ACCIDENT INVESTIGATION BOARD
WEDNESDAY, APRIL 23, 2003
3000 NASA Road 1
BOARD MEMBERS PRESENT:
Admiral Hal Gehman
Rear Admiral Stephen Turcotte
Major General John Barry
Major General Ken Hess
Dr. John Logsdon
Dr. Jim Hallock
Dr. Sheila Widnall
Dr. Milton Silveira
Mr. George Jeffs
Mr. Owen Morris
Mr. Aaron Cohen
Mr. Robert F. Thompson
ADM. GEHMAN: Good morning. The Columbia Accident Investigation
Board public hearing is in session. Today and this afternoon,
we're going to deal with various types of risks. We're going
to listen to a number of experts and talk about their view
of risk management and risk mitigation and how risk is looked
at from about five different angles, particularly as it applies
to manned space flight and the shuttle program.
This morning we're going to look at risk as it applies to
the original design and construction of the STS. Later this
afternoon, we're going to look at risk from the point of view
of experts on aging aircraft. We have a couple of experts
going to testify and talk to us about how risk migrates over
a period of time as aircraft are used. Then later in the day,
we'll have Professor Diane Vaughan who will talk about organizations
and how organizations deal with risky enterprises.
For this morning, the board is very fortunate to have a wonderful
panel with years and years, maybe decades and decades of experience
in this particular enterprise, the STS system. The Columbia
Accident Investigation Board would like to thank the NASA
Alumni League for organizing this panel -- and a special thanks
to Norm Chaffee, the president of the Johnson Space Center
chapter of the league -- for helping us to arrange this panel
that we have in front of us.
What I'm going to ask, Panel Members, is if you would, first
of all, go right down the row in some order or another and
introduce yourselves and including in your introduction, if
you would, say a word or two about the official position you
had when you were involved in either the Johnson Space Center
or the STS or shuttle program when you were actively engaged
in running it. Then when you're finished with that, I would
invite you all to make any kind of an opening statement that
you would like to make; and then we'll proceed into questions.
So if I could ask you to start at one end or another there,
and maybe with Aaron there, and introduce yourself, including
a little background of your involvement in the space transportation
AARON COHEN, ROBERT THOMPSON, GEORGE JEFFS, OWEN MORRIS
and MILTON SILVEIRA testified as follows:
MR. COHEN: Good morning. Thank you. My name is Aaron
Cohen and I was the first NASA space shuttle orbiter project
manager from 1972 to 1982. This period of time encompassed
the design, development, and the first four flights of Columbia.
I retired as the Johnson Space Center director in 1993 and
I taught at the Texas A&M University from 1993 until 2001.
I am now professor emeritus of engineering at Texas A&M.
During this period of 1972 to 1982, there were many design
challenges on the various subsystems and the integration of
the subsystems into the basic vehicle. This included the structure
system, the life support system, the environmental control
system, the thermal protection system, which were the tiles
and the carbon material, the thermal seals, the avionics system,
the auxilliary propulsion system, the hydraulic system, and
the many mechanical systems such as doors, actuators and tires.
I would like to say that we have a very good documentation
of this activity, and it was prepared in 1993. It was a compilation
of papers presented at a conference held at the Johnson Space
Center in June 28th to 30th of 1993. This documents the design
challenges of all the shuttle systems. The papers were prepared
by the NASA and contractors' subsystem managers, and the subsystem
managers were the backbone of the shuttle design.
This is my introduction statement. I will be happy to answer
your questions in the hopes that we will be able to return
the shuttle soon to safe flight.
ADM. GEHMAN: Thank you very much.
MR. THOMPSON: Okay. My name is Bob Thompson. My principal
reason for being here today, I was the shuttle program manager
from 1970 to 1981. That encompasses a time that we started
into what we called Phase B, the very early design activities
on the shuttle; and I remained the program manager through
the first orbiter flight, at which time I retired and went
to work in industry.
I'll be happy to answer any questions. I think certainly the
subject of risk management, I think we all recognize that
any vehicle that can fly to and from earth orbit is going
to be a risky vehicle by definition. So you're going to have
to deal with risk. I don't care how you design it. Of course,
the way you determine that you want to design it really sets
in the family problems you're going to have to deal with;
and it's very important in the early design phase to pick
the set of problems you're going to want to have to live with.
I think we were extremely conscious of that when we picked
the configuration that we picked, and we knew we had a lot
of problems to deal with. As long as we continue to fly the
shuttle, we'll have to have problems to deal with. So I'll
be happy to answer any of your questions as we go on through
MR. JEFFS: I'm George Jeffs. I've spent since the Sixties
in the space business, most of it with NASA, a lot of it with
the Air Force also. I was at one time the chief engineer of
the Apollo program, the program manager of the Apollo program.
