NASA Johnson Space Center
Oral History Project
Edited Oral History Transcript
Interviewed by Kevin M. Rusnak
Houston, TX – 24 May 2002
Rusnak: Today is May 24, 2002. This interview with Arnie Aldrich is
being conducted in the offices of the Signal Corporation in Houston,
Texas, for the Johnson Space Center Oral History Project. The interviewer
is Kevin Rusnak, assisted by Rebecca Wright and Sandra Johnson.
Thank you for taking the time out this morning to spend with us and
to share some of your recollections of your involvement with the Space
Well, with your permission, I wanted to make a couple of comments
about my last interview.
We developed some confusion about a man that I was talking [about]
in the last interview that was involved in the initial architecture
and design of the Mission Control Center at Cape Canaveral [Florida]
for Project Mercury, and I got that name wrong. The man’s name
is Frank [J.] Chalmers, and I confused that with Tom [Thomas V.] Chambers
and other names, and so it’s corrected in your transcript, but
I wanted to be sure that people know that I finally got Frank right.
Also, I made a little bit of inaccurate summary of the remote sites
in Project Mercury, the tracking stations we had. I was trying to
list them, which ones had voice contact with the control center at
the Cape and which didn’t and what those sites were, and those
are corrected in your transcript also. That was pretty accurate on
the tape, but a little confusing.
Finally, the people that were with me at the reentry sites for Mercury
5, Mercury-Atlas 5 and 6, I got switched. For MA-5, which was the
flight of Enos, I was out at California, and [L.] Gordon Cooper and
Dick [Richard P.] Rembert were with me, not the two people I mentioned
in the tape, and you’ve corrected that. Then for MA-6, John
[H.] Glenn’s flight, Wally [Walter M.] Schirra [and Ted White
were] … with me out at California at the tracking station. So
those were kind of glaring errors, and everything else, you know,
we kind of tuned up. But I wanted to be sure I had straightened that
So today I would like to talk about the Space Shuttle. In fact, we
finished the last interview talking about Apollo-Soyuz and the flight
and how successful it was and the very good involvement we had with
the people in the Soviet Union that we worked with and their part
of the program.
The way I got involved in Shuttle happened directly because after
the Apollo-Soyuz flight, that [activity] really stopped. The people
in our program and at the Johnson Space Center kind of assumed there
would be a follow-on set of activities, either maybe another Apollo-Soyuz
kind of flight or perhaps a Shuttle-Soyuz or a Shuttle-Mir. Shuttle-Salyut,
I think, was up at that time. Anyway, we thought we would continue
a program of working joint operational missions.
In fact, right at that time there was also a change in the administration
in our government, and, for whatever the overriding reasons were,
our engagement really stopped. We did not have any follow-on activity
from the mission or flight people in NASA with the Soviets. [It] ended,
and the only part of our [joint] space program that continued was
the people that worked life sciences in both nations. The Soviet Union
and the United States continued to work together, continued to exchange
data and had an ongoing set of relationships. But none of that involved
anybody else that I know of in the NASA flight program.
In fact, that went on for a long time. I can’t remember what
I said last time about working with the Soviets, but for me and for
the people that I knew that were involved in the Apollo-Soyuz Test
Project, it was seventeen years before we re-engaged in terms of working
together in 1992, and maybe we’ll get to that today. I’ll
talk some more about that. It was a very interesting re-engagement
and led to a lot of things that are even yet going on today.
But, anyway, the way that led to me becoming involved in the Space
Shuttle Program was that we now had an Apollo-Soyuz Test Project [ASTP]
and an Apollo Program Office that no longer had any Apollo missions.
It was time to terminate the office. So what the Center did, what
NASA did, at that time was to—this project office was under
Glynn [S.] Lunney and I was his deputy, and they assigned the whole
office to work Shuttle payloads. We were coming into a time where
the initial development of the Space Shuttle was pretty far along,
and we were approaching the time for the first flights of the approach-and-landing
vehicle, Enterprise, and it was time to start thinking about payloads
and who would fly and what the rules of engagement would be and what
facilities the Shuttle needed to support payloads
So it was big piece of work, and it was very timely, and Glynn Lunney’s
whole office was moved off Apollo-Soyuz, off Apollo, and into Shuttle
payloads. In fact, it was called SPDP[O] [Shuttle Payload Integration
and Development Program Office], Shuttle Payloads Operations Development
Office, something like that. But that’s what they did, except
for me. And at that point in time, Shuttle development was pretty
far along, and as in any major development program, there were a lot
of technical challenges the program was dealing with, and things were
coming along, but there was a lot of interest to be sure we were doing
Bob [Robert F.] Thompson was head of the Space Shuttle Program, and
he asked me to come to the Shuttle Program Office and start a little
office there that was called the Program Assessment Office, and that
was something that interested me a lot. I was a lot more interested
in working on the Shuttle flight vehicles and systems than I was on
planning for payloads, and so it really resonated with me.
I went to take that job. I didn’t know anything about the Space
Shuttle. I had been too busy with all the Apollo things. I mean, it
was a brand-new thing to me. But over the next year, I put together
a team of about half a dozen good people, and we started to look at
all of the Space Shuttle systems and the Space Shuttle plans for how
the missions would be conducted and how we’d support those missions.
And so we had a very engaging time. It allowed me to get to know the
Shuttle very well, and it allowed me to get engaged with all the people
that were working on it.
Some of the people in that office, one of them was Gary [A.] Coultas,
who worked for me in a number of jobs after that. He was a very competent
guy. I believe I had Frank [C.] Littleton [and Bob White] in it for
a time, and also Bob [Robert W.] Moorehead, although Bob Moorehead
was in the Orbiter Avionics Systems Office. I think he came and worked
on some of the projects. And Don [Donald H.] Peterson, one of the
One of the interesting things that I was going to tell you about that
time period was, I decided if we were going to do that, we ought to
do it right. So I went after some fairly good people at the Center,
and most of the people I wanted to consider being in the office were
from the Engineering Directorate. And so I ran smack into Max [Maxime
A.] Faget, who supported the idea of having a Program Assessment Office
and looking at the issues, but he didn’t support the idea of
giving any of his key people to that initiative. So, in the end, I
got enough talent to do the job, I think, well, and I had a good time
Two of the things that we looked at during that time period had a
lot of impact on my later career. Probably the most significant thing
that I found in reviewing the Space Shuttle design had to do with
the solid rocket boosters. … Always before in our manned space
program, the vehicles we flew had more capabilities during ascent
and ways for the crew to get off, at least during the first part of
the flight and hopefully recover from a booster failure or launch
So when I looked at the Shuttle, we studied how it all came together,
we found that the solid rocket boosters on the Space Shuttle ignited
[at] liftoff and they both have to burn for two minutes. There’s
no way to separate them. There’s no way to get off. They have
to burn for two minutes. Then at two minutes they separate after they
stop thrusting, and they fall away. But if you try to separate them,
there was no way to shut them down, and if you try to separate them
while they’re burning, the momentum of the thrust holds them
in their places. They’re connected. So even you blow the bolts
to separate them, they won’t separate because the thrust keeps
them where they are.
I was quite alarmed that one failure could cause a catastrophe. There
would be no way to recover from a failure [of an SRB] during the first
two minutes of flight. So that was one of the findings, and I took
it and I briefed it to Bob Thompson, and I brief[ed] it to Chris [Christopher
C.] Kraft [Jr.] and others, and what I found was that everybody knew
that. They’d already dealt with it. They’d already accommodated
their thinking to the fact that that was going to be the way the Shuttle
ought to be put together and that we would simply rely on the very
high reliability of that system and press on in that manner.
So it was a big surprise to me, and, of course, you know where that
story leads [(Challenger)]. We did talk about ways you could maybe
shut a solid rocket booster down, and you can blow something off the
top, and it will dissipate the thrust. But no system like that had
been developed that was operational, that you could use. The analysis
of it was that it would take some period of time after you blow it
for the thrust to phase down, and it would cause an [unacceptable
thrust] imbalance. In the end, it was thought that that was probably
not something that was really reliable to implement that would be
any safer than just trying to do the rockets right and keep them like
they were. So the Shuttle system did not change, and that was probably
the biggest surprise to me when I first started looking at the whole
set of things that were going on.
The other area that I looked at and found a lot of interest in trying
to make some changes was in the Space Shuttle avionics system. The
avionics system in the Space Shuttle, I kind of felt at the time,
and I actually still do, is one of three major technological developments
that we had to do to make the Space Shuttle system work. Probably
the most complex technical challenge was the Space Shuttle main engine,
which is a liquid-oxygen, liquid-hydrogen engine that operates at
very high pressures and temperatures for the materials that were available
to be create it out of at the time, and it has to be of a size and
weight that fits on the back end of the Orbiter.
So it was a huge design challenge to make such a high-performing engine
and make it so reliable and make it with the materials we could make
it with, and I think it’s the most complicated, most elaborate
machine I’ve ever seen. It’s a wonderful piece of equipment,
and, of course, it’s been very successful. But it had a long
development program, a long test program, and there were problems
and difficulties that were solved during the test program. It didn’t
just start out being wonderful, but it’s been wonderful ever
The second technological challenge was the thermal protection system
on the Orbiter. The Orbiter now had to fly up like a rocket and return
like an airplane, and so it couldn’t have a blunt heat shield
on the bottom like the earlier spacecraft had. It had to have an aerodynamic
shape to fly like an airplane, but it had to have thermal protection
all over it. And to meet its performance requirements, the thermal
protection had to be as light as it could possibly be, because we
had a goal of what the cargo capacity could be in terms of weight
for the Space Shuttle, and so the Orbiter had to be lightweight, thermally
protected all over, and be aerodynamic in its configuration.
And that led to the system of tiles and tile mounting to the structure
that was developed for the Shuttle, and it was a wonderful system.
It also had development problems, but it has come along well as well,
not just the tiles, but the way that they’re mounted to the
aluminum skin of the Orbiter is quite complex. There’s a strain
isolation pad that goes under it [each tile], that’s called
SIP, and then it’s held down with RTV [room temperature vulcanizing
rubber], and there are, … there[’s] gaps between each
tile so that when the vehicle flexes you don’t break the tiles,
you just spread the gaps a little bit. But in some places you can’t
stand to have a gap, so there’s a thing called the gap filler
made out of several different things. It’s a very elaborate
system that we developed, and it has worked very well also. We learned
a lot about it as we flew it, but it’s a very good system.
Now, the third technical challenge is this avionics system. The avionics
system in the Space Shuttle was very revolutionary for its time. The
way the Space Shuttle works, everything in the vehicle, just about,
is actually controlled by a bank of computers. It’s not controlled
by the manual controls the crew has. There’s a bank of four
computers in the Orbiter that, in parallel, operate through data buses
to receive information from a whole variety of sensors on the Space
Shuttle and then send signals out to all the different actuation systems
that control the Shuttle. The aerodynamic … [surfaces] are controlled
by these computers, and the main engines and the separation of the
tank and the rockets and the nose wheel steering and the landing gear,
all those things that control where the Shuttle goes and how it flies,
are all controlled by these computers.
Even the cockpit where the crews have control sticks and switches,
they don’t go back to the engines or the aerosurfaces; they
go to the computers, and the computer then tells these things what
to do. So it’s a very complicated system, and the fact that
it’s all done electrically through the computers is a kind of
a flight system that’s called fly-by-wire. And I believe the
Space Shuttle was the first operational vehicle to be a fly-by-wire
Not long after, the Air Force had a fly-by-wire version of the F-16
fighter plane, but I think the Shuttle was built first and flew first
with this fly-by-wire capability. In fact, we had a test demonstration
airplane out at Edwards prior to that time that we used for research
on how the Shuttle system would be, and we test-flew that [system]
and then put it in the Shuttle.
So even though today it’s thirty-year-old technology, at the
time it was very at the forefront of the design of avionics for flight
vehicles. And for the Shuttle, because of the nature of reliability
and safety we wanted, we wanted this four-computer system to be what
we called fail-operational, fail-safe, which meant you could take
any failure in any part of this avionics system and still press on
and complete the mission. And you could take a second failure and
maybe not complete the mission, but you could safely return with what
was remaining. And that’s a big complexity in design and in
the architecture of how the system’s laid out. But that was
the intent, the goal, and that’s what was created.
So I was very, very interested in this avionics system, and at the
time I was doing work in the Program Assessment Office. The avionics
system was coming online for the approach and landing test [ALT],
the test program we did out at Edwards Air Force Base [California]
with Enterprise. The Enterprise was built to separate at 20,000 or
30,000 feet from a 747 and fly down and demonstrate the approach and
landing part of the Shuttle mission, but didn’t fly any other
part of the mission.
So when I came on, Enterprise was just about built and just about
ready to move into that test program, and it had the four-computer
suite and the fly-by-wire, but all it had in it for its software program
was just the approach and landing part. It didn’t have all the
things you would need to fly a total space mission, the ascent and
the stuff on orbit. And so it was a smaller, simpler software program
that went into this avionics system.
So when I came onboard with the Program Assessment Office, what I
found was this avionics team was now struggling to step from what
they’d created to go on Enterprise for approach and landing
to the much bigger job to build all of the software and do all the
testing to create the programs that would fly the Space Shuttle in
orbit. And so that’s a little background on the avionics system.
Now I’ll switch back to Program Assessment. What I found was,
and I guess it probably wasn’t obvious just to me, but there
was quite a lot of tension in the team that was doing the avionics
system, because Rockwell International was the prime contractor for
the Space Shuttle Orbiter, and they had the team of people doing the
avionics out there that had recently been in their Autonetics Division,
and they had built the avionics system for the F-111 Air Force fighter
plane, fighter bomber, I think it was. So they were very confident
that they knew how to do this avionics system and how it ought to
be done, and they were the prime contractor. But the software was
a subcontract to them by IBM Federal Systems, and IBM Federal Systems
really had a heritage of complex and very capable software. There
had been quite a lot of tension about who had the say about how this
[Shuttle system] would be architected.
So right about the time I came onboard, that kind of friction between
the prime contractor and the software developer had caused the Johnson
Space Center to change the contract, and they had taken IBM off being
the sub to Rockwell International and put it in as a prime …
contract for the Center out of Johnson. So now we had two separate
organizations rather than one reporting to the other, one doing the
software and one doing the hardware.
At the Center, our Engineering Directorate people had a long heritage
of working with Rockwell, and they were very kind of familiar and
in the groove with the way Rockwell did things, and so they were sympathetic
to how Rockwell wanted to approach the development of the avionics.
But the software, the IBM organization, had spent a lot of years working
with our operations directorates doing the ground systems for the
earlier flight programs, and so their connection into Johnson was
with another set of the management and a team of people that were
really very confident in them and had worked with them.
So we had this tug of war going on while we were trying to build this
very elaborate system for the first Space Shuttle flight. So I jumped
right in the middle of that and kind of got caught up in it in terms
of trying to understand what could be done. I’ll talk some more
about that in just a minute. But that’s one of the findings
in the avionics. The other two were more technical.
The second was, as people started to look at all of the software we
were going to need for the orbital flight tests and the orbital missions
to follow, it was clear that the programs we had to build for ascent
and for entry would not fit in the memory of this computer. This is
an AP-101 computer, also built by IBM. It was built up in Owega, New
York, by IBM, and it was called an AP-101. It had a graphite solid-core
memory, and that memory could take 64,000 sixteen-bit words. That’s
what the approach and landing test computer was, and it worked fine
for that software program. But as we looked at all of the things that
this software had to do to manage the Space Shuttle and control the
flights, do the mission, do the steering, it couldn’t fit. So
one of the problems was the computer memory size was not adequate
to do the job. Something had to be done about that.
The third thing I found was that this fail-op, fail-safe system that
the avionics system was supposed to [be] for the vehicle, so that
you could take two failures and still recover, was quite well architected.
