely just as somebody who had expertise in the wing as a
previous subsystem manager. I joined the current subsystem manager.
I joined the Boeing and the NASA KSC guys that were doing that work,
just as a part of that product team. I knew all the guys and I could
read the drawings. So we went and I helped with that.
I was there for a few weeks, and then I came back to JSC, and I was
sitting in Ralph’s MMT [Mission Management Team] meeting, whatever
they were called, the NASA side of that failure investigation. I was
sitting in the meeting, and somebody was telling me why the wing had
disintegrated. I’m sitting in the back of the room, and I said,
“Well, I have a problem with that. That doesn’t work for
me, because I just spent two weeks at Kennedy, and that piece of debris
is on the floor at Kennedy. Next?”
The next person comes up and they tell a story. I said, “Well,
I have a problem with that, because I just came from Kennedy and that
debris item is item number 6,” blah, blah, blah, pull out my
little notebook, “See, and it’s right here. It’s
on the grid at Kennedy.”
About the third time or fourth time that happened, Ralph turns to
me and he says, “I think we have a communication problem between
Kennedy and JSC.”
They were doing the right thing, they were trying to keep the guys
at Kennedy isolated. They were trying to bring the debris to Kennedy,
and the NTSB [National Transportation Safety Board] had very strongly
encouraged us not to poison the investigation at Kennedy. “Don’t
tell them what you think happened. Let them go through the debris
and figure out, based on the debris, what happened,” which is
the right premise from aircraft investigation perspective.
But the problem was it’s a big agency, and so everybody at JSC
is sitting here and they’re just churning, [demonstrates], particularly
once the OEX [Orbiter Experiments] data recorder showed up and they
had all this data right there. You could run all these scenarios and
all these possibilities, and everybody’s evaluating video, and
we have all these great ideas about what could have happened.
I got together with Ralph and talked about it and said, “Look.
You need a gateway, a one-way valve, between Kennedy and you that’s
telling you what’s on the floor to help you moderate what you’re
doing, because you’re wasting a lot of time. We got a lot of
people that really want to do good stuff, but they have no visibility
into what’s going on at Kennedy. I can’t have them traveling
to Kennedy en masse, because they’ll absolutely poison the investigation
as well as getting in the way. So how do we set up a construct that
will allow us to flow information from Kennedy to here to help us
focus the investigation?”
They said, “That sounds like a great idea. Why don’t you
go to Kennedy?”
So I went to Kennedy and stayed there for about five months, basically
working with the reconstruction guys, which was done under Steve [Stephen
J.] Altemus at the time. He was at Kennedy. So his job was to get
the airframe on the floor. My job eventually was as a communication
channel. As they were putting debris out on the floor, it was a communication
What it eventually evolved into was leading the Failure Analysis Team.
We need to go do this failure analysis. We need to go find this part,
cut it up, do this test, whatever. So somebody needed to prioritize
all the failure-analysis work that was being asked for, get the hardware
cut up, get it to the right lab in the agency or outside the agency
that could do the work, make sure we were being responsible with the
hardware, because it was one of a kind. Once you cut it up once, you
couldn’t do it again. You had to make sure it was really the
right thing you wanted to be doing.
So, okay, I need some help. I called this friend of mine from Langley
that I had met back in the day, and I knew he had a long experience
with hardware failure investigation. I called him, and said, “Could
you come help?” So he came to Kennedy, a guy named Bob [Robert
S.] Piascik. He came to Kennedy to help me.
I tried to get some Rockwell guys, old Rockwell guys, so I ended up
with a gentleman named Mike [Michael] Ehret and a gentleman named
Larry [Lawrence] Korb that I knew from way back at Rockwell. Mike
had since retired, and if Larry hadn’t retired, he was on the
verge of retiring, but I’d known him from back in the day and
Mike was the M&P lead, Director of Materials and Processes, during
Orbiter build for many years. So they came to Kennedy and helped me,
and were certainly invaluable. They helped me negotiate that pathway
between CAIB [Columbia Accident Investigation Board] and NASA, because
CAIB wanted to do certain things. They had an agenda. The NASA team
wanted to do certain things. They had an agenda. They weren’t
always the same agenda.
My job was to go figure that out. Negotiate that and figure out and
get the data that NASA wanted and get the data that CAIB wanted, and
if there was a conflict over a piece of debris make a recommendation
and figure out what we were going to do with that piece of debris.