I was the Apollo program manager and the shuttle program manager
at the same time for a while. I ran the space division that
also had the global positioning satellites. The Rocketdyne
division reported to me. The energy activities reported to
me at Rockwell. I ended up running that part of Rockwell that
was sold to Boeing.
I've enjoyed working on the space program with the NASA because
we have thought alike. We have been after the basic cause
of problems rather than Band-Aiding problems. We've left no
rock unturned to try and get the right answer to these things,
mutually. We may have missed a few, but they were unknown
to us or we would have fixed them. All those years I have
spent in the middle but between NASA and industry and making
those teams work because the teams are just as important a
part of making these big programs happen as the hardware itself.
I find myself again in the middle here, with NASA fine people
on both sides of me, a thorn amongst roses; but at any rate,
I will try and also answer any of the questions you might
have that we may recall the answers to. We're all very proud
of the hardware and its performance. Some of the best memories
that I have are the astronauts telling us, after flights,
what beautiful hardware it was to operate. Thank you.
ADM. GEHMAN: Thank you, sir.
MR. MORRIS: My name is Owen Morris. I was with NASA
throughout the Apollo program and worked on the space shuttle
from 1972 to 1980. Initially I worked with Aaron as his assistant
orbiter manager, and then later I was in charge of systems
integration at the Level 2 of the program. I worked with Bob
Thompson there from late 1972 to 1980, retired in 1980, and
then formed a company of my own for the next 15 years, working
on conceptual design. I'm very happy to be here and look forward
to answering your questions.
ADM. GEHMAN: Thank you, sir.
DR. SILVEIRA: Hi. I'm Milton Silveira. I first became
involved with the shuttle in March of '69, before we landed
on the moon. I was involved in Phase A studies; but even prior
to that, I was involved in the design of the systems, support
systems on Mercury, Gemini, and Apollo. I went through the
Phase B studies; and when we started into the hardware studies,
I moved from running a shuttle office in engineering and development
over to become Aaron's deputy as orbiter project manager.
I was involved with the shuttle up until about '80, when I
moved to headquarters to become NASA chief engineer. I retired
from NASA in '87, after 36 years with NASA.
I currently serve as a technical adviser to Lieutenant General
Ron Kadish in the Missile Defense Agency. I'm glad to be here
and hope we can help you.
ADM. GEHMAN: Thank you very much. Did you all get to
make any opening statements that you would like to make before,
basically? Okay. That's fine.
Okay. What we'll do is start a round of questioning here and
I'll go first and then I'll open to any one of my panel members.
I'll address my question -- and all of us will follow this
procedure. We'll address our question to somebody, but I hope
that any of you who wants to piggyback on the reply or elaborate
or anything will please feel free. We would love to have two
or three answers to the same question because you all approach
this thing from slightly different angles. Some of you were
more intermittently involved with systems and some of you
were more project manager and integration related. So I'll
start the first question.
Mr. Thompson -- and others, too -- I notice that in addition
to being involved in the STS system in the Seventies, which
was in the program design definition phase, that you had previous
experience in Gemini and Apollo also. Could you in any way
contrast the engineering development, the project managership,
the rules under which you operated of those two systems? Is
it possible to draw for us any differences or similarities
between those two systems? And then I would invite anybody
else that would like to comment on that.
MR. THOMPSON: Well, I would give you a broad, general,
off-the-top answer. I think the processes and procedures and
the management approaches and techniques were better in shuttle
than they were in either of the two programs previously, mainly
because we in government and we in industry had matured a
good deal by working through those programs. For example,
all through Mercury, Gemini, Apollo, Skylab, we kept a "Lessons
Learned" document. 8086 or something. I can't remember the
number. I think it was the 8086 document, and we made the
8086 document an applicable document on the shuttle program.
Let me pick a specific example. We lost a main propulsion
test article during the shuttle development period because
we used the wrong weld wire in a critical weld joint. That
wrong weld wire came about because the vendor had mixed two
metals on the weld wire reel. We had learned in an earlier
program that, in any critical welds, you ought to test the
weld wire you're actually using before you make the critical
weld. We missed that early in the shuttle program. We came
back and corrected it, but that lesson learned came out of
the previous programs and fed on into the later programs.
So that's just one of many, many, many examples I could cite
and I think, frankly, both the government management team
and the contractor management team was more experienced and
probably was able to take on the shuttle design and development
job and in many respects the shuttle design and development
job was considerably more difficult than Mercury and Gemini
and probably more difficult than any single element of the
Apollo program. So I think I would say that we were better
prepared to manage and develop a critical risk program in
shuttle than we were previously.