There were some issues with it. There were some places where they’d
tried to make a choice of four things for the four computers flow
into or out of three sensors or actuators rather than four, just for—I
always thought it was just to make it more complicated-but some things,
like the inertial measurement units—I think there were three—because
they were big and they took space and someone thought we could cleverly
work them in a way that you could get a voting choice of four out
of the three if you worked it cleverly and used the right kind of
But, anyway, the system that got created really was fail-op, fail-safe,
except there was quite a worry that what if there’s a failure
in this software that works in the four computers so that one software
failure would cause you to lose the whole thing? The way these computers
worked together was that they synchronize with each other twenty-six
times a second. They each tell the other where they are and check
to make sure they’re all marching in parallel and computing
the same thing, and they each take independent inputs and then they
provide independent outputs over these four data buses that go out
and control the things I was talking about. But they all check each
other’s outputs, and if the bank of computers sense that one
of the computers doesn’t sync right, twenty-six times a second,
or doesn’t compute the right output, it’ll vote that [one]
out, just instantly take it out of the loop, and so that computer
… ceases to be part of the voting logic unless it’s re-initialized
and put back in by the crew.
So that logic that’s in the four computers where they crosscheck
each other and crosscheck their outputs and decide who’s good
and who isn’t, it’s … [a set] of logic we called
fail-to-sync logic. The computer, it’s dealing with one of the
computers that fails to synchronize with the rest of the set. And
this fail-to-sync logic was unique and very difficult to prove that
you knew all of the possible bugs that could be in it and it could
never make a mistake when it was doing that. And that’s the
software that I was talking about in this other issue with the avionics
This team of people and other people I talked to, many of them were
concerned that there might be something in the software where instead
of … [voting] out a bad one, somehow they’d all get out
of sync and it wouldn’t work anymore, and then the Shuttle couldn’t
fly. So it was really determined that we ought to have fifth computer
with a separate software system in it that was independent, and so
if for some reason one failure took this big elaborate set of redundant
logic down, you’d have something to switch to that could fly
So at the time of the Program Assessment Office, I talked to you about
the solid rocket booster and what we found there, and then in the
avionics system there were three things. They were having difficulty
with the program team converging on a flight architecture, and the
computer was too small, and there was concern that even though it
was supposed to be two-fault-tolerant, it might not be. So that led
to my next job.
What frequently happens when you turn up a spade with stuff like that
in it, was that I was asked to become head of the Orbiter Avionics
Systems Office in the Orbiter Project Office and take charge of those
… things in the avionics that I just talked about. So I was
in the Program Assessment Office maybe for a year, probably a little
less than a year, from mid-’75 to mid-’76.
Then I moved to the Orbiter Project Office, now knowing a lot about
the Space Shuttle and also kind of having a focus on these problems
in the avionics system. I think one of the reasons I was selected,
I’d had quite a long history of working with Rockwell and North
American, as they were called before, on the Apollo Command Service
Module vehicle for quite a few years. I had good relations with Rockwell,
and also I’d worked in the Mission Control Center for a number
of years, and I was very familiar with IBM and with the software kind
of things that they did. And so I was kind of a neutral person for
the two, and so that was the job I was given.
It was interesting that right at the time that I started doing that
work, it was just the time when the team of Rockwell and IBM and [NASA]
engineering and flight operations were all getting together to define
in detail the content that we would have to have in the computers
to fly a full Space Shuttle mission. Within a matter of a couple of
weeks from moving into that office, I was in charge of conducting
reviews out at Rockwell where we would define and create the documentation
for the detailed requirements for the software for the ascent flight
phase and another package for the on-orbit flight phase, and a third
one for the entry flight phase, and a fourth one for systems management.
These were multi-day reviews at Rockwell, and all of these different
elements in the program would present what their system has to do
and what the mission is and what the software would have to have in
it to operate.
That was also a very strong learning process for me. It was the kind
of thing that I was familiar with in terms of those kind of systems,
and I knew a lot about space missions and space vehicles. So, I mean,
it was totally new in terms of all the uniqueness, but it was just
exactly the kind of thing I was familiar with dealing with.
So I had a very good time conducting those reviews. I think they were
very effective, and what we created out of them was a series of documents
that are called flight software [system] requirements documents, or
FS[S]R’s, and they were written by IBM, and they had these requirements
at a very detailed level, in English so you could read what the software
was supposed to do.
One of the mistakes a number of software initiatives did in the early
years, maybe they still do, is not documenting the requirements well.
You can document the software requirements at a pretty high level
and give it to the software team, and then they’ll define what
they’re going to build in software language. And so there’s
no English language for the people that might want to get engaged
in exactly what was being created to look at. Well, in the Shuttle
we didn’t do it that way, and I think I credit IBM for the process
and the way … [to cause] it to be done. These FS[S]R’s
were very detailed. They talked about all the subtleties that had
to be built into the logic in each of the different areas, and they
were really excellent documents, and we built this whole set of them.
The way the Shuttle works, I’ll talk about how we got the computer
size bigger in a minute, but even with more size in the computer,
you still couldn’t load all the software in the computer that
you needed for flight and have it all in there at once. What you had
to do was, during the countdown there was a preflight checkout load
that would be in, and then when you got close to the time you’re
going to launch, there’s a mass memory unit. There’s a
couple of those. They were redundant mass memory units in the Orbiter,
and you roll the software out that was doing the checkout and roll
in the ascent load from the mass memory. You fly ascent, and then
when you get up, you roll out the ascent and roll in the on-orbit,
same thing for entry.
So each of these was really a separate stand-alone computer program
to do those different phases. It still is today. It’s still
the same way. You could have the systems management in there, because
it managed the systems and the vehicle the whole time and it wouldn’t
change. But it would be resident in one part of the computer, and
then these other loads would come in and out, depending on the mission
So we had a flight software requirements document built for each of
these phases, because it was a separate computer program, and after
these reviews out at Rockwell where we defined the broad capabilities
that we would have and tried to define the details, in the time frame
after that these documents were continually refined.
What got to be my role was Chairman of the [Orbiter] Avionics Software
Control Board [OASCB], that we had at Johnson, and all of the organizations
on the Shuttle at Johnson were members, and IBM was a member, and
Rockwell was a member. We met every week for a day-long meeting or
longer, for years, and what we were working on is the baseline of
Once we built the FS[S]R’s, their configuration was frozen,
and the only way to change anything in there, whether it’s the
detailed color of a light or the rate … [that a surface] moves
… [at] or, anything you wanted to change had to come to the
Software Control Board with a documented change, review it, get it
approved, put it in. And each change was a change packet that I would
sign. I think I must have signed thousands of change packages for
the software, but the configuration control was very rigid. The documentation
was very thorough, and I give that quite a lot of credit to how well
the avionics system [has] worked over the years, because it was a
very rigorous process that was created.
One other thing I’d really mention about that time. When we
had those first reviews of the requirements … [for] the Orbiter
software, I met a person that I respect about as much as anyone that
I’ve worked in the space program with, and that’s Sy Rubenstein
out at Rockwell International. He was the man, at that time, in charge
of the avionics system for Rockwell, and later on he became head of
the whole [Rockwell] Shuttle Program, and then he was head of Space
Station initiative that Rockwell had. They didn’t win [Station],
but he was the leader of that.
Sy was a very capable, competent man, still is, and I enjoyed working
with him a lot over the years. He added a lot to our system. In fact,
he became the [Rockwell] Shuttle program manager at a time later when
we were having problems with the tiles on the first vehicle, and he
was involved in solving that problem and many others. But Sy was really
a very outstanding technical leader for the Space Shuttle and deserves
a lot of credit for it.
So now I had my new office, the Orbiter Avionics Systems Office, and
I was well into the software requirements, and I was running the Software
Board, but we still had these other two problems I talked about. We
had a computer that was too small.
During that time frame, Aaron Cohen was the Orbiter Program Manager,
the fellow I worked for, Project Manager, Orbiter Project Manager,
and he didn’t want to buy new computers if he didn’t have
to. So this was probably the Christmas of 1977, I think. It turned
out at that time he had Fred [W.] Haise assigned as a special assistant
to him, and Fred told him that he could kind of re-formulate these
requirements so you could skinny them down and get them into the 64,000-word
computer. And Fred took, like, a ten-day leave over Christmas, and
I think he went to Big Bend [National Park, Texas] or somewhere and
sat and mulled on the computer requirements and came back and made
a pitch about, you know, “If you did this and did that and did
the other thing, you could cram them into the box,” and it was
a great effort on his part. But, really, too many things fell off
the table, and we couldn’t step up to some of the loss of redundancies
and some of the loss of capabilities that it would take.
So we had to have a bigger computer, and at that time IBM came forward
and said they could do what they called a double-density version of
the computer. It essentially doubled the memory in the same box, still
the core memory. So we worked that change, and Aaron approved it,
and we increased the memory from 64,000 sixteen-bit words to 104,000
sixteen-bit words. And then we got a shot at cramming these ascent
and entry and orbit programs in. It was still a challenge. We still
had to do some scrubbing and be careful with growth, because it wasn’t
an easy fit, but it fit, and that’s how that problem got solved.
We took this issue with perhaps needing a backup system, an alternate
to the four-computer bank, all the way up through Center management.
A lot of people felt strongly that we ought to have a backup system,
and we worked on what it could be, and what it turned out to be was,
it turned out to be an identical computer, but a fifth one, and a
switch, kind of a lever switch sort of thing that the crew could operate
that would take the four sets of data buses that went in and out of
this big primary computer bank and switch them over so they would
all be controlled by this one computer, and it wouldn’t be redundant,
but all four of them would get driven by that single computer.
Then it was proposed that the software in there would be a skinnied-down
[single load] version, just for a safe recovery, of these bigger programs
that had to go in one by one into the four-computer bank. We would
have that software built by somebody else so it would be new, unique
software, and it would be physically separated from the other set.
We took that forward as a proposal for implementation. Chris Kraft
didn’t really think that was the right thing to do, and we had
a very intense set of meetings telling him the preponderance of the
people who actually thought that was a failure mode worth providing
coverage for. [I believe that Chris felt that we had to work the primary
computer system until it was highly reliable and once we’d achieved
that a backup was not only unnecessary but perhaps a distraction from
doing the primary system right.] In the end, he acquiesced, and we
did built the fifth computer, and we had Rockwell build the software,
this new program that would do all the things you needed to abort
and get home, and they tested it independently, and it was different.
It was the same requirements but different software, and that became
part of the now five-computer set of orbiter avionics system that
controls the whole Space Shuttle. So those were the things that went
on at that time.
The other part of my job was when you’re doing, particularly
doing software, but also doing these kinds of systems, you have to
do a lot of test verification. So I was also in charge of the test
verification programs for both the software and for the full avionics.
The software was verified by IBM. I think they had a very, very capable
way of doing that. There’s been a lot of talk in the years since
that time about the right way to do software is to have independent
verification and validation [IV & V]. Whoever builds the software
would build it and test it as they build it, and then you’d
turn it over to some other organization, and they would do a test
program to prove that they couldn’t find any bugs that were
in there that you didn’t … [find].
The way we did the Space Shuttle, the way IBM did it, they had independent
verification, but it was within their organization. They had a development
organization to build the software, and they had a verification organization
to test it, and they were physically separated, but they were all
IBM and they were all in the same building, and they used the same
The way that works, you agree on these requirements that are in these
FS[S]R’s, and then that’s what IBM takes away to build
the flight software code from, and they have an organization that
will create software logic to meet those requirements, and they’ll
do development testing as they build it. This verification organization
takes the same requirements, and they construct a test program that
they would have to show that the software would do what it’s
supposed to do. So when the software is coded and the development
organization thinks it’s correct, they hand it over to the verification
organization, and they start all over, and they run this test program
they’ve created, and they do their tests. So that’s the
way the flight software was done, I think it’s still done that
way, and it certainly was rigorous and thorough.
But that’s not the end. After you get it all verified and deliver
it as the flight software package, then you have to test it with the
hardware. The way the Shuttle Program does that, they have a very
elaborate test facility over in Building 16 at Johnson called the
SAIL, the Shuttle Avionics Integration Laboratory. It is a physically
accurate representation of the Space Shuttle Orbiter in terms of the
wire runs, the electrical power, the data buses, all of the avionics,
the electronic boxes in the front end and in the back end and in the
payload bay. It’s all there laid out in the specific configuration
that it is in the Orbiter vehicle, and it’s got a big test laboratory
around it, and you load the software in it.
So now you have a full representation of the Orbiter. You have all
of the hardware in the avionics, all of the software in the avionics.
In the laboratory around it, you have a series of math models. The
math models simulate the flight environment that you’re going
to fly through, and they simulate the other elements of the Shuttle.
We don’t have a solid rocket booster or a tank there or a main
engine, so they simulate those. But the electronics are all there,
and so you can fly a full mission.
There’s a cockpit station, and you can fly a whole mission in
the Shuttle Avionics Lab, and because of the physical layout you can
test things like EMI [electromagnetic interference] and electrical
transients and things that you find in the integrated system in the
vehicle that wouldn’t show up in the design testing of these
individual components. And you can fly the whole stuff together and
see how the software and the hardware works and check [response to]
faults and see if the right things occur the way they’re supposed
to. So it’s a very elaborate test program in the SAIL before
you’re done, saying that the whole avionics system is flight-worthy.
So that program was part of this Orbiter Avionics Systems Office [responsibility]
I don’t know how enthusiastic this discussion has sounded, but
this was one of the nicest job[s] I ever had. I really enjoyed this.
I did it from about 1977 to 1981, and it was really a fine job. Maybe
1982, because what happened after that—yes, it’s 1982.
What happened after that was Aaron Cohen moved on. I think he went
to be head of [JSC] engineering before he got to be head of the Center.
But in any event, he left the Orbiter Project Office. So in 1982 I
was asked to become head of the Orbiter Project Office and moved up
from this avionics job. But by then all of the software had been developed,
and we’d flown four flights with the Space Shuttle, maybe five
flights. I think … [I] started around flight six, was the first
flight I was in charge of the Orbiter and not following in the avionics.
So that was a nice package of work, ’77 to ’82, when all
of this came together, and we made this avionics system work, and
it’s really worked flawlessly all these years since. It’s
a wonderful system. People always talk about needing an avionics upgrade
and how it’s old technology. Well, it’s very thorough
technology and it’s very elaborate, and for the time it was
cutting-edge, and it’s still darn good stuff.
In fact, as they talk about upgrading the avionics, particularly the
crew cockpit, you have to be very careful when you go in to modify
any part of this system, because it’s all interrelated, and
if you’re not careful, some of these very elaborate protections
that have been built in and tested and redundancies that are in there,
you could perhaps alter in some way that’s subtle, that could
make the thing less reliable rather than more reliable, if you go
in, just because it is so complicated, and there are so many thousands,
millions, I guess, lines of code that have been created and tested
over the years.
One of the current astronauts was just telling us yesterday what the
software may lack in flexibility now, it makes up for in its essentially
100 percent predictability, because you know exactly how it’s
going to react in any given situation. And so to mess with that, as
you point out, can be unpredictable.
This fail-to-sync logic in these four computers was very complicated
to build, and there was a fellow at IBM named Lynn Killingbeck that
was, I guess you’d call the technical guru of that. And when
we first started doing that in the lab, we would have fail-to-syncs.
We’d [have] the computers break up and come apart functionally.
But over a few years of intense testing and work by this IBM team
and this Lynn Killingbeck, we may not have ever found all the bugs,
but we found all the ones that were likely to occur, because when
it finally got solid, it was rock-solid.
The first Shuttle flight had some sync issues, though, before it ever
I was going to talk about that. Actually, I’ve got that listed
down as one of the things to make a point of. It relates back to this
Rockwell and IBM contention for who’s the real expert.
We’re now at the time of the first Shuttle flight in 1981, and
so all the software’s been built. It’s all in the mass
memory, or in the computers, and the backup system’s there.