So that’s what I wound up doing.
I had several failure analysts that worked with me and tapped into,
I think, every lab inside the agency and some outside, and did several
hundred, if not thousands, of failure analyses in that five-month
window to support the investigation and the eventual final reconstruction
of what happened. But, again, just a perfect combination; I was one
of the few people that understood the construction of the wing, as
well as could speak to the M&P community because my background
was in M&P as well. I knew what tools they were wanting to use.
I knew what processes they were wanting to use. I knew what was going
to happen to the samples when they went to the lab, and I knew all
the M&P guys. I had a foot in both communities to start with and
then also had a preexisting relationship with the Orbiter Project
Office, so it just worked out to be perfect that I could do that for
them. It was interesting, but definitely it was not how I intended
to use my academic background when I got it, but that wound up being
very invaluable with my experience to go do that for them.
Were you involved at all in the tests out at San Antonio [Texas] with
the foam and the wing?
Not so much so. Not so much so. That was a different group within
the Structures Division here at JSC that supported that effort. We
did more just the actual cutting up and forensic work on the airframe
itself, if they would find something out there. By that time there
was no stopping flow of information. It was everywhere. It was in
the press. But early on, people would have a theory outside, and there
were a few of us that would know what those theories at JSC were,
and we would come into KSC and we would help gather data and try to
capture some of that data without telling everybody at Kennedy, “JSC’s
latest theory is this,” blech, because then they would start
seeing things in the debris that maybe were or weren’t there.
Being able to keep that flow of information going from KSC back to
JSC and be cross-comparing that with what other efforts we’re
finding, like the efforts that were going on at San Antonio, to say,
“Yes, we’re similar attributes.” Or they would show
us physically what they found after they shot the test, and then we
could go to the debris and we could find those same attributes. “Well,
that’s a unique attribute. I’m looking at a whole bunch
of debris, and I’m only seeing that attribute here.” So
you would try to correlate that.
I think we would have gotten there even without the OEX recorder,
but certainly in terms of getting a very crisp understanding of how
things devolved within the wing, the OEX recorder was invaluable.
But certainly when you looked at the debris field, it was very obvious
where the issue had started. We were just able to put those two stories
together and able to provide physical evidence to corroborate what
the OEX recorder was saying.
Then you moved into the NASA Engineering Safety Center. Can you tell
us about that?
So Ralph Roe was previously an Orbiter Project Manager. When he left
that role, they put him at Langley to stand up this NESC thing, and,
of course, I had a preexisting relationship with Ralph. He says, “I
know what your background is. I need a mechanical analysis person.
I know that’s your background. Will you come and do this?”
I was on maternity leave, because I had my daughter the August after
Columbia. I said, “Well, I’m not coming back from maternity
leave early, but if you’ll wait, I’ll come back.”
He says, “Okay.”
So I wound up coming back in, I think in February, a year after Columbia,
into the NESC as their mechanical analysis lead. That was a really
interesting time for me, because it was taking me, for the first time,
out of the mainline project and putting me in an independent capacity.
It was really the first time I’d done any kind of independent
work. Everything I’d ever done up to that point had been very
in-line programs, so it’s a little bit different. You’re
a little more sensitive to cost and schedule [in an in-line role].
You get out there in the independent world, now all of a sudden people
are like, “Wow, you don’t need to be sensitive to cost
I’m like, “Whoa, wait a minute, wait a minute.”
So that’s kind of different. Not that Ralph would ever say that,
but there was just definitely the thought with some folks in this
group that you’re in this independent capacity and you don’t
need to be sensitive to those kinds of things. So it was a real—I
won’t say clashing of culture, because that has a negative implication,
but it was definitely culture shock. NESC is composed of all ten Centers.
That’s one of its things. When you put the research guys in
with the manned spaceflight guys, it’s just a totally different
culture, totally different way of problem solving, totally different
way of looking at risk acceptance, all a good thing. I think the research
center guys really pulled at the manned spaceflight guys and said,
“Come back to your basic engineering roots. Don’t be so
programmatic.” And the manned spaceflight guys went to the research
guys and tried to pull them out of their ivory towers. “That’s
great, but let’s talk about applied engineering.”