MR. COHEN: I'd like to add my comment. It's almost
the same as Bob's but maybe a little different emphasis. I
was on the Apollo program. I wound up being the manager of
the command and service module on Apollo. The heritage we
had from Apollo was a very strong subsystem manager concept,
both at the government and at the contractor. It turned out
to be a very, very good system. Our subsystem managers, in
all honesty, were not peak ticketed, so to speak, to the program
office. They actually worked for the head of the engineering
directorate, which was Max Faget at the time, but the subsystem
managers essentially did do their daily work for the project
office and there was a very good check and balance. They had
a very good relationship with their counterparts at Rockwell
or at Grumman or in the Apollo program, but in the shuttle
program at Rockwell.
There was just a very good check and balance in the system.
I felt very comfortable with that because if there was a disagreement,
the subsystem manager could always go to Max and Max could
then go to Chris, who was the center director, or Bob, and
we could resolve the issue. So I felt that that was a heritage
from the Apollo program that made it very good.
MR. THOMPSON: While we're on this subject, let me make
another point that I would like to call to the board's attention.
At the time we were moving into Phase B on the space shuttle
program, we still had not decided what configuration to build.
So the Phase B management was still led out of Washington
with almost identical management roles at Johnson Space Center
and the Marshall Space Center because it had not developed
exactly what vehicle we were going to build. Once we got to
the end of Phase B and it became apparent the vehicle we were
going to build, we went into a somewhat new management structure
for NASA, which set up a program manager at what we called
If you aren't aware of it you need to understand what Level
1 was in shuttle, what Level 2 was, and what Level 3 was.
The agency, NASA, and within the manned space flight, decided
to set up a Level 2 program manager having agency-wide responsibility
for the design, development of the vehicle but to locate that
individual institutionally at the Johnson Space Center so
that he could take advantage of all the institutional resources,
but he did not have any program per se responsibility to center
director. He had, of course, a desire to keep the center director
informed, but he did not responsibly report to the center
director. He reported directly to Level 1 in Washington; but
in working in Houston, then you had to work across two other
centers to work the other project elements.
In addition to the subsystem managers that were set up within
the project elements, one of the key things that I feel that
we set up to manage across the program were what I call ten
key technical panels. We picked a key NASA individual to chair
those panels, and we made those ten key technical panels all
report into Owen Morris' office that was part of my Level
2 program office. Those key technical panels then had membership
put on those panels of experts all around the country at other
NASA centers, within contractors, within universities; and
those technical panels worked specific technical issues that
cut across the total vehicle. They reported in to Owen and
then any issues came from there to my control board and I
had the responsibility to sign off or approve or implement
the things that came out of that integration process.
If that process has been allowed to weaken, I would be very
concerned because that's the heart and soul of working issues
across the vehicle of a technical nature. For example, if
insulation is coming off the tank, the tank project manager
cannot approve that. He cannot allow that to happen. That
violates a systems-level spec. He has to come to the program
manager at Level 2 and ask the program manager to approve
a bunch of insulation coming off the tank. If the system isn't
working that way and if the Problem Report And Corrective
Action procedure is not working and if the program is not
bringing the collective intelligence to deal with those kind
of problems that you do if you work through the system properly,
then you've got a problem in the program and you need to fix
ADM. GEHMAN: Let me follow up on that. I don't want
to hog the microphone here. So I'll let my panel get a word
in here edgewise. For me to understand the chain of command,
did any of you work for the chief engineer at JSC?
DR. SILVEIRA: For the chief engineer at JSC? In reality,
although he did not have that title, Max Faget, who ran engineering
and development, was basically our chef engineer; and, yes,
I was on his staff during the Apollo program.
ADM. GEHMAN: During the Apollo program. What about
DR. SILVEIRA: During the shuttle program, we started
out that same way, yes, sir, until I became Aaron's deputy.
ADM. GEHMAN: To get to Mr. Thompson's point then, as
I understand this -- and I'm beyond my level of expertise
here. If you were trying to resolve an engineering program
-- of course, that's all you did for ten years was resolve
engineering problems -- but the engineering section or the
engineering division, would you describe for me the checks
and balances between a fix, an engineering solution that Mr.
Faget had responsibility for, versus either the shuttle integration
office or the shuttle program manager?
DR. SILVEIRA: Well, probably our biggest disputes were
always between operations and engineering as to what operations
wanted and what engineering was capable of doing. I think,
in general, the thing is, you know, we as a team had been
working all through the Apollo program together and I think
as a team we realized that we were all friends, we knew each
other, we knew who to go to, and we knew how to resolve any
issues we had. And we usually, you know, came to a compatible
solution as a result, without having to be dictated to as
far as what approach we ought to use.