It’s been built and provided, and it’s been certified
and tested also. The way the launch sequence would work is that the
four-computer bank and the checkout program was in there from the
time you powered the vehicle up, twenty-four hours or longer before
the launch would occur, and it would do all the things it’s
supposed to do down to about minus twenty minutes. At minus twenty
minutes, the primary system is required to synchronize the [backup
flight] software with a logic update for guidance and positioning,
so the backup system can start in the right place, because it hasn’t
been [engaged]. The way the primary system works is, even on the launch
pad, it navigates during this whole period pre-launch. So it really
know[s] exactly where the Shuttle is, what its position is, because
the Earth’s rotating.
In space, you call it the state vector, about where the vehicle is
and where it’s pointed. Well, you still have a state vector
when you’re on the launch pad, and the primary system’s
doing this, but the backup one is not involved in this until minus
twenty minutes. So at minus twenty minutes, the primary system has
to tell the backup all this information so it can start, and from
there it tracks and does its own thing.
That was the flaw on STS-1. At minus twenty minutes, when it was time
to load the backup with this initialization from the primary system,
it didn’t take. So we had to stop. The preponderance of the
colleagues thought all of this wonderful elaborate IBM software and
this big four-redundant-set system that was so fine was in great shape,
and this little backup with the software built by Rockwell had some
problem that they didn’t get right. And so over the course of
the next evening, the next twenty-four hours, about halfway into the
reviews we conducted to find out what was wrong, it turned out that
the backup system was fine, and the primary system did not send the
initialization in the way that was correct, and it was an IBM software
problem, which was easy to fix.
I think we just went one day later. We found the little thing and
adjusted it. But it was interesting that the initial reaction was
that Rockwell didn’t do this right, but in the end IBM kind
of had to step back and take credit for having something they hadn’t
quite gotten in the right place. And that was easy to fix. It didn’t
have any further impact. It pressed on from there.
Jack [John R.] Garman was telling us that that episode caused him
a few headaches, I think.
Well, he led the team that worked all of those aspects of that, all
the issues related to it. I can’t remember how long it was.
It wasn’t more than a few hours before everybody was not blaming
Let’s see what I’ve talked about here, because that was
interesting. I wanted to tell you about that, and I think maybe there’s
one other thing about that time period.
When you had first become involved with the avionics, at least in
this capacity, had the approach and landing tests flown?
Tell me the time frame again.
Well, the approach and landing tests were about mid-1977. I think
they finish up in October, and I was wondering if you were in charge
of the avionics at this point, or if you had any involvement of the
results of tests of avionics that were uncovered through the ALT Program.
I really was in charge at that time, but I’d just come onboard,
and all of my focus was on the first orbital flight. The computer
programs for ALT had been built, and they were available to be in
the vehicle, and they’d been tested, and they really weren’t
a big issue. This redundant-set thing had had some testing, but the
whole computer program was smaller. It was more simple, and so there
weren’t issues with that. I did participate in monitoring the
flights and the flight readiness reviews, but I didn’t really
focus on doing much for the ALT. [Earlier there had been some fail
to sync and maturity issues with the ALT software but they had primarily
been worked by people in the JSC Space Software Division (SSD) on
the NASA side and by their directorate head, Bill Tindall.]
As far as the avionics go, how was the pacing and the scheduling of
that through the late 1970s, I guess, comparable to other areas of
the Orbiter, the Shuttle Program, as a whole?
Well, when we found out that we had to put in these bigger computers
and we found out all this software that was going to get built and
we had some of these problems, like this fail-to-sync problem, it
looked like the schedule was very tight to make the first flight for
that. I’m sure we thought we were behind schedule. But I’m
sure you have talked to other people about the fact that when Columbia
was built, and it was built in Palmdale, like all the Orbiters were,
and then it was flown to Florida, when it took off on the back of
the 747 from Palmdale, a whole bunch of the tiles just came off as
they went down the runway. I mean, they didn’t have to fly a
long time. They just fell off.
That led to the requirement to have a lot better understanding of
how the tiles were attached and how to know they were well attached,
and that problem took two years to solve. The Orbiter did go to Florida,
but it got down there, and it had some large number of tiles missing,
and, not only that, we had to figure out why they came off and how
to be sure you could mount them so they wouldn’t in the future.
That was a two-year problem, so the launch of the first flight slipped
from 1979 to 1981, which, for people who were behind in the software
and behind in getting the computers updated, was a great window of
additional time. So by the time the tile problem got back on track,
all of the stuff I’ve been talking about in the avionics also
was on track for first flight, and I’m sure we would not have
made the initial flight date with the things we were doing in the
What sort of direction were you receiving from the Program Office,
from Bob Thompson, for instance?
Well, the Orbiter Project Office was a very strong JSC project office,
and Bob Thompson and Aaron Cohen were almost equals in terms of where
they were in the management structure and the scope and the size of
their responsibilities. And so, I mean, I worked daily, hourly with
Aaron Cohen during all this time. But I didn’t get a lot of
interfacing with the Space Shuttle Program.
We were doing the software that … [interfaced with] the main
engines. The main engines had their own computer[s] provided by Marshall
[Space Flight Center, Huntsville, Alabama], and they were mounted
on … [each] engine. But then there’s an interface unit
in the Orbiter for each engine that is part of the Orbiter equipment,
and it was part of this avionics system that I … [worked] on.
So there was quite a large interface with the main engine and a smaller
interface with the boosters and the [external] tank, and those, of
course, are Bob Thompson, other projects at the Marshall Center. So
we had to have the software requirements right and tested for them
But there wasn’t anyone who did avionics in Bob Thompson’s
office. I essentially did it for both Bob and Aaron, but did it primarily
in the focus from Aaron with an engineering team that included people
from those projects as well. So I didn’t feel closely guided
or directed by the Shuttle Program, but we knew very much what we
were doing in conjunction with them and how we were doing it and how
the teams worked.
What sort of involvement did the flight crews, the astronauts themselves,
have with development of the software?
They had a big role in it. This Orbiter avionics stuff, we had a control
board that always had a representative from the flight crew, and during
most of this time Bob [Robert L.] Crippen was the representative.
He would come to every meeting, and most of the things we’d
do would have some crew-related consequence or impact or requirement,
and so we had to work those things carefully, and he had to be sure
that they were worked in the way that would suit the crew, but yet
also that the vehicle could actually accommodate.
One of the things I found in these detailed requirements we put together,
even after we had a pretty good set and we were working forward, was
that the crew interface was very complex. You know, the crew has a
lot of indicator lights that turn various colors for different things.
They have lots of switches, and they have other controls, and then
they have cathode ray tube displays with data on them, and all of
those things operate with the computers. They don’t operate
around the vehicle. They go to the computer, or they come out of the
If you got into the “what if” discussions about some of
those controls and displays, many things in the Shuttle can occur
automatically, or the crew can override and do them themselves, or
they can take some other action based on the data they see. So you
have a very intricate involvement with the crew procedures and the
crew checklist and what the crew’s choices and options are.
So when you build the software, the software doesn’t just do
it one way. It has to also interact with the crew and the displays
and the controls, and it has to give them alternate manual control
in many instances, if that’s the way we choose to proceed. So
that makes a very intricate set of requirements on lots of details
that go on in the crew area.
I’ll just tell you one example that we found along the way.
In the past programs, I spent quite a lot of time in my role as the
operations systems person for the control center in analyzing these
space systems in detail and how they worked. So analyzing these kind
of things came kind of natural to me, although I was now running [this]
Software Board. If you probed in a little bit about “what happened
if” with some of these things, you could find they didn’t
cover all of the situations that might occur.
For example, when the Space Shuttle does an entry, after the de-orbit
burn, the first phase, well, it turns around. It does a de-orbit burn
back end forward, but then it turns around and starts in like an airplane.
But there’s no atmosphere yet. So it controls with its little
rocket engines to approach the atmosphere, and as it enters the atmosphere,
you get to start having some control with the aerosurfaces and you
still have the rocket engine authority and the two work together.
And then after you get down heavier into the atmosphere, and you get
into the high-heat-load phase, it’s all aerosurface. The little
rocket engines don’t have enough authority anymore in that atmosphere
to control it.
Then when you get through the heat phase and you’ve taken off
a lot of the energy and … [you’re coming into the vicinity
of the] landing field, you get into a thing called the heading alignment
circle, where you approach and you fly at a high altitude around the
runway to get lined up with just the right amount of speed and altitude
to do the right kind of approach for landing.
So there are phases. You don’t have just one entry sequence.
You have a series of modes in entry, and they’re called 301,
302, 303, 304, [etc.] and you want to do them in sequence. I mean,
you want to do 301 first, and the flight system can step through those,
or the crew can call them. They have a capability to engage these
modes, and they can manually step through them at the right time.
And that’s the way the software was built.
But what we found was, if for some reason the crew inadvertently presses
the wrong one, like when you’re supposed to go to 302, you press
303 instead, it would go to 303, and you don’t want that to
happen. So that’s the kind of thing that I was saying is subtle
with the crew interface that we had to think through. We had allowed
the crew all the things that they had to have, but you also wanted
to have all the protection so they couldn’t do something that
was really inadvertently bad and could get them in trouble. So that
took a lot of talking through and analysis and thought, and some of
the times it just took another line of questioning. You’d say,
“Well, but what if such-and-such were to occur?”
“Oh, well, that wouldn’t happen.”
You’d say, “Well, we don’t want the software to
let it happen.”
So we did a lot of that, and it did involve the crew, and Bob Crippen
was a wonderful person to work those kind of things with, but we worked
them with other crew people, too. Sometimes there’d be more
than just one person at the meeting.
In fact, you talk about first flight, I was in charge of the avionics
system, and because of my role at that time, I wasn’t as sensitive
to some of the other things in the Space Shuttle that are very challenging
things to do. I didn’t know much about the main engines at that
time or the ascent. That wasn’t a focus of mine. But I was really
concerned about this entry because we had worked on entry and the
control logic. The avionics controls the vehicle during that time
frame. It does during ascent, too, but it is so much more, I guess
you’d call it automatic, during ascent. The entry, there was
really a lot of concern about how much control authority the vehicle
had to have with these little engines and with the aerosurfaces and
the amount of atmosphere.
The biggest anxiety I had during first flight was how well this vehicle
would fly through that entry sequence, because we’d never been
able to fly it. We’d only done it in laboratories and testing
and analysis. And so the biggest moment for me for STS-1 was waiting
to see it come out of blackout, because if it came out of blackout,
it was now in a very easy, straightforward—… [landing’s]
still a challenge, but we’re now in a lot more certain regime
of the flight. Approach and landing testing had done landing. But
the blackout thing was my time of most concern during first flight.
And, of course, it came out of blackout fine. It was just fine.
The other thing we did wasn’t really part of my job, but we
had been concerned about this thermal protection system on the Orbiter,
the challenges of creating it, and then this problem we’d had
with the tiles not being mounted as thoroughly as we thought they
would be. A lot of us were of the opinion that you probably couldn’t
afford to lose a tile during flight. Just the loss of one tile would
create a place that would overheat and burn through, at least on the
So we didn’t know a lot about the flight performance of the
thermal protection system. The fact is, it’s much more flight-tolerant
than we initially thought it was, which is good because we’ve
had damage on many flights of the tile system.
But, anyway, after the first flight, Chris got together a group of
the senior people on the program, and we all flew out to Edwards,
where the vehicle had landed, to look at the tiles. There was a lot
of damage on the underside, but there was one tile on the body flap
that had a really big ding into it, and when you get a big hole in
a hot area, then the heat kind of melts like an inverted snowcone.
It just kind of melts into the tile. And so there was this one that
was really quite damaged in a very heat-sensitive area, but it didn’t
melt all the way through. It had survived. So that was interesting
to go out and look at that, and that got to be another feature of
my next job, worrying about the tiles.
That was the first flight, and during the first half dozen flights,
we found that there was quite a lot of material coming off the solid
rocket boosters and the [external] tank during ascent, and invariably
it would come off. It would hit the bottom of the Orbiter and make
little dings that now would be threats to the entry phase, to cause
That was during the time Glynn Lunney was the Space Shuttle Program
Manager. He’d replaced Bob Thompson. We made a number of changes
in the Shuttle Program to the external tank and to the solid rocket
booster mounting mechanism for things, so that some of the stuff that
would come off no longer was there.
For example, the big external tank had a lightning-rod protection
system on it. It had a metal ring around the top of the tank. I think
it was copper and then … [a wire] down so that if it got hit
by lightning, it would have lightning protection. And this ring was
under the thermal protection system on the tank so you couldn’t
see it, but it was there. This was probably the most glaring example
of things that got fixed during that time period, but we found out
that during ascent, in fact, the ring was breaking apart and coming
off, and the copper was hitting the Orbiter and making some of the
damage on the Orbiter that we were seeing. So it turned out you actually
didn’t have [to have] a lightning protection system on the tank.
The whole launch configuration is protected for that, and so we took
Another thing that caused damage was ice on the tank, and our monitoring
and controls for being sure not to launch with ice on the tank got
worked extensively during that time period, because little pieces
of ice would come off.
So the early flights were kind of test programs for the Shuttle vehicle
and the thermal protection system and how to tune it up, and I worked
on that both before I got to be the Orbiter Project Manager, which
started with flight six and then during the next fourteen flights
or so, working with Glynn Lunney on the whole Shuttle system and the
effects on the Orbiter thermal protection system as a major activity.
The top of Columbia initially had a lot of very small white tiles,
and they were hard to maintain, and they were damage-prone, and the
upper surface, the big areas that weren’t in a high heat mode
had this—it’s kind of a felt pad with a white layer on
top. It’s called FRSI; it’s flexible surface reusable
insulation, I think, is what FRSI is. But in the higher [upper surface]
heat areas, they had these little tiny tiles, and it was hard to maintain,
and they weren’t a very good thing. So during that time period
Rockwell developed this thing, this material that’s a blanket
called advanced flexible reusable surface installation, advanced FRSI,
and we started to use that.
But we found the surface of that would erode and break up during the
flight phase, and so they had to get a surface coating to paint it
with. So now the high-heat areas in the top of the Space Shuttle orbiters
have this advanced FRSI on them, and it’s a nice blanket you
can put on [in] big pieces, and then it’s coated with a material
that toughens the outside surface, so it has a long life to it. So
we made that change.
The other thing, if you look at an Orbiter today, you’ll see
most of the top surface is white. But on both of the OMS [orbital
maneuvering system] pods, on the front of the OMS pods, there’s
a black area of the tiles on each one. That’s because those
two areas stick out kind of in the jetstream flowing across the front
of the vehicle to the back, and we were continually getting damage
on the OMS pods on the tiles there, and these black ones are a tougher
tile that Rockwell developed. They’re still tiles, but they’re—I
can’t today remember what we call them, but they’re significantly
tougher than the white ones you see on the rest of the OMS pod and
the rest of the vehicle. And they were put there specifically because
we always get some amount of little material coming by that in most
flights was damaging the front end of the OMS pods, to the point where
if you had a really big piece of damage there, that could be flight-critical,
because the OMS pods are needed to control the vehicle [and they contain
the tanks and lines of high energy propellants].
That’s sort of the TPS story, and I think I’ve kind of
moved into this new job. In 1982 I got to be the head of the Orbiter
Project Office, starting with the sixth flight, and from there through
the twentieth flight, I was in charge of the Orbiter system, and some
of these things I’ve just been talking about with the tiles
was after I moved on from avionics, although I think it started while
I was still on the avionics job.
Another thing we did during that time period I was in charge of the
Orbiter, the first four flights were called development flight test
[flight] flights, and there was a big rack of instrumentation. Essentially
the payload was an instrumentation package that had many sensors all
over Columbia to sense temperatures and pressures and dynamic loading
all around the vehicle to really get much better analysis of what
the real flight loads and temperatures were. So about the time that
I took over the Orbiter, the data from those flights was becoming
available, and what we were seeing is that in many areas the Orbiter
was over-designed. It was too strong, too beefy, and what we could
actually do was take about 8,000 pounds out of the Orbiter by redesign.
That was very desirable because that would be directly related to
So those changes were under way when I took over. Both Columbia and
Challenger were built to this heavier design because they existed,
but Discovery, Atlantis, and then Endeavour weren’t yet created.