So it’s really an interesting balance of how they evolved over
those first couple years of trying to pull the best from both of those
orgs. Engineers that were good in their core engineering skills and
still true to their core engineering skills, but could be aware and
understand what the constraints were from a cost and schedule perspective,
so that you could drive out legitimate technical options that a program
manager didn’t feel like they were just backed into a corner
where you had no option at all. If you gave them an option that was
so politically, schedule, or cost prohibitive, what are they supposed
to do with that? It was a real interesting melding of the cultures,
is probably a better way to think of it. There was some clashing at
first, but melding in the end, of those cultures and trying to pull
the best out of the different Centers.
Great experience, allowed me to reach outside the manned spaceflight
Centers. I’d had a ton of experience with Kennedy, and even
in the latter years of my time in Orbiter had a lot of exposure to
Marshall because of the last several instances we had on Orbiter with
the kind of aging Orbiter things, flow liner, cracking of the flow
liners, BSTRA ball [Strut Tie Rod Assembly] cracking. Those are all
mechanisms inside the main propulsion system that I was involved in
once I moved into systems engineering, Vehicle Engineering Office.
They weren’t necessarily strictly structural issues, but everything
that fails in the end is a structural issue. So I was working with
these guys on the fatigue aspects of the flow liner and the BSTRA
balls, and so a lot of that expertise, that MPS [Main Propulsion System]
expertise, came out of Marshall. I dealt with people like [Robert
J.] Schwinghamer and dealt with all these guys that you’ve interviewed,
Otto [K. Goetz], and all these guys from Marshall.
I had a lot of exposure to that element, but never really any exposure
at all to the other Centers. Once I moved into the NESC, a lot more
work with [NASA] Glenn [Research Center, Cleveland, Ohio]. I’d
had limited exposure at Langley, mostly in that aging aircraft area,
so I worked a lot more with Langley and a lot more with Glenn and
the other Centers. Then, of course, in the capacity I’m in now
as Orion chief engineer, I’d work with all the Centers. Because
we have resources from most of the Centers, all except [NASA] Stennis
[Space Center, Mississippi], and a big contingent of our team is at
Glenn. I was able to build on relationships within the NESC. People
that I knew from NESC are now within positions within the institutions
at the Centers, other chief engineer functions, other institutional
positions, Director of Engineering kind of positions at other Centers.
Now you know these people, so it’s much easier, much easier
[to get work done]. Very different than where I grew up twenty years
ago where it’s much more parochial to JSC. I think it’s
a good thing.
Did you do any work on the Return-to-Flight effort at the NESC?
I did. When I was in the NESC, obviously, a logical segue would have
been Return-to-Flight for me, although part of the NESC, too, was
to try to get some broader exposure than that. Clearly a big emphasis
for NESC was Return-to-Flight. Yes, I did some of the work. Basically
when they were working on the debris risk, the Debris Assessment Team
type work, where they were trying to establish what probability of
impacts from ice and things of that nature were, and foam, I worked
some of that independent assessment. I did some of the independent
analysis on ET ice, which actually was umbilical ice, which was quite
large, so looking at some of the historical data on that and modeling
and doing some statistical analysis, trying to establish if I thought
the program models for debris damage were appropriately predicting
what the risk was. I worked a lot with them on that.
Did you have any contact with the Stafford-Covey [Thomas P. Stafford
and Richard O. Covey Return-to-Flight Task Group]?
Stafford-Covey, no. No, actually, I don’t think I did. I’m
trying to remember what, if any, NESC more formally engaged with them
on. So I don’t know if any products I did ever fed anything
further up. I don’t remember.
I don’t think so.
I think I might have picked your brain [enough], but is there anything
else you would like to talk about about Shuttle?
Do you think there’s anything that we have overlooked that you’re
just dying to tell us?
Just dying to tell you? No, I don’t think so. Like I said, based
on the questioning, the only thing if you’re interested in the
evolution of some of the tools, some of the analysis and things like
that, Glenn Miller, I don’t know if he’s on your list,
but he’d be great person to talk to. He certainly is well aware
of how the evolution of the analysis went, including some of the bigger
analysis challenges in terms of structural thermal mechanical analysis,
tiles in conjunction with the primary structure, probably one of the
bigger more unique elements of how we did analysis on Shuttle. He’s
well aware of all that.
That’d be great. And I thank you for your time today.
Oh, you’re very welcome.
It’s really amazing.
White: Like I said, it was a little trip down memory lane,
which is good.