ADM. GEHMAN: The point I'm trying to get at -- and
thank you for that answer. The point I'm trying to get at
is: Would it be incorrect for me to characterize Mr. Max Faget's
role as being essentially an equal to the program manager?
DR. SILVEIRA: Yes, sir.
ADM. GEHMAN: That is correct.
MR. THOMPSON: I don't understand why you would use
the word "equal." No, Max Faget could not make a within-the-program
ADM. GEHMAN: I understand that.
MR. THOMPSON: He could come to me and make his wishes
known. He could come to my control board and argue until we
got to midnight, pro and con. If he did not like what I did,
he could go to the center director, who could go to my boss
in Washington and straighten me out; but when it came time
to decide who made the decision, there was no doubt who made
the decision and who was responsible for it.
DR. SILVEIRA: But there were few decisions that went
MR. JEFFS: You need to put this in the right perspective,
too. The majority of people worked for the contractor. On
Apollo we had 40,000 people on Apollo. We worked for these
guys, but those guys worked for us. On shuttle we had up to
20,000 people. So you've got a whole engineering structure,
both in the contractors' level and the different contractors
with the subcontractors. So those technical issues were being
massaged with great care, and they were being interfaced with
the NASA so that we had a team working. But the drawings came
out of the contractor. The detailed decisions on how to do
things on change control within the contract were done with
the contractor. So you've got to look at both these things
together to see who's making the decisions and how they're
MR. THOMPSON: And you have to really be a little more
specific. Ask us any detail you want and we can tell you how
that would be managed and handled. For example, if it was
a stress-level issue down in designing what an allowable stress
somewhere internal to a wing, you'd have to go deep into the
contractor organization and check that work to really find
out whether it was pro or con. And the subsystem managers
in the government actually checked that work, not number by
number, but looked at the procedures used, looked at the decisions
made, looked at the allowables and the materials and this
sort of thing. But now if you ask who's responsible for not
having an abort system on the vehicle, you have to ask me
that question. You cannot ask George Jeff or you cannot ask
Milt Silveira that question.
MR. JEFFS: But if you would ask who, why it didn't
work, then you can ask George Jeffs. (Laughter)
MR. THOMPSON: Well, if it didn't work, it's a combination
of the government and the contractors.
MR. MORRIS: Yeah, I think, getting back to how decisions
were made, we probably ought to talk about the Change Board
that Bob Thompson chaired. That board was made up of all of
the element managers. The orbiter was Aaron Cohen. The tank,
the boosters, the engine. Reliability. Max Faget sat in on
that board. He was a bona fide member of the board. Operations
was a member of the board. And there was no significant decision
made that that board did not understand. Now, as one of the
program managers in Apollo once said, you know, "The board
is here and this is a Democratic organization but I have 51
percent of the vote."
MR. THOMPSON: But there was never a significant decision
made in the shuttle program that Max Faget didn't have plenty
of opportunity to sit in my board while we were discussing
it, make his wishes known as many times as he wanted to, and
he knew exactly why I made the decision I made. Whether I
agreed with him or not, he knew why and he knew and by the
next day I had signed off on the decision and written up why
it was made.
MR. COHEN: Let me hitchhike on one more thing. The
orbiter also had a Change Control Board, and on that board
we had Rockwell sit in on the board, we had a contractor sit
in on the board, and we had each directorate, like Gene Krantz
from Flight Operations, George Eddie from Flight Crew, Max,
and R&QA and so forth. So we also had a board. Now, if it
went outside our envelope boundary, then we would take it
to Level 2; but if it was inside, then we make the decision.
MR. THOMPSON: And you can say the same thing for the
other project elements -- the tank or the engine or the SRBs.
MR. JEFFS: As Bob says, the other elements, whether
it's the SSMEs or the orbiters, these are engineering focus
operations. The engineering is the head of the snake. So engineering
had a key voice in almost every decision that was made down
the line on these programs. And a free voice.
DR. SILVEIRA: And I think, importantly, the heritage
of the organization, most of us came out of the Langley Research
Center and we moved to the Manned Spacecraft Center when it
came down to Houston. So we had a heritage of working together.
We knew each other, and we respected each other. Once we arrived
at a decision, everybody supported. There was no hassling
afterwards. We were sort of really, in looking at a lot of
organizations today, we were sort of unique in that regard,
in being able to work together and make decisions together.
MR. THOMPSON: You never strive for 100 percent agreement.