So they could take advantage of this knowledge of areas where we could
take some of the weight out, and that was in the plan, and that was
in work when I took over. But I managed the time frame that Atlantis
and Discovery were built and delivered, and they were lighter vehicles.
In fact, today Discovery and Atlantis and Endeavour go to Space Station,
and they fly Space Station missions because their performance is suited
to it. Columbia isn’t used for Station missions because it’s
almost 10,000 pounds heavier, and it can’t carry the payload
to Station that the other vehicles can. I mean, this is hard fact
in … [metal] that happened in the time frame I’m talking
about, but it’s still part of the way things are.
On the other hand, not everything was over-designed. What we found
was that there were certain areas on the underside of the Orbiter,
I think forward of where the main gear doors are, that actually were
under-designed for thermal protection. The loading on the vehicle
when it flies ascent and then entry, the most stringent design requirement
is not the structural load of flying the vehicle, it’s the thermal
load when the temperature rises. It puts a stress in the vehicle,
and, of course, the load you have to deal with is with the combined
thermal and structural load, but the entry thermal load on the underside
of the vehicle is the most strenuous thing that the vehicle see[s].
These areas forward of the main landing gear doors, the thermal protection
system wasn’t quite up to the design requirement, and so we
either needed thicker tiles or we needed some other kind of correction
to keep from overstressing the structure inside.
You couldn’t just exactly change the tiles, because you had
to have the aerodynamic smoothness of the vehicle to fly the thing
aerodynamically. So I think we might have changed the thickness of
the tile a little bit, but we also added heat sinking on the inside,
and we beefed up the structure on the inside of the Orbiter. So for
Discovery, and later Orbiter[s], while, for the most part, we were
taking weight out, in some areas we were doing beef-ups, and this
was all based on what we learned from the flight instrumentation on
the first four flights, before we took the DFI [Development Flight
Instrumentation] rack out of the payload bay.
How did building the second two Orbiters compare with Challenger and
Well, of course, I wasn’t … [actively] involved in building
Challenger and Columbia, but we’ve talked about some of the
development problems that happened for the first flights, which was
Columbia. I talked about some of the avionics problems. We talked
about the tile problems. I’m sure there were many, many difficulties
with developing these vehicles and building them, and it was much
more smooth for Discovery and Atlantis. I mean, we were now building
a known configuration, even with these changes. They … [flowed]
quite smoothly, and they were in excellent shape when they were delivered.
Kennedy [Space Center] would always complain about some of the unfinished
work that would come when you deliver an Orbiter, because for various
reasons, it got to be time to send it, and you’d send it, and
there’d be this big package of stuff that was still [to be done]
—but that was really trivial in terms of the scope of creating
One of the things that I enjoyed a lot, I accepted both Discovery
and Atlantis for the government at the roll-out. I went out to the
ceremony for the roll-out [for each], and I made a speech on the viewing
stand to take note of delivery of the two Orbiters to the government.
The second speech I did, the one after Atlantis, at that time we thought
that was the last Orbiter that was going to be built, so it was kind
of an emotional speech about the wonderful job and all the things
that the Rockwell team had done in creating these orbiters. And they
liked it so much that they used my speech in their HR [Human Resources]
program for literally several years afterwards. People would come
up to me and tell me they’d seen my speech, because they used
it as part of their motivational program for their new employees.
That was a good time, and they were—well, they were good vehicles.
They still are good vehicles. They were well done.
Then Endeavour was a replacement and a carbon copy, and it was even
cleaner. I mean, it flowed through and came out. That was after my
time, but Rockwell got a lot of kudos for how well Endeavour came
ahead of schedule. I don’t remember the cost story, but it was
all very successful.
Do you remember if there were any discussions of building a fifth
Orbiter prior to—
Oh, yes. The reason that we were able to do Endeavour in the time
frame we did is because we had a program called the Structural Spares
Program. The thing that takes the longest time when you go to build
an Orbiter is to build all the major structural pieces.
The cabin, for example, is a very elaborate module that’s made
up with complex shapes and complex welds, and it takes like a year
to make it a cabin. The cabin fits inside the outer shell. There’s
a lower and upper shell. It’s what you see, but inside is contained
almost a spacecraft in itself, which is the pressurized cabin. Then
all the other structural elements are also massive. Well, they’re
not all massive, but what I was about to point out is massive, and
that’s the mounts for the main engines. They’re a very
large structure in the aft end to take the main engines, and they
take a long time to create and build.
So during this period that I was the Orbiter Project Manager, we were
concerned about maybe having to have another Orbiter, and no one wanted
to move out to build one, but we did move out to build structural
spares. And so at the time of the Challenger accident, we had these.
These major structural pieces already existed, and all you had to
do was decide to put them together and then outfit the thing and do
all the stuff it takes to make an Orbiter. But the long lead time
things were already in hand.
Interesting. I think I had read somewhere that the Endeavour was one
of something like three government projects that year to come in on
schedule and under budget, so it’s quite an accomplishment.
But we’re about at the time to change out our tapes. So this
might be a good place to take a short break.
All right, we’re back on.
Aldrich: Okay. I was going to mention a couple of other things. I
was head of the Orbiter Project from flight six to flight twenty.
But during that time, [NASA Associate Administrator for Space Flight]
Jim [James A.] Abrahamson decided that our … [numbering] system
was not elaborate enough. So starting with the tenth flight, he caused
us to start using a numbering system which would include the last
number of the year, like 1984 would be four, 1985 would be five, and
then a second number which was the launch site. One would be the Cape,
and two would be the West Coast, and somehow there was maybe going
to be three at some point. I can’t remember how that would work.
And then a letter for the number of flight in the year, like, the
first flight in the year would be “A,” and the second
one would be “B.” So flight ten was not STS-10; it was
STS 41-B. And from that time up through the twenty-fifth flight, which
was the Challenger accident, we had this new numbering system.
Anyway, I was in charge of the Orbiter up either to or through flight
twenty, and then I became the Space Shuttle … [Program] Manager
at Johnson, where I was managing the whole Shuttle system.
Another thing I’d talk about, though, during the Orbiter time
frame was, I talked about how we had used the data from the first
few flights to tune the Orbiter design to be a better match for the
actual flight environment. But the other thing we were also doing
was doing a series of things at Rockwell that are called “load
cycles” for the Shuttle vehicle. A load cycle is an even more
elaborate modeling of the flight environment as we were getting to
know it, where you would operate the vehicle in this math model. You’d
simulate the vehicle with the best data you now had from what its
real flight parameters were, and play it against the environment you
knew that it would fly against, which we were also learning more about,
and then you’d introduce all the variances that you could see.
You know, the vehicle has to have the capability to fly through with
margin in all of the things it does, and so you’ve got a series
of margin parameters.
Each time you do that in a very elaborate way and test all of the
components, you’d come up with a further refinement of the vehicle
capability, and then you could change the software—this is primarily
an ascent problem—you’d change the software parameters
so that you could optimize them, and you’d know what your margins
were and where they were and so forth. We did several load cycles,
which were many-month activities by the Rockwell systems integration
team, and they were quite expensive things to do. I mean, it was a
big project to essentially re-do the entire design with updated known
parameters from the vehicle.
We also would use that information to create the day-of-launch parameters
we’d load in the computer. They’re called I-loads, initialization
loads, and there’s a fairly large number of parameters that
you load into the vehicle right before flight, so that it’s
tuned up for the specific winds of the day. You know exactly what
the environment is it’s going to fly through on the way up,
and you, again, tune the vehicle to have the most margin as it flies
through in terms of angle of attack and what it sees.
So this knowledge of the instrumentation from the first several flights
and these refined load cycles we were doing and this evolution of
the I-load process all occurred during this time frame I was talking
about, and it led to a very mature design over time. But there was
a lot of engineering and analysis to it. It didn’t happen all
in one step. It really kind of evolved from flight to flight and period
So then I took over with STS 51-J in October of 1985. I became the
Shuttle Program Manager, and I was in charge of the Shuttle Program
from there through the Challenger flight and all of the analysis that
followed the Challenger flight.
Then in the fall of 1986, I was asked to come to Washington to be
the Director of the Space Shuttle Program, which was the next phase
in my career, and I moved to Washington and led the program during
the recovery from the Challenger accident and all the changes we made
and all the fixes.
So I don’t know much more to say about that. I’ve been
talking about these jobs that I had and how really wonderful they
were and how much I enjoyed them, and one of the jobs that I enjoyed
the most of all of this was, in spite of the terrible tragedy of the
Challenger, the period that followed that, where I led the return
to flight, was probably as great an opportunity as anything I’ve
I took on an approach to that that I’d seen George [M.] Low
take. I think I mentioned this in the last interview also. After the
Apollo fire, George Low started leading a series of program reviews
with all of the technical people on the program, both NASA and Grumman
and Rockwell, then North American, and looked at everything in the
vehicle, didn’t look at just what had caused the problem, but
he looked at all of the systems and all of the flight performance
and teased out those things that people had some uneasy feelings about
or that were known to be not quite as good as they could be.
After the Challenger accident, I started having reviews like that
with the team here [JSC] and with Rockwell and with the other projects,
the tank and the engines and the solid rocket boosters, and we made
a list of all the things that we were concerned about in terms of
the Space Shuttle systems and what we thought were things we would
want to have changed but we never had the time or the money or it
didn’t fit [the schedule], and also areas where we felt there
was real risk.
We made over 200 changes to the Space Shuttle during the period after
the Challenger accident, and only a handful related to the solid rocket
boosters. Others were these Orbiter systems. They were things with
the main engine, a few things with the tank. We changed a lot of the
software. We added new abort modes. We created new down-range landing
This was an opportunity to look at the whole system and make it as
good as it could be, and I think the Shuttle benefited tremendously
from that for the period that’s followed. There’s been
other improvements since that time, but nothing of that magnitude
has been done since, and it made a much more solid vehicle.
What we found was that [with] all of this pressure to get to first
flight of the Shuttle, there’d been a lot of decisions made
about, you know, “We’ll live with this for now, but we’ll
fix it later,” and so there were a lot of things that troubled
people, that we would have wanted to have fixed, but there was never
time. And then once we started flying, the flights came so quick one
after another, you couldn’t stop and fix anything.
So we got all of this on the table, and we made a lot of changes and
made the Shuttle a lot better, and some of them we required them to
be done for first flight, for the first flight after the Challenger,
and some we allowed over a longer period of time. In fact, some of
the changes to the main engine have literally taken a decade to make.
They were just completed some ten years later. They’ve come
into the upgraded versions of the main engine in terms of the manifold
and some of the other things that were done to it. So it was a time
to take a lot of stock, and it was really good. [Two of the more major
were the development of the new Pratt and Whitney LO2 and LH2 Turbopumps].
I was going to talk a little bit more about the period before Challenger,
and I jumped [forward]. The period before Challenger, we were increasing
the flight rate quite dramatically, and we were also just about to
increase the scope of what the Shuttle could do. The launch site at
Vandenberg Air Force Base [California] was just about completed. We’d
actually had a ribbon-cutting. It was all in place. There was one
concern with trapped hydrogen in one of the ducts under the launch
pad that we were still debating. But other than that, it was ready
to go, and we were planning to launch out of Vandenberg sometime in
the first half of 1986.
So that had been a big piece of work, and in addition to having the
launch site, we also had to have a launch trajectory out of Vandenberg
that would work. Out of Vandenberg you do polar orbits, and so we
were going to fly it straight south over the ocean, past Los Angeles,
and down over the South Pole, and we wanted to have the equivalent
to our trans-Atlantic abort sites on the East Coast. We … [wanted
to] have a trans-Pacific abort site, and we looked at, you know, you
could come all the way around and land in Alaska, but you wanted one
sooner than that. And we looked at Diego Garcia in the Indian Ocean
as a place.
What we finally picked was Easter Island. The Shuttle Program actually
enlarged the runway at Easter Island to be able to take a Space Shuttle.
Of course, we never landed there, but I’m sure that the tourist
industry has benefited greatly from that maneuver while we still thought
we were going out of Vandenberg and when we increased [the runway
size]—in fact, one of the things that bothers me is that all
the time we were doing that, I had multiple opportunities to go down
there, and I would love to have gone to Easter Island, but I kept
putting it off and then it never happened. But the runway happened.
The other thing that was going on at that time, we were working on
two solar system exploration payloads, the Galileo spacecraft and
the Magellan spacecraft, and they needed a high-energy upper stage
to fly their missions, and so we were working on a Shuttle version
of the Centaur upper stage, the liquid-oxygen, liquid-hydrogen upper
stage, that General Dynamics makes out in San Diego. There was an
Atlas-Centaur version and a Titan-Centaur version, and now we were
working on a Shuttle-Centaur version, which the tanks were changed
in shape to fit nicely in the payload bay with these other vehicles.
We were really a long way along with that, too. We were having acceptance
reviews out in San Diego for the Shuttle-Centaur at the same time
we were having readiness reviews at the Vandenberg launch site, and
these two missions, the Galileo and the Magellan, were going to be
also flying during 1986 with the Centaur. That wasn’t quite
as far along as the launch site, the West Coast launch site, because
there was still questions about the safety of the flying the Centaur
in the Shuttle payload bay. We’d kind of committed to fly it,
but we weren’t sure for some of the abort modes, if you came
back with it in the bay, what to do with that problem.
So there were some issues like that still in work, but these things
were built and ready to deliver. It was really just a question of
finding ways to take care of those certain flight phases that we were
still concerned about. So that was a busy time.
Then when the Challenger accident occurred, we elected not to continue
with the Centaur Program. We elected not to continue with Vandenberg,
and we moved into this period I talked about where we made all these
changes to the Space Shuttle to make it a more sound system, in our
What was your personal opinion of the use of the Centaur?
Well, you know, up until the time of the Challenger accident, we were
on a pretty good roll for doing things. We had a lot of confidence,
and I felt like we could pretty much do what we set out to do. So
we were concerned about all these questions we could come up with
about how to operate it, but until the Challenger accident, I think
we thought that was one of the right roles for the Shuttle, to fly
high-energy upper stages and do missions like those.
We actually did those missions, anyway, but we used the solid IUS
[inertial upper] stage, which doesn’t have nearly the energy,
and it caused the flight times to the planets they were going to to
be much longer. But we still did the missions. They just changed the
After the Challenger accident, so many things changed about how we
looked at things, and it was pretty clear that flying the Centaur
was not a very good idea, and we stepped right away from it.
In this pre-Challenger period, what sort of coordination were you
yourself doing with the Department of Defense?
I wasn’t doing a lot of interfacing with the Department of Defense.
I’m trying to think about what our relationships were for working
the Vandenberg issue. But we had our own teams of people there. Primarily
Kennedy Space Center people out at Vandenberg was the government team,
and then there were several additional contractors that were working
out there. But then we had our Shuttle contractors as well. So it
wasn’t a strong interface.
The thing I remember most about the Department of Defense was that
Pete Aldridge was down at the Center [JSC] training to fly, I think,
on the first flight over at Vandenberg, and he was there over in the
crew training facilities a lot, and we saw him.
Did you have much involvement with the couple of Shuttle flights that
were dedicated Department of Defense missions?
Yes. I was cleared for the missions we were flying. But, you know,
we were sort of a delivery system. We didn’t get into what their
missions were going to be afterwards and that sort of thing.
Sure. You know, I find most people still can’t talk about what
their involvement was there, so I understand that.
What were some of the other perhaps significant missions from this
early period prior to Challenger that really stick out in your mind?
Well, one of the missions that sticks out in my mind a lot is STS-9,
John [W.] Young flew, and the thing that sticks out is that I was
talking about the redundancies in the Space Shuttle, in the Orbiter,
and some things we had four of and some thing[s] we had three. One
of the things that we chose to have three of was the auxiliary power
units [APUs], and the auxiliary power units we used for entry to work
the aerosurfaces and they’re the primary means, I think, of—I
guess gravity does the landing gear, but they’re involved in
some way in the landing gear also. I guess the brakes. You have to
have the auxiliary power units for the vehicle to fly the entry phase
and the landing, and you’ve got three of them. You turn them
on in orbit. You kind of do a little check to be sure they’re
all right. You also use them for ascent to control aerosurfaces during
ascent, for [balancing dynamic] loading.