If you get 100 percent agreement, there's something wrong.
ADM. GEHMAN: Right, you're missing something.
MR. JEFFS: I'd like to add one more thing I mentioned
earlier, and that is the issue of organization and developing
organizations. I was fortunate to have, with the Apollo program,
a source of great depth of capability of people, experienced
people. They came from the aircraft areas. They came from
P-51s. They came from SMJs. They came from across the board
on how to build aircraft. A great base.
That base was trimmed and kind of honed during the Apollo
program. That same base fortunately was maintained on the
shuttle program. Trimmed and maintained. So we had not only
the same kind of people but the same people, the same procedures
had been smoothed. The knowledge of what each element could
do and couldn't do within the organization and between ourselves
and NASA was understood. That doesn't exist to the same extent,
as I see it, in these different companies today, probably
because a lot of these people are gone and you can't put everything
in the data base. You've got to have with the people. So there
MR. THOMPSON: George just read part of his proposal
for the contract.
DR. LOGSDON: I want to go back to the period of '69
through January of '72. At the policy level, the decision
whether to approve the shuttle was being debated; and you
folks at the engineering and management level were getting,
I think, changing signals of what kind of shuttle was going
to be politically acceptable. I guess the question is, Bob,
you said you started as shuttle program manager in '70 and,
Milt, you said you were involved in the Phase A studies. Phase
A studies produced a particular concept, a fully-reusable
straight-wing shuttle. So first question: Did that first design
have the large payload base, the 15-by-65 payload bay?
MR. THOMPSON: The answer to that is yes; and the answer
to what came out of Phase A, what came out of Phase A, those
of us that were given the responsibility to go implement the
program felt that that was a very dumb way to go about it.
The two-stage fully-reusable system, as we looked at it in
detail about going to build it, a lot of people argued that
politics made us change it; that is absolutely not correct.
We changed from that vehicle because we found, as we dug into
it, that was not a very smart way to go about the job, for
many, many reasons. I could spend half a day here explaining
it all to you, but the concept that politically we wanted
to build a two-stage fully-reusable vehicle but couldn't afford
it, that is not correct. The vehicle we built is the vehicle
that the NASA people that came into the program starting in
Phase B that had the responsibility for building it, we built
the vehicle that we wanted to build, not the one that the
politicians told us we had to build.
DR. LOGSDON: Fair enough. In 1970, a new set of requirements,
I believe, appeared in terms of what was required to get the
Department of Defense support for the program -- with additional
cross range, I guess, being the most important of those new
requirements. Tell me if I'm wrong, that that had a link to
shift from a straight-wing to a delta-wing configuration.
MR. THOMPSON: You want me to answer that?
DR. SILVEIRA: Let me make some comments on that, John.
Of course, you know, a few of us got cleared on what the Air
Force programs were; and once we understood what the Air Force
requirements were, then we understood how that affected the
design and changed over to meet those requirements.
MR. THOMPSON: I'm not sure I would agree with that.
I think the myth that the straight-wing two-stage fully-reusable
orbiter was a good system to build is strictly a myth. You
don't want any wing on the orbiter while you launch it, and
the only benefit of the straight wing is in the terminal approach
and landing phase. The fact that what Max was proposing was
to hold that straight-wing vehicle up above the stall level
all the way down to 10,000 feet above the runway, then whip
it over and land it on the runway and to carry those straight
wings all the way to orbit and back, and to have a fly-back
booster, that whole system crumbled when you began to look
NASA did not put cross range in the vehicle because the Air
Force forced us to. NASA put cross range in the vehicle because
we thought that was the right way to build the vehicle and
it just happened to give the Air Force some capability they
wanted. But we wanted it for abort capability during the launch
and we wanted to start flying the vehicle right at entry.
We didn't want to keep the thing above stall all the way down
to landing area and then flip it around. So the myth that
the Air Force made us do something we didn't want to do is
absolutely a myth.
DR. LOGSDON: The implications of that design for thermal
protection came along with the NASA engineering decisions.
MR. THOMPSON: We got the same thermal protection the
way we fly the shuttle that we were going to get with the
straight wing. The straight wing was not any benefit thermally
I guess it's awfully interesting to me, look back over 20,
25 years, the myths that have grown up and where they have
come from. But I'll go on the record today saying NASA built
exactly the vehicle it wanted to build.
DR. LOGSDON: I guess the final thing I'd like to talk
about a little bit is the cost estimates for development and
operation that were provided, again, to the political level
of decision-making. OMB gave you a budget ceiling, I believe,
in May of '71 that said you had to build the system with a
billion-dollar development cost; and the ultimate presentation,
at least to the White House level, said you could do that,
or 5.5 billion, with an operating cost of $118 a pound. I'm
curious where those numbers came from, particularly the operating
MR. THOMPSON: Well, I'm not going to answer just the
operating cost; I'm going to answer the whole question.