Anyway, we flew STS-9 and brought up the APUs for landing and flew
the entry, and just as the vehicle touched down, two of the APUs failed.
Two of them. At that time, and probably still today, we’re not
certain you can really fly with just one, at least not some of the
most dynamic requirements the vehicle … [ought to see]—so
that was about as close to a very serious incident as anything I can
imagine, that I can remember during that time frame.
It was so close to touchdown, the vehicle just touched down and rolled
out, and there was not much made of it, but we in the program office
and in the technical areas were very concerned that two of these things
could go down at once. Of course, there’s some very high-energy
little components in these APUs that produce so much horsepower from
a small unit, and there was some kind of an injector stem that had
a degrading failure mode that we didn’t know about that had
to do with its lifetime, and the two of them had failed almost at
the same time. We had to redesign that part of the APU, and it was
okay. But if it had happened on the way up and we’d lost two,
we’d been very concerned to try to come all the way down with
one and whether it would have failed also.
I think we also had one flight where we lost several computers for
some reason, but then I think we got one of them back online and I
don’t remember that problem as well.
Then on one of the flights, too, there was an issue with the landing
gear, where they had both blown out a tire and locked up the brakes.
Oh, I had a terrible time during this time period with landing gear
and brakes. The landing phase is a phase that the crew is principally
involved with, and it requires expert performance, and they all train
heavily for it. But they’re all nervous about it because they
know they have to execute, and it’s a very high-energy landing
that comes in steep and fast, and you only get one chance. So the
crews have always been very concerned, I guess, about landing. It’s
something they have to do and they’re skilled to do, but it’s
something they think about a lot.
We were having these problems with the brakes and with the tires,
seeing a lot of damage. The brakes, when you used them, would tear
up to some degree or another on most of the flights, and the tires
were getting damaged. In fact, we changed the surface of the runway
at Kennedy to make it more smooth during the touchdown period to keep
from having such a really rough wear on the tires. That period went
all the way up into this post-Challenger period I’m talking
about. We were trying to find out what you could do to the brakes
to make them better. There were material changes and there were dynamic
changes in the hydraulics that operated the brakes that come from
these APUs, and I think there was another set of changes, also, with
different materials for the whole brake system.
We were still trying to figure out which change to make and which
would solve the problem when we had the Challenger accident. So during
this period where we fixed things, we made all the changes at once,
and when the vehicle flew again, the brakes were fine, but we don’t
know today which change is the real one that was the biggest part
of making it right. It was kind of an interesting thing, but, yes,
every flight had some, just like the tile[s], some degree of brake
damage and a lot of concern because the crews weren’t sure what
the worst thing might happen with these tires and brakes. The one
that was the worst is the one you mentioned, where the tire blew,
but, you know, that one turned out all right, but you didn’t
want to have tires blow during this landing process.
While we’re on the topic of landing, had you had any involvement
at any point with developing the auto-land capability for the Shuttle?
Yes, we developed auto-land early in the program. We thought it was
a mode the vehicle ought to have. In fact, we as an avionics organization
felt it was certified for use starting with flight four in the Shuttle
Program. There … [have] been some changes and refinements since
then, but the vehicle has this auto-land capability in the software,
and it could be used, and it’s been tested and certified to
the point where we think it would work with a high degree of success.
It’s never been used, but it’s there.
How do the flight crews feel about that?
The flight crews, again, it’s one of these aspects of this approach
and landing phase which they’re very concerned that go well
every time, that they have everything that they can at their control
to make sure it goes well, and what they were worried about is not
that the auto-land system wouldn’t fly the vehicle right, what
they’re worried about is if there was some glitch in the auto-land
system right at a critical [point] of approach, and they had to take
control back over, the transient of getting off the auto-land and
getting back into manual control might be something they couldn’t
deal with. So they just didn’t want to commit the vehicle to
this auto-land mode when they were more comfortable having full control
and being on top of the problem all the way down onto the runway,
which I can certainly understand.
It does bring up another thing that I was really interested in during
this time period. We were talking about new launch systems and new
capability during this time period. I’m not sure exactly where
it fits in this sequence of things we’re talking about. For
example, there was a time period over at Marshall where they were
working on a version of the Shuttle called Shuttle C, which was a
tank and solid rockets and then the engines mounted on a cargo carrier,
and that was to fit the kind of thing I’m talking about, where
you could fly cargo missions and you wouldn’t have to have a
What I was really interested in doing was making a few lash-ups in
the software and flying the Space Shuttle without a crew, because
it’s got just about everything you need. So you could fly a
mission all by itself. The whole ascent, it can fly today. You’d
have to fix it so the payload doors could be opened, because right
now the crew opens the payload bay doors and closes them. The vehicle
can do the de-orbit burn. It can fly the landing; it has the auto-land.
The crew drops the main gear with a switch, because that’s the
way it is, but that could be automated. So, with not too many quite
simple changes to the software, you could link all these things together
and you could fly the Space Shuttle just with some amount of ground
control and do a mission and deliver payloads. I was quite interested
in seeing if we could cause that to happen, but it was enough of a
change and it didn’t have the right priority. I never could
find a good enough reason to sell that idea to be done.
What were the advantages of that?
Well, you could fly missions without a crew and you could make it
a cargo delivery vehicle that you would fly. There was a lot of discussion
about why you would risk a Shuttle crew to fly a commercial satellite,
for example. So I don’t know how far we’ll get today,
but during this time period and my later work at NASA Headquarters,
I spent quite a lot of time looking at ways that we could have a different
kind of launch system or other launch systems, and one of the quickest
ways you could do that, if you wanted to, would be to fix the Shuttle
so that you wouldn’t have to have a crew onboard.
Clearly these discussions in terms of value of payloads versus the
risk of the crew’s lives came into great play after the Challenger
After the Challenger accident, we stopped flying missions like that,
because the value proposition didn’t make sense, that you’d
risk a crew to put up a satellite you can put up lots of way[s].
In your position in Shuttle management, what were your thoughts on
having these non-astronauts, like the Teacher in Space or, later,
the Journalist in Space, that kind of thing, flying on the Space Shuttle?
Again, pre-Challenger, this seemed like absolutely a great thing to
do and the right direction to move in. I mean, we were expanding the
use of the Space Shuttle in many ways during that time, and the Challenger
caused a lot of pulling back in. But before Challenger, you know,
we talked about the West Coast launch. We talked about the Centaur.
We talked about the Teacher. Our flight rate was increasing. I mean,
we were viewing the Space Shuttle as just a series of opportunities
yet to be stepped up to in many directions.
Did you have any interaction with Administrator [James M.] Beggs?
Yes, but not very much. I was occasionally at meetings with him, but
I didn’t work [directly for him]. I mean, there were several
people between me and Jim Beggs, and I didn’t go to Washington
too much in those days. This was really a Houston- and Kennedy-centric
kind of activity during the years I was working Orbiter with Rockwell.
Can you explain maybe practically how the levels of management worked
at this time in the Shuttle Program?
Well, the Shuttle Program is made [up] of a series of projects. There’s
the Orbiter Project that I’ve talked a lot about. It’s
a Johnson project, and it really fit the tradition of Johnson in terms
of a [manned] spacecraft development organization.
Marshall had three projects. They had this main engine I talked about
a while ago as being such an elaborate, refined, high-technology machine.
They had the external tank, and they had the solid rocket boosters
and motors. The motor is the solid rocket before you put the controls
on the back. So the booster and the motor are not the same thing.
But when you talk about the booster, you talk about a little more
stuff than what Thiokol makes.
Then there was a launch project at the Kennedy Space Center, and there
were the various other elements here at Johnson, the astronaut training
and the flight simulation and mission control and flight control teams.
They weren’t called a project, but they’re another big
aspect of the organizational structure.
Then the job I had, and Glynn Lunney before me and Bob Thompson before
him, was the Space Shuttle Program Manager, and that was a Johnson
job that integrated all those projects and caused the Shuttle to be
one element as a total set of all those things coming together. Then
that job reported to the Associate Administrator for Space Flight
Let’s see. During the time, John [F.] Yardley had [that job
for] many years, but then—I’m trying to think of the time
frame—… [it] changed to Jess [Jesse W.] Moore after John
Yardley left, and that was just a fairly short period of time before
the Challenger accident. Jess Moore, before him, Yardley, reported
to the Administrator, Jim Beggs, then Jim [James C.] Fletcher. Also
the Center Directors, Chris Kraft, and Bill [William R.] Lucas at
Marshall, reported to Beggs. So Beggs had the Center Directors and
the Associate Administrator. The Associate Administrator had the Shuttle
Program Manager, and then the Shuttle Program Manager had all these
projects across the various NASA Centers. That is the way the organization
was wired together, I guess you would say.
Perhaps we can compare that with the changes made organizationally
after the Challenger accident.
The biggest change that was made after the Challenger accident—well,
there were two changes, two significant changes. One affected the
Shuttle and one affected the Space Station. The change that was made
in the Space Shuttle is the one that I talked briefly about. Early
in the program, the Associate Administrator for Space Flight had had
a Director for the Shuttle Program in Washington. It was Mike Malkin.
At some point in time, he left and that job was never filled. So …
[there] might have been on the org chart, a director, but there wasn’t
a director during most of this time that we’ve been talking
The most major change in the Shuttle Program was the re-creation of
that director job in Washington, and that was the job I was asked
to come to Washington to take, which I did. Dick [Richard H.] Truly
was the new Associate Administrator for Space Flight, and I moved
up to work for Dick and be the Director. The Shuttle Program Manager
job here still existed, and I think it was Dick [Richard H.] Kohrs
became Shuttle Program Manager here at Johnson, still integrating
all these projects at the other Centers. But now there’s this
director that daily, full-time, is managing the program from Washington.
That was one of the recommendations of the Rogers Commission, to strengthen
the Washington control. It essentially moved the control of the programs
to Washington. That’s a change now that’s evolved to a
new era now, where in the recent time frame they’ve moved control
of programs out of Washington and back to the Centers. But after Challenger
that was the recommendation, and I think it was the right one for
the time. There needed to be a lot of attention to everything in the
program, a lot of tightness and pulling together. So that was the
change for Shuttle.
For Station, they also said it ought to be run out of Washington,
and the Station Program, Space Station Freedom, had been working well
for a number of years here at the Johnson Space Center and had the
program office that was in charge and the other centers reported to
it here. The decision that was made was to move the Space Station
Freedom Program to Washington, and so they created a whole new program
office at a new place out in Reston [Virginia]. They had a hard time
attracting key people, because Washington is not a big aerospace-centric
employment area, so they had to bring people in and create a new team.
It was a very difficult and laborious thing to do, but over a couple
of years, the program was moved from the Johnson Space Center to Reston
as the key top program office for Space Station, even though there
were various Station projects still out at the centers.
So those two things came out of Challenger, and actually the change
to Space Station was bigger than the one to Space Shuttle, and they
both happened in the same time frame.
We’ve talked a few times about some of the effects of the Rogers
Commission [Presidential Commission on the Space Shuttle Challenger
Accident] and the Challenger accident, but can you share your memories
of the flight of 51-L?
Well, I was certainly directly involved in all the flights during
that time frame, and I was there in the launch control center, we
launched, and you could see the vehicle up, and it came apart just
within—it was actually at 70,000 feet, but it looked very close,
and we didn’t know what had happened, and it wasn’t for
a couple of days of looking at data and looking at various videos
that we finally realized what had caused it to come apart.
There was a team immediately created there at the Cape to start to
analyze those things, and we met full-time for days to do that. It
was actually video footage from whatever cameras, it was a very clear
day, and so the video coverage was great. When we started getting
those in and looking at them, you could see the flame coming out of
the joint on the solid rocket booster. Then … [when] you more
subtly went back, you could see at liftoff a little puff of black
smoke came out of the same joint.
But, no, I was right in the middle of all of that during that time
Can you describe for us the events of the next few months, your participation
in the investigation and that whole process?
If I can remember. We had this team at the Cape but then after, well,
they created an independent team, and they took it away from the team
that had been responsible for the program up to the time of the launch
and put a new set of teams. I think there was a team at each of the
Centers, as a matter of fact, plus an overall team. It got to be a
much bigger set of people involved. I wasn’t involved in the
downstream analysis. I think those teams worked for nine months maybe.
There was a lengthy period of assessment that went on after Challenger.
What I did was turn directly to this thing I was talking about, which
is, what are we going to do to recover? What has to be fixed? How
do we get the program together? I was working on where we’re
going, and all these other teams were looking deeper and deeper into
the data and they’re finding what the analysis was.
What were your thoughts on the conclusions and recommendations of
the Rogers Commission?
Well, I don’t know if you’ve looked at the Rogers reports,
but I was very involved in the hearings for the Rogers Commission,
and I testified quite a lot at the hearing, and that’s well
reflected in the reports. I think there’s five books, and there’s
a lot commentary from me there, mixed with a lot of other key people,
and I thought their findings were thorough and accurate and well done,
and we tried as hard as we could to work with them to cause them to
be that. I mean, no one was—I don’t know, maybe some people
were negative. I was not negative about the Rogers Commission at all.
In fact, I enjoyed working with Chairman Rogers and several of the
other people that were part of that review.
At what point were you asked to step up into this new role as the
You know, the Rogers report came out in, I think, June. I think that’s
when we first saw what they were going to recommend, and I was busy,
again, working on what we were going to have to do and how we were
going to do it, [and who was going to do it], and working with Dick
Truly by that time. I was still a program manager here; I wasn’t
in the Washington structure.
I can’t remember exactly when Dick moved out to make that—I
don’t think the Rogers Commission change was specific to create
a director in Washington. It said, “Take control up from the
Center and bring it into the Headquarters,” and so I didn’t
know what change that was going to be or how it would play out.
I think it was like three, four months later that Dick Truly called
and said he wanted to come down and talk to me, and he flew down to
Ellington Air Force Base [Houston, Texas], by himself, on the NASA
plane, and I met him out there, and he told me that I could either
take that job, or if not, he was going to have to fill it with somebody
else. I wasn’t at that time really thinking I wanted to leave
the Johnson Space Center and leave Houston, and so I didn’t
give him a really positive answer. Within a day or so, I could see,
you know, I was either going to be in that job or I was going to work
for someone who was in it, and so I went up there, and I said, “Yeah,
I want to take that. That’s what I want to do.”
Had you worked for Dick Truly much prior to that point?
Yes, I’d worked with Dick over the years. That takes me all
the way back to Skylab, when Bob Crippen and Dick Truly were kind
of new green astronauts. They were assigned to some of the Skylab
systems in the … [Orbital] workshop out at Huntington Beach
[California], and I was involved in Skylab in that time frame. I remember
going to several reviews out at Huntington Beach, where they were
there, and they were working on the things that we were working on
as a project. So I got to know them in that time frame, and so I knew
him quite well.
Obviously he must have had great faith in your ability, asking you
to step up into this position.
Well, it’s always hard for me to sound immodest, but I think
based on all this Shuttle work I’ve talked about, I probably
knew as much about the Shuttle system as any one individual that might
be available to him at that time. I knew the people. I knew the programs.
I knew the mission operation side of things, and I knew these-I’ve
explained how the whole Shuttle system is flown by the avionics, and
I was a principal in developing that. Then … [I’d] been
head of the Orbiter Project for a number of flights, and the Orbiter’s
really the most elaborate part of the Shuttle, and so I think I was
probably a very useful person at that time, and I did have a good
relationship with Dick, so it worked out.
How did you go about consolidating this level of management at Headquarters?
Well, I started doing all of the reviews that I had done here [JSC]
as the Shuttle Program Manager. Since the time I had become the Shuttle
Program Manager, I was head of the Shuttle Configuration Control Board,
and so I just changed, and I just managed it from there [Headquarters].