DR. LOGSDON: Good.
MR. THOMPSON: Again, one of the big myths on the shuttle
is that it was way over budget. That's an absolute myth. In
December of '71, when Jim Fletcher and George Low went to
San Clemente to present the final recommendation to President
Nixon, we prepared a letter that George and Jim took with
them, a one-page letter. That letter said that we felt we
could build the configuration that you now know as the shuttle
for a total cost of $5.15 billion in the purchasing power
of the 1971 dollar but that it would take another billion
dollars of contingency funding over and above that to handle
the contingencies that always develop in a program like this.
So you need to budget 6.15 billion in the purchasing power
of the '71 dollar and that we could build it and fly it by
1979 if everything went perfectly, but the $1 billion and
18 months ought to be planned in the program because that's
probably what will really happen and we'll probably fly it
in early '81. That was in the document.
Jim Fletcher and George Low went to San Clemente, had a little
model of the shuttle. President Nixon approved it. He came
back into the agency at NASA. Bill Lilly, who was the comptroller
of the agency at that time, took that letter and started his
negotiations with OMB. When he finally got around to getting
it through the OMB cycle, they took the letter and said we'll
take the 5.15 billion but we won't give you the 1 billion
because we never budget contingencies. We'll hold you to the
1979 launch date because we never launch budget contingencies
there, and we'll put it in the '73 budget at those numbers.
So we lost two years of inflation in that little maneuver
in OMB. I went back and talked to Bill Lilly. He said, "Shut
up. You got your program. Go on about your business." So we
did. During those years of the shuttle development, inflation
got as high as, what, 20 percent, 18 to 20 percent some years.
We would usually get maybe two thirds of that out of the Congress.
Also, the shuttle was picked as a program to be monitored
by OMB and they actually put five or six people out of the
OMB into my office level here at the Johnson Space Center
and they monitored for several or probably two years exactly
where all the spending was to try to keep an accountability
in the program.
One of the fellows who worked for me in the financial area,
named Hum Mandell, kept a very accurate level of the spending
in the shuttle program. When we finished the program, his
record showed that the orbiter actually underran our original
budget, including the 1 billion-dollar contingency and the
18-month schedule. Our schedule was right on. The other elements
of the program were slightly over. The total cost of the program,
when you account for inflation, account for the under-commitment
of the '71 to '73, you account for the deliberate schedule
that OMB asked to us do with their funding, he came to me
after the first flight and says, "Here. We can prove you met
your cost and schedule goals." I called John Yardley in Washington
and John says, "Hell, why don't you put it in a filing cabinet.
No one's interested in that." We put it in a filing cabinet.
Hum took it and got a Ph.D. thesis on it at the University
of Colorado. So you can get his thesis and read it if you're
really interested in the true funding.
One more thing. I remember being called on television at the
time, not knowing that Jules Bergman was going to be on. After
they introduced me, Jules Bergman says, "Hey, Mr. Thompson,
you said you could build this thing for $5 billion. You've
already spent 8 1/2 billion. That's a terrible overrun. What
the hell you going to do about it?" Inflation doesn't mean
a thing to the people who write in the papers, and it's a
pretty complex job to keep up with the true cost of a development
program like the shuttle. In fact, after three years, OMB
quit and went home. So the myth that the shuttle was way over
budget is another myth.
DR. LOGSDON: Bob, you didn't answer the question about
MR. THOMPSON: All right. Operating costs. (Laughter)
I had a better answer for development costs.
At the time we were selling the program at the start of Phase
B, the people in Washington, Charlie Donlan, some of them
got a company called Mathematica to come in and do an analysis
of operating costs. Mathematica sat down and attempted to
do some work on operating costs, and they discovered something.
They discovered the more you flew, the cheaper it got per
flight. (Laughter) Fabulous.
So they added as many flights as they could. They got up to
40 to 50 flights a year. Hell, anyone reasonably knew you
weren't going to fly 50 times a year. The most capability
we ever put in the program is when we built the facilities
for the tank at Michoud, we left growth capability to where
you could get up to 24 flights a year by producing tanks,
if you really wanted to get that high. We never thought you'd
ever get above 10 or 12 flights a year. So when you want to
say could you fly it for X million dollars, some of the charts
of the document I sent you last night look ridiculous in today's
world. Go back 30 years to purchasing power of the '71 dollar
and those costs per flight were not the cost of ownership,
they were only the costs between vehicle design that were
critical to the design, because that's what we were trying
to make a decision on. If they didn't matter -- you have to
have a control center over here whether you've got a two-stage
fully-reusable vehicle or a stage-and-a-half vehicle. So we
didn't try to throw the cost of ownership into that. It would
have made it look much bigger. So that's where those very
low cost-per-flight numbers came from. They were never real.