I took charge of the Shuttle Control Board. We had a daily noon meeting
on the Space Shuttle, every day. They probably still have it, and
I ran that from Washington. I had two secretaries. I had one in Washington,
and I had one here [JSC], and I would come back. I’d spend like
three days in Washington and two days here for a long period of time.
So Dick and I were up there, and we were totally in charge of everything
that was going [on]. It was being run from Washington, but I was really
still also operating with all of the scope and interfaces here at
Johnson that I’d been involved with for many years.
How did the rest of the Shuttle Program respond to this change?
I don’t think it was a big change for them. I mean, I was the
same person, and I was doing most of the same things. Whether I was
on the phone from Washington or on the phone from Houston was a little
subtle. There were changes in personnel at each of the projects, and
the new people were of the same motivation I was, you know, “We’ve
got to solve this terrible problem and make things right and get back
to where we need to be with the Shuttle Program,” and so it
was a very cooperative, energetic team that came together to do that.
I didn’t have difficulty having a tight—and we would meet
face to face once a month somewhere as a Shuttle Program management
team also. In addition to having the weekly change boards and the
daily telecoms, we would come together.
I remember the first meeting we came together after the accident,
and we’d moved out of this mode where we were just analyzing
what happened, and we were now starting on building, putting the project
back together, we met at Marshall, which was a very emotional meeting.
But everyone was wanting to go forward and ready to do whatever it
I had the opportunity to speak with J.R. Thompson at some length about
the changes he made at the Marshall Space Flight Center and the kind
of attitude going on there. This is the same idea as you’re
The people that were most impacted by the Challenger, other than the
family members themselves, were the people at the Johnson Space Center.
They were literally devastated for a long, long period of time. It
was more here than anywhere else.
Why do you think that was?
Well, I think it was partly because these crew members were our friends
and neighbors, part of the Johnson family, and I think because we
felt here like we were entrusted with their well-being, and we’d
let them down. It was a very emotional time, as you can see.
But completely understandable, though.
And the people here at Johnson had that attitude I was talking about
earlier, about we had been on such a roll. We really thought we knew
how to do things. We knew how to do things right. We could do just
about anything we decided we wanted to set into. This was a big shock.
Did you find the technical changes necessary to getting the Shuttle
system where you wanted, these things that you had mentioned a little
bit earlier, were more or less difficult than the organizational or
attitude type of changes that needed to be made?
Well, I didn’t mean to strongly indicate attitude changes were
a problem. I mean, everyone was so impacted by this, and then there
were certain personnel changes that got made, and when the team finally
came back together, there was no problem getting the team to work
together and the team to pull in the right direction. It was a unified
effort right from the start.
The changes that had to be made to the vehicle, some of them were
hard changes to make, and some of them we wanted to make before first
flight. Some of them we just wanted to make and get them when we could
get them, like these—I can’t remember the several changes
to the main engine. But if you think about the last ten years, changes
have come online in the main engine with new turbo-pumps for oxygen
and hydrogen, and they’re made by Pratt & Whitney. They’re
not made by Rockwell, Rocketdyne. That change was instigated at this
time. It took years to develop these new pumps and then certify them
with the engines and get them delivered.
Also, the manifold on the engine had some number of ducts leading
into the thrust chamber. You either went from three to two or two
to three, I can’t remember which. That change took half a dozen
years to get developed and get certified, and it’s come into
the main engine. So we talk now about main engine upgrades, and they
happened over the last four, five years. But those are changes we
approved during this post-Challenger time frame, and it took that
long to develop them.
Others, the number one concern on our list, other than doing whatever
had to be done to the solid rocket booster, is there’s a seventeen-inch
valve in the line that goes from the external tank into the Orbiter
for the oxygen and another one for the fuel [hydrogen], and that valve
is open at launch and then when you separate from the tank, it closes,
so that you have a … closed-[off] place. You don’t have
an open duct. There was great concern that that valve might flap closed
during the boost phase, while the propellants were flowing through
it. It was a valve that was moved open with a small amount of tension,
and then the dynamic flow of the propellant going through it would
kind of hold it open, but there was no latch. It was just open and
flow. It was a little hard to rig for each flight. It had to be set
up and rigged so that it held open with a certain amount of spring
tension. So the whole thing seemed a lot less certain that it was
going to perform correctly than we’d like to have it.
We went to look back at the design to see why it was we were so certain
the flow would always work in such a way the valve was held open until
all the propellant had gone, and then you could close it with the
closing system. Rockwell couldn’t really produce the certification
in any depth that would give us any more confidence. This maybe …
[wasn’t] a risky thing, and the fellow that designed it had
died. So we ran a major test program to see if we could build some
certification data, and there was enough uncertainty in that that
we said, “We’re going to fix this thing.” It was
the number one … [change on] our list other than solid rockets,
and we were going to do two things. For the first flight, we were
going to put a little latch in so that it not only was held open,
but there was a little latch that kept it open. Then for later flights,
we were actually going to design it to a thing that took the flapper
right out of the system and it was going to be a whole new valve.
Well, we did build the latching system, and it’s on the seventeen-inch
tank shutoff valves today. But the new one was going to be a fourteen-inch
valve with a thing that moved out of the way. We couldn’t build
it so it would pass certification and be effective. So the change
we made for first flight worked and is in, but the more elaborate
change we couldn’t develop, and so we’ve stuck with the
one that—we did do that change for first flight.
I didn’t have to worry with the changes … [to] the solid
rocket booster, because the Rogers Commission created several review
teams to work forward of their efforts, and one of the teams they
created was the one for the redesign of the solid rocket boosters,
and they were charged both with overseeing the redesign and approving
it. So if whatever design change NASA was going to make that NASA
said was okay, we still couldn’t go forward unless they also
said they were satisfied. So it was a very tight process. There were
expert people on that review, and we had to do all the things they
thought were necessary before we could say the solid rocket booster
could fly again. [This team was led by Guy Stever and Jim Mar of MIT
was a team member.]
If you’ve talked to anybody about that, we kind of did a belt-and-suspenders
set of changes on the solid rocket booster. We changed the O-ring
material and the way—maybe we didn’t change the O-ring
material. We changed the way the O-rings can be tested so that you
could verify that they were redundant. We changed the way they fit
in the metal overlapping parts of the booster cases, so that when
you pressurized the booster, they would close on the O-ring rather
than tend to relax them, which the old system would do.
Then we filled in the flow through the … [liner] material on
the inside of the case. In the old design there was a gap there so
that if there was a path through the O-rings, it could go right through
the … [liner] material, and this [change] caused the lining
material to overlap. It caused the joint to pressurize to close, rather
than pressurize to open, and it changed our ability to test to be
sure both O-rings were perfect and good.
Then we put a heater system on so they would never get cold. We felt
we had to make all those changes. Probably just to change the way
the joint worked would have been enough reliability to expect that
it would work, but we wouldn’t stop there, and it was probably
driven because we had these expert people overseeing what we were
doing, what Thiokol was doing. So that got done.
The things that I was working on changing, like these changes to the
main engine and changes to the seventeen-inch valves, and those are
changes to other parts of the system that, I mean, we could decide
to do it or not do it, but we really did just about all of the ones
that we came up with. We either changed them for first flight or changed
them sometime during the next few years.
What about the re-investigations of crew escape systems?
Well, that was another thing. You know, there was a lot of speculation
that when the Orbiter broke up, that the cabin pressure maybe was
still intact until the cabin hit the water. So the crew might have
had an opportunity to do something if there was something they could
have done. One of the recommendations of the Rogers Commission, I
believe, certainly one of the recommendations we took away was that
we had to have some kind of an approach to give the crew some action
they could take if that kind of a situation occurred again.
We looked at something that we looked at a number of times before,
and that is, the first four, I think, flights of Columbia had just
two crewmen and they had ejection seats and they could come out through
the roof. But when we went to increase the number of crewmen, there
was no way to mount that many ejection seats effectively so they could
do that job and you’d have still a cabin you could use.
But on several occasions we looked at whether there was some clever
way to put enough ejection seats, or if you could have some kind of
an ejection rocket pack that a crew could wear and they could go out
two at a time through the holes that already exists. The windows …
[in] the roof pop out, and you can go out through the roof. We tried
hard to come up with some way we could provide for crew escape that
would get the crewmen out, and it never seemed feasible that a workable
design could be created.
But now we’d had the Challenger accident, and we had an event
[where] it really might have been good to have some way to get people
out. So the first thing we did was change the hatch so you could blow
it from the inside. You couldn’t open the hatch in the way the
Orbiter was before the Challenger accident. Now the crew could pyrotechnically
open that hatch if they needed to. We looked at how you get the crew
out, and what we found is that if you open the hatch and the crew
jumped out, they’re going to hit the back end of the Orbiter,
one of the tail surfaces, more times than not. So they can’t
just jump out.
We, again, talked about having some kind of a rocket pack that they
could wear, that … would propel them out. But no one really
wanted to fly every flight with seven rockets in the cabin. So what
we did, someone over here in the JSC Engineering Directorate came
up with this idea of this pole that once you open the hatch, you could
extend this pole out and you hook a ring on it, just like when you
parachute from an airplane, but you slide the ring down the pole to
the end, and then when you fall off, you’re going to miss the
back end of the Orbiter, and you can open your parachute.
So that was what was concluded to be the best compromise of providing
the capability, but not tearing the thing up so bad that you caused
other problems. I don’t think the crews particularly like it,
but it’s better than not having an option. That is what the
configuration of the Orbiter is today for getting the crew out [in
Were there any other significant changes than the ones you’ve
talked about to this point that were either examined and discarded
or in some other way perhaps saved for a later point in time?
Well, let’s see. Of course, one of the big changes, it got to
be a job I had a little later, for a long time, there was a lot of
emotion about building a new solid rocket booster instead of fixing
the one we had, even though these fixes from an engineering sense
were quite a thorough approach. A lot of people wanted a new solid
rocket booster. In fact, J.R. Thompson was a very strong proponent
of that. So we instigated this program called the Advanced Solid Rocket
Motor, and it was a contract to Lockheed as the integrator, even though
they’re not a solid rocket maker, and then Aerojet [General
Corp.] was the company that was going to build the rocket itself,
We were going to have a great new plant over in Mississippi, just
across the line from Alabama. It’s about an hour’s drive
from the Marshall Center in Iuka, Mississippi. That plant was mostly
built, created. Making a solid rocket motor plant with all of the
things you’d like is a big undertaking. I mean, we were going
to pour these things in pits, and there were big protection things
that if there was an accident, personnel would always be safe. It
was a very elaborate plant, and it was just about completed, and we
were about to start manufacturing advanced solid rocket motors when
we finally got back to flight with the redesigned solid rocket motor,
and you could see it was going to perform well.
So from the time we started to fly again, the support for this very
elaborate and expensive advance[d] solid rocket motor started to wane,
and it waned and waned and waned to the point where finally in one
of the budget years, it was decided not to proceed.
Since you’ve mentioned the return to flight, can you share your
memories of STS-26 with us?
In what [manner]?
The preparation, the building up to that flight, and then the actual
Well, I’m not sure what you want me to talk about.
Your personal feelings on that and what you were doing at the time,
that kind of thing, of the actual mission itself.
Well, the actual mission was—I mean, I’ve already talked
about it. It didn’t happen.
No, I mean STS-26.
Yes, the return to flight.
I thought you were talking about the Challenger.
STS-26, you know, we literally worked for years on these changes we’ve
been talking about, and worked very hard because some of them were
tough to do, and they took a lot of intensity in terms of reviews
and oversight. For a while, it seemed like we’d never get there,
you know, because part of what Dick Truly’s response to this
Rogers Commission and all of the other emotions for this is we were
going to do everything with total caution and total care, and everything
was going to be perfect.
We wrote down a lot of guidelines and ground rules about what it would
take to make this an absolutely the safest flight we could make in
terms of what the launch criteria would be. We just set about to do
it as perfectly as we could. So it was a long period of time. I think
it was two and a half years, something like that. I could probably
tell from this.
Right around thirty months, I think.
Yes, two and a half years, from January of ’86 to September
of ’88. But then we finally got it all at the Cape, and we got
it on the launch pad, and we got it checked out. The night it was
supposed to launch, we were in the [launch] control center, and we
had a tremendous thunderstorm come through at one or two in the morning,
and we thought, “You know, the weather’s probably not
to make it.” So all of the things it took to get to that point,
we still aren’t going to get to go. At about five in the morning,
it cleared, and all of a sudden you could tell that all of the stuff
we’d worked on so long was all coming together and was all going
to go, and it did. It was absolutely incredible, absolutely incredible.
Then we had this long period. I talked about the conservatism that
we set for ourselves, to have everything as careful and right down
the middle as possible. But we caused the whole Shuttle Program to
be very conservative, and then for years afterwards, after the time
I left the program, there was this tremendous feeling of caution and
concern and oversight and checks and balances, to the point where
the team had a very hard time getting to each flight, because no one
wanted anything that looked at all like it might have some issue with
it to go untouched. We really built in a tremendous amount of conservatism
that I think over some number of years returned to a proper balance
for that, but we kind of were overdriven for a long time in terms
of everyone being so cautious and conservative.
Some of the programmatic changes in terms of dropping the Department
of Defense flights, not doing the commercial satellites, these kinds
of things that we talked before, kind of give the Shuttle a different
focus on, like, these science missions and that kind of thing. What
impact did that have on things like the flight rate and the preparations
for flight, I guess, from a management standpoint?
Well, I’m not sure how to answer that. Some of the changes we
made during the down period and flowing out of the down period, I
believe we built another Orbiter preparation facility at the Cape
so you could process three Orbiters instead of two at once. We upgrade[d]
another crawler so that we would have three crawlers instead of two.
We got a second 747 aircraft so we could ferry. I mean, we were thinking
in that time frame, post-Challenger, after we got back and going,
that we could make a flight rate of fourteen flights a year, and some
of these changes we made to these preparation facilities and support
facilities were tuned to what you would really need to do that. I
think the launch pads had to be, not upgraded, but they had big maintenance
re-work on the launch pads, and one had been done and one hadn’t.
So we got both launch pads up. We got more Orbiter preparation capability.
We got more crawlers. We got this carrier aircraft; they have a pair
We did something with the solid rocket booster facility, but that
might not have been for flight rate so much as doing the assembly
in a more careful way. I can’t remember what we did there. It
was a new building of some magnitude.
So we were thinking that the Shuttle [flight] rate would build, and
I don’t think we were thinking we’d move back into commercial
satellite launches. Nobody thought that was a very good idea. I don’t
think the DOD wanted to be invited back. They wanted to [be] off on
But we still thought the Shuttle, the scope of what it could do and
would do and NASA would want to do, would require these kind of changes
in flight rate. Then over time, and this conservatism I talked about
in terms of being even more careful with testing and checks and anybody’s
concerns, we never really built the flight rate up close to that magnitude
again, and for a variety of reasons, it’s actually less now
than it was then, partly because, I think, of what the mission requirements
are right now. It certainly could fly a bigger flight rate than we’re
flying here in 2002 if that’s what the nation wanted to do with
Before you moved on to your next position, what sort of interaction
did you have with the Space Station Program?
In the Shuttle period I was in, I really didn’t have any interaction
with them, because we weren’t close enough yet that the two
programs had to come together. Dick Truly had both programs under
him, and—no, he didn’t. Dick Truly had the Shuttle, and
there was another program office, another Associate Administrator,
[Andrew Stofan], that had the Space Station Freedom. One of Dick’s
frustrations during that time period was, you did need to work Shuttle
and Station together at his level, and he had a co-equal. So some
of the things that he felt ought to be done, he couldn’t always
make happen the way he wanted to make them happen.
When he got to be Administrator, he changed that. He brought in Bill
[William B.] Lenoir and gave Bill Lenoir both the Station and the
Shuttle, all under the Associate Administrator for Space Flight. So
it was the reverse of what I started to say. They really were independent
during this time frame of Shuttle recovery. But then later Dick combined
them so that one person would have more opportunity to make the two
programs fit in a way that was good.