Let me make one other comment. In my judgment -- and no one
can either agree with this or disapprove it -- in my judgment,
it would have cost more per flight to operate the two-stage
fully-reusable system than the one we built, even though the
cost analysis didn't show that. When you get two complex vehicles
like that and all one vehicle does is help you get up to staging
velocity -- and the staging velocity is 12,000 feet per second
-- when you build a booster that does nothing but fly up to
12,000 feet per second, you've built something wrong. I think
that's what the two-stage fully-reusable system was; and I
think, had the system tried to build it, we wouldn't have
a shuttle program today. My feelings.
ADM. TURCOTTE: You've largely described what could
be in today's, I guess, modern management vernacular as a
matrix organization as it existed back in the Sixties and
Seventies, et cetera. You also described some complex relationships
between both contractors and the different center directors
and the program manager, element managers, subsystem managers,
MR. THOMPSON: There were no complications on the program
management channels. They were very clear.
ADM. TURCOTTE: Okay. Could you explain the difference,
as you see the organization today, in its relationships and
its matrix structure today, and compare and contrast it to
the Sixties, Seventies, and up to, say, the middle Eighties.
MR. THOMPSON: I could not, because I'm not in detail
familiar with what they're doing today.
MR. COHEN: I don't think I can either. I knew that
question was going to be asked, but I really don't know enough
about what they're doing today. I understand the system very
well. You described it as a matrixed system. It was. It may
appear to be complicated, but it was really very well defined.
The people, when they came to work every day, they knew what
they had to do; and both at the contractor and at NASA, they
knew what they had to do and they knew what their role was.
MR. THOMPSON: I want to try and make another comment.
A lot of the people at NASA had come from working in a research
center back at Langley, through Mercury, Gemini, Apollo, Skylab;
and when we got to shuttle and set up the matrix organization
for shuttle, it was clear to me then and it's clear to me
now that the primary responsibility for integrating that program
was the government's responsibility. So when we wrote the
RFP for the contract that Rockwell ultimately won, we asked
for them to build us an orbiter and to provide major systems
engineering support. We did not say you're responsible for
systems engineering across the program and we didn't say you're
responsible for integrating the program, because they had
no contract leverage over any other part of the program. They
had no responsibility for the tank or the booster rocket and
so forth, no direct responsibility. So it was the government's
responsibility to integrate the program.
Now, we used all of the hardware development contractors in
a very heavy support role. A lot of the ICDs were actually
prepared on assignment by Rockwell in Downey, but those ICDs
came into Owen's office for review. They went across the total
program for review and came to me for signature, and I had
the full control of those ICDs. Aaron couldn't change anything
that impacted the tank. The tank couldn't change anything
that impacted the orbiter without coming back to me at the
systems level. So it was no doubt but what the government
had the program management and the programs systems engineering
integration responsibility, but we plugged the contractors
in in a way to use their talent as effectively as we could.
GEN. BARRY: I've really got two questions, if I may.
One has to do with history, and one has to do with design.
On the history element, could you please give us maybe a characterization
of what I'm going to say here -- and correct me if I'm wrong
in any of it. It has to do with compromises.
Now, after, of course, when Apollo was coming to the end and
Jim Fletcher was administrator, there were plans, originally,
to put stations on the moon. Then that was backed off by the
administration and there was a space station design with a
shuttle. Then that was given up in place of the shuttle as
we know it today, which was a bit of a compromise to try to
put a space station capability payload to orbit, get down
to hopefully $1,000 per pound eventually at some future point,
depending on how many times you flew per year. The historical
question I'd like to ask is: What compromises were made on
the structure development on the shuttle in that time period?
Then I'll ask my design question here.
MR. THOMPSON: I hate to keep hogging the thing here,
but you're asking history and I guess I'm the oldest person
here. To answer your question, I've got to take you to 1968
or '69 -- I can't remember which year -- and the Space Council.
Do you know what the Space Council is?
GEN. BARRY: The vice-president.
MR. THOMPSON: In 1969, driven by the fact that the
government works on five-year budget plan, it was then incumbent
on NASA to put some dollars into the out years for where they
wanted to go post Apollo. So the nation then came to a fork
in the road or what are you going to do with manned space
flight, in 1969, because you could see the end of the Apollo
program. We had already decided what to do with the residual
hardware in what became known as the Skylab program. If something
wasn't done, we were going to go out of the manned space flight
business. That simple.