Why did you then move on from heading the Shuttle Program?
Well, after we returned to flight, I was the Shuttle Program Director
for six flights after—STS-26 was the first one and then five
more. That was the time period when Dick brought in Bill Lenoir and
put both the Shuttle and the Station under Bill, and I was really
of a mind that I would have liked to have been the head of one of
those two programs, but they both went under Bill, and so I thought
it would be time maybe to move on to do something else.
I talked to both J.R. and Dick about it, and Dick was looking for
someone to be the Associate Administrator for the Office of Aeronautics
and Space Technology, was the name. It’s had a number of names,
including a couple while I was there. But it was a long-term office
at NASA Headquarters that had been the Headquarters oversight for
the aeronautics program for a number of years, and some of the advanced
technology for space also was in it. So I was asked if I’d like
to do that, and it sounded like something I’d like to do.
So I moved to that office for the next three years. That was 1989
through ’91. It wasn’t quite three years, I was Associate
Administrator for Aeronautics and Space Technology. What that work
entailed was primarily the aeronautics programs … [in] NASA.
Now, you’ve heard me talk about my whole career. I hadn’t
done anything with aeronautics until this point. But I knew a lot
of the people, and a lot of work in aeronautics is not that different
from space. We were doing several really interesting programs at that
One of the programs I worked on was what we called the High-Speed
Civil Transport Program. That’s sort of code for a supersonic
aircraft, like the Concorde. We were building one back at the time
of the Concorde—“we” United States—and for
several technical reasons, we didn’t proceed. One of them was
that the engine effluent from a high-speed supersonic transport contributes
to deterioration of the ozone layer, and the other was that the sonic
boom is a problem flying over land, and if you don’t fly the
plane [supersonically] over land, you’d only gain half the advantage
of the mission of having a supersonic aircraft. So it was considerations
like that in the decades prior we stopped working on supersonic transport
in the United States.
But when I took over the Aeronautics and Space Technology Office,
there was a new program that was called the High-Speed Civil Transport,
which was a supersonic transport for commercial airlines, and it was
working directly on those two problems I talked about: how can we
deal with the sonic boom, and how can we deal with creating an engine
that won’t damage the environment. It was a great program. It
was being done in conjunction with Boeing, and that was in place all
the time I was there, these several years I was in the office.
But after that time we, again, moved away from doing a supersonic
transport, and that program came to an end. It might reappear. Right
now if you read the papers, I read that Boeing’s working on
a high-speed jetliner that’s almost supersonic with some different
configurations. So the idea’s still around, but this program
I was working on was quite an expensive program, and I think people
didn’t want to go at the end and actually create it.
Another one I was working on in that office is the National Aerospace
Plane Program [NASP] with the Department of Defense, and the NASP
is kind of like a single-stage-to-orbit spacecraft except it is built
on the concept of being air-breathing. So the NASP sits on the runway,
and it collects air and makes liquid oxygen out of it to get enough
to take off, and then as it starts to fly in the ramjet and in the
scramjet mode, it gets its oxygen out of the atmosphere. This was
thought in the mid-eighties on into the early nineties as being perhaps
the way the next space transport vehicle would be, and had a lot of
support and enthusiasm, but, again, the technical problems caught
up with it. It was very hard to do, and it got to be expensive, and
the DOD [stopped] wanting to fund it, and so at the time I was in
the office, we were working on it, but it also ended.
Another task that Dick Truly asked me to take on in that office was-this
was the period following President [George H. W.] Bush’s commitment
to the Moon-Mars Project, and I was in charge of the planning we were
doing and the program for that, out of this Aeronautics and Space
Technology [Office]. It’s got a name.
Space Exploration Initiative.
Space Exploration Initiative. So that was in this office as well,
and we didn’t get very far beyond the planning … [with]
that, but it did involve quite a lot of interaction with organizations
and people, and during that time frame I was going around speaking
to groups quite a bit about going to Mars and why and what it meant
and how we’d do it. So I’ve got a little stack of speeches
at home that we built during that time that were supportive of going
to Mars. It seemed like a great thing, but it was unaffordable.
The other thing that that Associate Administrator has the responsibility
is the reporting channel for the aeronautics centers of NASA. So I
was in charge of Langley Research Center [Hampton, Virginia] and Lewis,
now Glenn, Research Center [Cleveland, Ohio], and Ames Research Center
[Mountain View, California], and all their facilities, and that all
flowed up through me to the Administrator, who was Dick Truly. So
… [not a] small part of the job was that interaction, and I
spent quite a lot of time working with the Centers on their plans,
their programs, their budgets and so forth.
What kind of priority was being given to aeronautics at this time?
Well, I was going to say something about that, and then I didn’t
branch down into it, but these two programs, I particularly mentioned
the NASP and High-Speed Civil Transport, were big expensive programs,
and the Aeronautics Program had a big budget, and these were kind
of far-reaching advanced vehicle programs. They weren’t just
doing technology; they were working towards a program that would lead
to a vehicle.
There was also a lot of other aeronautics programs as well. There
was one to improve the civil transports we have today in terms of
making them more efficient and quieter. There was a program with the
FAA [Federal Aviation Administration] on improving the FAA’s
capabilities and roles in the national aerospace system. The FAA doesn’t
really have a research capability like NASA has, and so we had this
cooperative arrangement where some of the programs FAA wanted to do
were being done at Langley and being done at Ames. It was through
my office. So it was quite a robust program in those years. It’s
been sometime after that when NASA started this reduction in funding
for aeronautics and not featuring program[s] of this type so much
In fact, during this period also, we were building at least one wind
tunnel, and I think maybe making plans to upgrade several others.
We were investing in more wind tunnel capability, a new tunnel at
Ames, and I can’t remember the name of it now. So it was a robust
aeronautics time frame, and it was good for me because it was all
interesting and a whole bunch of new things.
One of the things that happened during this time frame was the United
States Air Force decided to no longer keep the SR-71 fleet, and so
they shut down the SR-71 program. But they gave NASA, which is my
office, three of them, one two-seater and two single-seaters, and
a whole building full of spare parts, and a whole tank farm full of
the special fuel it flies. So we were the curators of the SR-71 for
those years I was there, and then sometime after that, the Air Force
reactivated it again, and I think they’ve deactivated, but we
still have out at Dryden these SR-71s that came at that time frame.
I think it’s been a while … [since] you could have said
that NASA’s aeronautics program was robust. But actually this
may be a good time to pause again since we’re running short
Okay, that’s good for me.
So, to summarize, I really enjoyed this time working in the aeronautics
program, even though, as I say, it wasn’t really my background.
I’m an electrical engineer; I’m not an aeronautical engineer.
The things I found there that I managed were things I could relate
to. I thought it was a productive time for me to be there, and I certainly
enjoyed it a lot.
But in 1991 there was a new office created at NASA Headquarters called
the Office of Space Systems Development, OSSD, and I was moved to
be the Associate Administrator for Space Systems Development. What
Space Systems Development had under it for responsibilities was Space
Station Freedom, which now, of course, was all up in Headquarters.
Dick Kohrs was the Program Director for Space Station Freedom, and
he was at Headquarters, and the whole program team was out in Reston.
It finally had been pulled together and was making pretty good progress.
So that this new office had Space Station Freedom. It had the advanced
solid rocket motor … [that] we talked about…. It was still
a program that was going on.
It had a thing that was called the National Launch System. It was
a joint program with DOD and involved a lot of the launch industry
working on technology improvements in various components in launch
vehicles to create a new expendable launch vehicle for use by the
DOD and commercially to replace the current expendable launch vehicles
that were going on. So we had quite a few NASA programs in support
of that through Marshall and other places, and that was in the office,
and then a series of smaller technology projects of various kinds
across the Centers that were working space technology.
What made this a really interesting job for me, though, had to do
with the re-engagement of my career with the Soviets, now Russians,
that occurred during this time frame. In the Space Station Freedom
program, Dick Kohrs was a very competent manager, and he was right
there, and I didn’t really need to go in and over-manage him
or manage past him in his program, so I didn’t get heavily engaged
with managing Freedom, other than it’s just where he was part
of this office that I had.
But there was one project that I did take on that interested me, and
it came about in the following way. This was just after the time when
the Berlin Wall came down and the Soviet Union came down. There got
to be quite a lot of interest in the United States in terms of engaging
or exploiting or taking advantage of what we could do with the Russians
that might be of benefit to us.
During the same time frame, the Space Station Freedom Program was
wanting to create a crew rescue vehicle, and they had initial contracts
with both Lockheed and Rockwell to study what we called at that time
an ACRV, Advance[d] Crew [Return] Vehicle. But we had the same kind
of problems with the Congress at that time frame that we’ve
had in the recent time frame, and that is that nobody really wanted
to give us any money to do it. They would give Space Station Freedom
the money it needed to do Freedom, but [when] you said, “I want
to have a crew rescue vehicle, and that’s another so many hundred
million dollars,” it would never get funded. So we were struggling
with—I mean, we thought we ought to have a rescue vehicle for
Freedom. We were struggling with how to get one.
Barbara [A.] Mikulski [D-MD], who was the head of our Senate authorization
committee, called a hearing in the fall of 1991, and she called Dick
Truly and she called Yuri Semenov, who was the head of NPO Energia
in Russia in Moscow, and she called a high-level official from the
State Department. I think she asked the number one guy, but he sent
the guy directly under him. The purpose of that hearing, and Semenov
came over and we had the hearing over in Congress, over in the Russell
Building, and her question was, why is it we couldn’t use the
Soyuz to be a rescue vehicle for Space Station Freedom.
Of course, Semenov thought it was a great idea, and Truly said, “You
know, okay, we probably can do that.” The State Department guys
said, “Well, you know, I don’t know.” So she kind
of lashed out at State and why … [they] wouldn’t let us
start working this program. So that happened without directly connecting
I think I talked earlier today, and I think also in the last session,
about how after Apollo-Soyuz we stopped working with the Russians,
but the medical team kept working. The life sciences team kept exchanging
data and analyzing human performance in space. So we had a group there
at NASA Headquarters that had maintained a contact with the Russian
people, and the head of that group was Sam Keller. Sam was in the
science organization at NASA Headquarters. I can’t remember
what his title was at that time. But he had a lot of current experience
and engagement with the Soviets, now Russians.
We had this hearing with Mikulski in the fall, September, October
sometime, and in January I got a call from Sam Keller. He said he’s
been asked to go over to Russia and meet with Semenov and his team
and talk about this Soyuz thing, and it occurred to him with my background
in Apollo-Soyuz and with the Russians, that I’d be a great guy
to be part of his team. So I thought it was a great idea, too, and
what I did was get a little team of key people in the various system
areas from the Johnson Space Center that I knew, people that knew
about the propulsion systems and knew about the life support systems,
and so forth, somebody from the crew. I’m not sure the first
trip we had somebody from the crew, but we added a crew group to it
later. We took about seven or eight people. We went over, and we had
a meeting in Moscow.
Now, during Apollo-Soyuz when we would go and meet with the Russians
in Moscow, with the Soviets in Moscow, we would always [meet] in some
rented facility downtown, and they would say they were from the “Soviet
Academy of Sciences.” That’s who we thought they were.
We didn’t know where they actually made things. Even when we
sent our docking system over to be tested with theirs in a test facility,
it was some rented facility downtown that we went to.
Well, the day we arrived in 1992, Sam Keller and I and this team of
Johnson spacecraft engineers, we were driven straight out to Kaliningrad,
to this place, huge factory, huge brick wall around it, great gate,
through the gate and up and right into the office that used to be
Sergei Korolev. But these people that we went and met there are all
the same people we worked with back in the early seventies, the same
people, but now we know they’re NPO Energia. We know where they
hang out, what their place is, and we started these discussions about
what you could do with the Soyuz.
It became obvious to us almost immediately that there’s no reason,
if you fix the docking mechanism, you couldn’t use the Soyuz
for a rescue vehicle. But we didn’t treat it lightly. We reviewed
each system and what their criteria was and [what] we’d have
to do and how it would work. It was a several-day review over there
The only problem we found that would be a limitation was that the
Soyuz, like our capsules, has one propulsion system when it’s
in orbit, but when they drop the service module part off and just
bring in the reentry part, it has another little attitude control
system, and that attitude control system on the reentry part of the
Soyuz uses a monopropellant, hydrazine. Hydrazine you can only store
for so long, and it starts to deteriorate, particularly if there’s
any kind of contamination at all. Over some period of time you may
not have as good hydrazine as you want to have. So their limit on
the length of time they could … [leave a] Soyuz in space was
six months. So we were making quite a lot of progress on how we could
change that system to put in a bi-propellant or some other propellant
and make the Soyuz stay longer. I’ll talk some more about that,
and we never made that change.
The current Soyuz that’s the rescue vehicle on the International
Space Station is an unmodified Soyuz and has to come home every six
months to be replaced. So we didn’t get to make that change.
There weren’t any other significant changes to be made to the
We had a series of meetings over the next couple of years to work
the details of providing this rescue vehicle, and it was very successful.
But some other interesting things happened during this time frame
also. When we were there, now we’re at Energia … [in]
their big factory buildings and things, and they took us down to this
mockup area, and in the mockup area was a Buran spacecraft sitting
there, and it’s a mockup of a Buran, but it’s a full-size
Buran. Up until that time period, we were still debating what kind
of a mechanism we were going to use to dock the Shuttle to the Space
Station Freedom. There were various concepts about what to do. The
Shuttle didn’t have a docking capability in its early years.
It wasn’t until the Space Station came along that it needed
one, and that design hadn’t been firmed up.
Well, these Russian[s], Yuri Semenov, Viktor [P.] Legostayev, took
us down and showed us this Buran. The payload bay is configured exactly
to a Shuttle payload bay, the same kind of rails, the same dimensions,
the same latches. So any payload that could have flown on the Shuttle
could fit in the Buran bay. And on those rails there’s this
airlock sitting, attached to the back of crew cabin, pointed up at
a docking mechanism on top, exactly what the Shuttle needed to dock
with the Space Station, except we hadn’t committed to do it
yet. They’ve got it built. It’s sitting there.
So this was also the same time period when Dan [Daniel S.] Goldin
replaced Dick Truly, and this was probably, I don’t know, the
second or third trip over. They didn’t show us this the first
time, but I think after we started talking to them more and talking
beyond Soyuz about other things that they might contribute to a Space
Station program, they decided [to show] this to us. So I had pictures
When we got back, Dan Goldin had been moved in. I wasn’t reporting
to Truly anymore, and I went in and told Goldin. I said, “Look
at this thing that they’ve got.”
Well, that caused a great uproar of activity, and Rockwell got sent
over to look at it and see why it is we couldn’t use theirs
[Energia’s]. Well, I think what Rockwell decided was that it
was too heavy or they could make one that was a better fit rather
than just take the one the Russians already had. But they created
exactly the same thing, exactly the same thing, and that led to the
docking mechanism that went on the Shuttle that was used—the
airlock tunnel and the mechanism-that was to Mir, Shuttle-Mir, and
now for the International Space Station, and it came directly out
of the NPO Energia mockup area. We … [might have picked] the
same thing after we decided, but we were having these technical debates
about different ways to do that and, you know, there was no forcing
function. We hadn’t decided which way we were going to go, and
this forced it. So that was a very interesting thing.
We had other discussions during that time period with Semenov and
the team about other things that they might have that could work on
Station. But I didn’t even get to the point where that happened,
because we redesigned the whole thing [Space Station] after a short
Another job I had in this Office of Space Systems Development was,
we were still talking about wanting another launch system. I talked
a little while about Shuttle C would be a way to lift very heavy cargoes
and use the Shuttle. We’d talked about ways you could upgrade
the Shuttle itself, and we were talking about other things.
We wanted to have a different launch system for two reasons. One reason
was that we were already seeing that with the expense that the Shuttle
system is and the expense the Station was going to be, too much of
the NASA budget was encumbered in these [two] programs, and it was
limiting what NASA could do. So we really wanted a new system that
was a lot less expensive to deliver payloads to orbit, and we set
as a goal a ten-time reduction in the cost of a pound of weight to
orbit. I had those discussions with Dan Goldin and with Mike Griffin,
who was now in our office and I was working with him, not in my office
but in Goldin’s set of managers.
So I got asked to lead a study to define an alternate launch system
to the Shuttle that would have the same capabilities of performance
but would be ten times less cost of payload to orbit. I put together
three teams. I put together two teams at Marshall. One of them looked
at these expendable launch vehicles that were in use and this new
National Launch System we were talking about building, and their role
was to define what you might create with that type of hardware. You
might put a little airplane or a cargo [carrier], a little spacecraft
or a capsule on top, but you’d use these expendable launch vehicles
or maybe they’d be partly recoverable, partly reusable.
The second team was to look at a single vehicle that was like the
NASP, but was rocketed and [didn’t] have this air-breathing
thing to it, and it was called a single-stage-to-orbit vehicle. That
was a team at Marshall also.
The third team was here at Johnson, and they were to look at the Space
Shuttle and what changes you could make to it to have a ten-time reduction
in cost of launching a payload.
These teams worked for three or four months. They were big teams.
They worked these things in a lot of depth. They worked not only the
different configurations you might have in those three categories,
but what the costs would be and then what the end performance would
be. They all came together with their pieces of this study, and we
produced a fairly elaborate report that’s got the three studies
in it and the recommendations.
The recommendations were that either going down the Shuttle path,
you can improve a lot of things but you can’t make it ten times
cheaper, and with these expendable vehicles the same thing. Too much
of that gets thrown away and has to be replaced. You don’t meet
the financial goal that way either. Even if you have two fully reusable
pieces in a two-stage vehicle, you don’t get the financial return.
The way you get the return is this single-stage-to-orbit vehicle,
and we recommended NASA move out with the program to develop a single-stage-to-orbit
vehicle. That’s what the report says. The report’s still
around. People used it up until a few years ago. Now we’ve gone
past that time frame.
So this is interesting because it has another interconnection with
these Russians. One of the things that we were doing to make the single-stage-to-orbit
concept work was to have a different kind of engine. Of course, it’d
be a reusable engine, but people that work on launch systems and propellants
often endorse the idea that a liquid oxygen-kerosene engine is good
for a first stage, performing down in the atmosphere, and a liquid
oxygen-liquid hydrogen [engine] is good for an upper stage that performs
in space. So here we had this little single-stage-to-orbit vehicle,
and we wanted to have liquid oxygen-liquid hydrogen, but that’s
not optimum for the first phase of the flight.
So we were talking about building an engine that was a tri-propellant
engine, and this single engine would take kerosene during the early
time period, and then it would blend and flow with liquid hydrogen
and use liquid hydrogen on the upper stage.
We want to do a test program with that, so at some point we got into
a discussion of the Russian technology in engines. What they’ve
actually done over these years that we were so enamored with the Space
Shuttle engine and we went to oxygen-hydrogen, they’ve continued
to build more modern versions of kerosene-oxygen engines, and they’re
very good, very high-performing engines, and they have a capability
that we had not perfected in the United States, which is to burn oxygen-rich.
I can’t tell you in a propulsion sense why that’s better,
but it creates a higher-performing engine, and it’s hard to
do because if it’s oxygen-rich, you have the risk of an explosion
if you don’t do it right, and they had developed materials and
coatings that allow them to run oxygen-rich and have these very excellent
The company that builds those engines is called NPO Energomash, and
they’re in Khimky, Russia, just outside of Moscow, and they
built a modern version of these high-performing kerosene engines for
the Energia vehicle that flew the Buran and also could fly with other
payloads. You could fly either the Buran as a cargo or a cargo package.
There’s four strap-ons on the Energia, and each strap-on has
… [one] of these RD-170 kerosene-oxygen engines that are new
and modern and high-performing, that are built at Energomash.
Well, somehow we got onto the thread of talking to NPO Energomash
about using their technology to create one of these tri-propellant
engines where we can do kerosene and we could do hydrogen, so we went
over to visit and talk to them about that, and with Marshall, and
we actually got a little program framed up about doing a test program
with a tri-propellant engine. In fact, they had a full-size mockup
of what it would be, and they also had a test article under way.
We did that in conjunction with Pratt & Whitney because at some
earlier time frame Pratt had gotten over there after the Soviet Union
had ceased to exist and had worked a deal with Energomash that they
would have all rights on this Energomash RD-170 engine. So if somebody
else wanted it, they’d have to buy from Pratt & Whitney.
So we had this little team go over. We had Pratt, we had Marshall,
we had a few people in my office there at Headquarters, and we were
working this tri-propellant engine with Energomash.
However, given my national launch system job, I’d also been
talking to all of the rocket and rocket engine contractors in the
United States about what they had and what they could do [to create
a lower cost, higher performing system]. One of the people that I’d
had in and we’d talked to were the Atlas people, General Dynamics
in San Diego. I’d had them in, and we had talked about re-engining
the Atlas, because it didn’t have an optimum engine system.
It had a 1950s configuration engine system, and you could easily see
ways that you could upgrade the Atlas and the Atlas-Centaur to make
it a better performing vehicle, and I was looking for things we could
do in all camps.
Well, when I got over to Energomash to talk about the tri-propellant,
we also talked about their RD-170 as it exists, and we got into these
discussions about why you couldn’t put—the RD-170 has
four valves with a single thrust chamber, and we got into these discussions
why you couldn’t—it’s too big to go under an Atlas,
but you could cut it in half and you could put half an RD-170, which
became an RD-180, under the Atlas, and it’s the perfect engine,
perfect engine for that vehicle. That didn’t happen during my
time frame here, nor did we get to the end of the tri-propellant program.
But that caused me to get really re-engaged with Energomash, as well
The other thing we saw down in the mockup area at Energia was one
of these RD-170 engines that Energomash had made, and Energia was
making the electromechanical actuators on it, and they were running
simulations in their lab with this big engine system and their actuators,
and so I got all wrapped up with both of those companies and these
propulsion initiatives, these transportation initiatives we were working.
So this was also the time frame I was meeting a lot with Dan Goldin.
I found this docking thing and brought it to him, but he was also
wanting to make changes to the Space Station Freedom. Initially he
didn’t like the big solar arrays, and he wanted to make something
that looked more streamlined, and so we looked at a lot of concepts.
There was a one concept where we could have made a station out of
Orbiter, just modified it in certain ways and it could have stayed
up for some period of time. I was working these things as head of
this Space Systems Development Office. But he also had other people
he was calling in to look Station designs. Langley had some designs.
Almost from the beginning he was wanting to change Freedom and maybe
the Clinton administration wanted to change Freedom anyway. It may
not have been that he was so technically unhappy about it. But that’s
one of the things … [we] started right on.
I had some ideas that were in my head. One of them was that I thought
the Space Station should fly at a higher inclination. The Freedom
flight plan was to fly due east out of the Cape at twenty-eight and
a half degrees [inclination]. So it would fly over things that are
as high a latitude as Florida but no higher. A number of the launch
sites in the other parts of the world are at much higher inclinations.
It makes it very hard for them come down to a Freedom orbit. So if
you wanted to launch [to] the Space Station Freedom, say, from Baikonur,
it would take a very large vehicle and be very inefficient. The other
launch [sites] in China and Japan, the same story. We didn’t
know at that time, things might fly out of there. But it seemed to
me with the Space Station, it was international, it was going to be
permanent, it would make much [more] sense to be accessible from launch
sites anywhere in the world and to overfly much more of the populated
part of the world on a regular basis.
So I … [recommended] to Goldin that he change the inclination
of the Space Station to 51.6 [51.2 degrees], I think is what Baikonur’s
at, and, in fact, I don’t know who else might have been recommending
it to him, but he made that change. Created a big problem. The Shuttle
is most efficient, or any launch vehicle is, when you fly due east,
and if you fly at a higher inclination, it takes more propellant to
deliver the same cargo. So [we] now … [had] created a deal where
the Space Shuttle, as it existed, couldn’t fly the missions
to Space Station with the heaviest cargoes, because it was now going
to be at 51.6 [51.2].
So the next task I got was how could we upgrade the Space Shuttle
for more performance. We looked at additional engines, re-engining
schemes in some regard, putting engines under the tank. Out of all
those studies, there was one change that was most clear that it was
something we could do and something that the program could accommodate,
and it would give us an additional eight or ten thousand pounds [payload
capability], which is just about what we lost on the inclination change,
and that was to change the material that the external tank is made
out of, and that’s what I recommended to Dan, to change it from
aluminum to aluminum-lithium, which is a metal that’s hard to
work and we didn’t have huge amounts of experience with it in
Marshall had done some studies on how to weld it and how to repair
it and things you have to do when you build the tank, and over a six-month
period of time, Marshall got comfortable enough to say, yes, they
could buy into that. So that’s what I recommended to Goldin,
and that’s what we’ve done. The tank is now aluminum-lithium
for most flights, and now I don’t know if they still make any
that are aluminum, but they could.
So that change … [was] something that happened in this time
frame also, while I was working with all these propulsion people.
My report that looked at replacements for the Shuttle said single-stage-to-orbit
is the way to go, and Marshall studied that for another year after
I left. But that report led to the instigation of the X-33 out in
Palmdale [California] with Lockheed that has since—it didn’t
make it, but it was the evolution of all the thinking that was going
on during this time frame I’m talking about.
Two other stories about Russia during this time frame. By coincidence,
I work for Lockheed Martin now, and these two stories relate to this
time frame in Russia and Lockheed Martin.
While I was over there during these visits talking about making the
Soyuz a rescue vehicle, I met an old friend of mine in the hotel lobby
that we stayed in, the Radisson Hotel. It’s called the Slavjanskaya,
and it was a good place in that time frame in Moscow to stay. This
fellow Frank Martin that used to work for NASA up at Goddard was now
a Lockheed employee out in Sunnyvale, I met him the lobby at this
hotel in Russia, and he said he’d been asked to come over and
look at their Proton production facility at Khrunichev, another company
in Russia…. So we talked about that, and he was going the next
Well, the next day when he got back, his eyes were this big. He said,
“I went to this factory, and they had all these huge rockets,
and they’re just out there. They just build them out there in
this factory, and there were cats running up and down the aisles,
but they had these tremendous launch vehicles.” So he went home
and reported that in, and it was not many months later that Lockheed
Martin formed this International Launch Services company with Khrunichev
to market the Proton internationally. That’s now one of the
principal launch vehicles that’s in the Lockheed Martin suite.
The International Launch Services offers both Proton and Atlas launch
vehicles, and if one of them gets delayed for some reason, you can
go on either one. So that became an initiative that Lockheed Martin
happened to capture right in the same time frame I’m talking
The second one is, I told you about General Dynamics recognizing that
the Atlas could be upgraded to be a more effective vehicle. Well,
in the next year or so after that, Lockheed Martin bought the Atlas
business from General Dynamics in San Diego and moved it to Denver.
They started manufacturing Atlas vehicles in Denver, and not long
after that, the Department of Defense decided to create two new heavy-lift
launch vehicles, one with Boeing and one with Lockheed. The Boeing
vehicle is called the Delta IV, and the Lockheed vehicle is called
the Atlas V, and what the Atlas V is, is an upgraded version of the
Atlas. The old Atlas has a very flexible set of tanks. They can’t
stand under their own weight unless they’re pressurized. The
new Atlas V has a rigid tank structure so it can be transported horizontally.
It can be raised vertically, and it doesn’t have to have a pressurization
system with it all the time.
But the big change is the engine on the Atlas V. It is half of the
RD-170 Energomash engine that we were over there a few years before
talking about the ideal fit, to put it on the Atlas through Pratt
& Whitney, and Lockheed has put it on the Atlas, it’s going
to have its first flight here within a month, and they got the RD-180
engine through Pratt & Whitney, which is built in Energomash.
So the last three years of my NASA career was heavily re-engaged with
the former Soviet Union, the Russians, the ones that I had known from
years past, and it was directly related to the technical interest
areas that I was trying work in the United States in any regard, and
it was a really very interesting period. I enjoyed it tremendously,
doing all that.
It sounds like you had quite a number of different projects going
on at that time.
Well, you know, there’s only half a dozen Associate Administrators
at NASA Headquarters, so each one has a series of major programs under
them, and I didn’t do all this work myself, but I pulled on
mostly Marshall and Johnson in this time frame and the aeronautic
centers in the previous job I had. But you can get down into the details
quite nicely, and we certainly did with going back to Russia. We’d
gotten a lot going on.
You asked me over the phone about my role in Crystal City when we
redesigned the International Space Station, and I really didn’t
participate in that. All of these thing[s] I’ve just talked
about are the last things I worked on, and I was still there when
some of this redesign at Crystal City started, but there was a whole
new team of people. My office was left to keep Space Station Freedom
on track and going until we decided to move to something different.
So I didn’t really get engaged in the redesign.
Interesting that you got the Space Station Program without the Space
Shuttle Program, whereas previously you said that they had consolidated
those two under a single Associate Administrator.
They did, but then they broke them out again. It’s a lot like
this thing about whether you ought to manage programs from Washington
or from Centers. You know, there’s good arguments to both and
whichever one looks like it’s not working very well, well, they’ll
say, “Well, the other one will fix it,” and that goes
Were there any other projects from that time period before you left
NASA that you want to cover before we wrap things up today?
I don’t think so. I think I’ve just about hit all of this.
Oh, I did want to say a little more about the X-33. After I turned
in that report that recommended that we try to build a single-stage
… orbit vehicle, NASA studied it for another year in terms of
what it could be and how feasible was it. I was gone, but then they
moved into instigating the X-33 project. I was working at Lockheed,
and Lockheed happened to win the X-33 program, and so my very next
job was, I was moved back to our headquarters for Lockheed Martin,
which is in Washington, and the director of the X-33 Program was T.K.
Mattingly, who had been brought in by Lockheed to run that program,
and T.K. asked me to come and be his deputy on the X-33 program, and
I spent the next year and a half working on the X-33 for Lockheed,
out at Palmdale and back, which was a direct flow-out from these things
that had begun in the early nineties that I was talking about.
It’s only, I guess, somewhat relatively recently that that program
came to a conclusion that wasn’t perhaps, I guess, the intended
Well, it had technical problems and it had funding and political problems.
The first few months that T.K. and I were working on that program,
we went out and reviewed it, and our conclusion [was] that it never
did start with enough funding to do all the things it had to do, and
we predicted many of the difficulties that did happen later. We predicted
that that’s kind of the way it would flow, because while it
sounded like a lot of money, it was a billion-dollar program, building
a brand-new single-stage-to-orbit launch vehicle was bigger than a
billion-dollar problem, even you’re just building a prototype.
I’m sure you discovered that with the National Aerospace Plane,
Well, the same problem. Some of these things, we have grand ideas,
but we want to fund them in a more digestible manner than probably
what it takes.
I’m looking over my list of questions, and we’ve covered
most of the topics that I had planned for today. I did want to check
with Sandra and Rebecca if they had any questions for you that they’ve
come up with.
I think you’ve covered my questions.
Okay. Looking back on your career as a whole, are there any particular
moments or specific jobs that you really recall as the highlights
of all your years at NASA?
These jobs were all so marvelous that, you know, you couldn’t
downplay any one of them. I tell people I worked for NASA for thirty-five
years and I enjoyed it every single minute. It was really wonderful.
The ones that I feel like I made the most contribution to are these
Shuttle years that I talked about today. I was so heavily involved
with what the Shuttle was and how it came to be and the things that
went on with it, that I just had a great experience with that.
We appreciate your contribution to the space program and to our project
here over the last interview and this one. But I want to give you
the opportunity to make any final remarks before we close today.
No, I think that, what I just said, is probably my final remark.
Okay. Thank you very much.