So the vice-president at the time, Spiro Agnew -- and this
thing never really got advertised very much maybe because
of that -- in any event, he chaired the Space Council and
they worked for about six months and they looked at where
this nation should go post Apollo, so-called post-Apollo planning.
I'm sure those are in the records and you can go back and
That Space Council looked finally at four major options. They
looked at a manned Mars expedition, they looked at a follow-on
lunar program, they looked at a low earth orbital infrastructure
program, and they looked at getting out of the business. They
looked at those four things.
They made the decision to have a low earth orbital infrastructure
program. It wasn't we'll build a shuttle or we'll build a
space station, you know. We will have a low earth orbital
infrastructure program. It never got announced like Kennedy
announced the lunar program, but that decision was made by
the President on the advice of the Space Council.
Now, up until that time there had been a lot of debate in
this country about whether space station should be a great,
big, artificial-gravity rotating wheel launched on Nova-class
boosters or whether it was to be a zero-G station built on
orbit in modular form with something like the space shuttle.
The desire for a zero-gravity, modular space station prevailed
at that time. It was a commonsense, logical thing to do; but
before you can go that way, you obviously have to have something
called a space shuttle. You have to have a truck and a personnel
carrier and a work machine to go up there and do that work.
Also, at the time the President was giving the head of NASA
instructions to come down off the 3 1/2 percent spending that
we had peaked at in Apollo, down to about 1 percent spending
for the agency. As Jim Fletcher looked under his 1 percent
spending -- with Apollo ongoing, with Skylab ongoing -- he
felt that he couldn't have but $1 billion annual funding expended
on low earth orbital infrastructure development.
We then undertook obviously to build the shuttle first and
then the modular, zero-gravity space station second; and the
low earth orbital infrastructure gave the nation a capability
to operate from the surface of the earth up to 600 nautical
miles, operating shuttles and space stations and interim upper
stages that would take payloads from that low earth orbital
up to geosynchronous orbit. As the thing evolved, we started
with the shuttle; and the requirements for the shuttle were
driven 99 percent by what we wanted to do to support the space
station. It also happened to give the Air Force the kind of
payload volume and the kind of capability they wanted, although
they really wanted to be at higher orbits for their work.
So the Air Force came in and said we will plan to use the
shuttle and we will also take on the task of building the
interim upper stage, which was part of the low earth orbital
infrastructure. So NASA embarked on the shuttle. It wasn't
necessary to commit to a space station at that time because
the shuttle had to be built and operational before you commit
to space station, and the President at that time, Nixon, had
other things on his mind. He didn't get up and make a great,
big speech about low earth orbital infrastructure.
So now a lot of myths have grown up about we stumbled between
space station and the orbiter and we wanted to do an orbiter
this way and an orbiter that way. That's not the way it happened
at all. It was pretty orderly planning. It was a decision
to go to the low earth orbital infrastructure. Let's have
a shuttle, then let's have a modular zero-gravity space station.
Once the Challenger accident occurred, the Air Force got off
of the ship and stuck with their original vehicles, which
I think was probably the right decision for them all along
because the nature of their missions don't fit the shuttle
quite that well but they could have done some of their work.
But they actually developed the interim upper stage and they
built a bunch of launch facilities at the West Coast that
we ultimately phased out.
GEN. BARRY: Let me ask the following question based
on a historical perspective. Can you give us an understanding
of the design specifications for the orbiter to take debris
hits? When you finally settled on the design after going through
these ramifications of alternatives and finally settled on,
as we know, the space shuttle system to be today, our question
from the board repeatedly is: Was the space shuttle designed
to accept debris hits from foam, either at the RCC or at the
belly with the tiles?
MR. THOMPSON: The answer to that is no. The spec for
the tank is that nothing would come off the tank forward of
the 2058 ring frame and it was never designed to withstand
a 3-pound mass hitting at 00 feet per second. That was never
considered to be a design requirement.
MR. COHEN: You've got to recognize in the first early
flights we were concerned about ice coming off the tank. That
really was our big concern, was ice going to come off the
tank, because we knew ice would do very serious damage. Ice
would do serious damage.
MR. THOMPSON: But usually ice under insulation was
our principal concern where you would get a crack in the insulation,
you had cryopumping under there, you'd get ice formed up under
it, and a chunk of ice and insulation come off. We must have
-- Owen, you can estimate, 15 -- we had so many meetings on
trying to make sure we didn't have ice, we called them the
ice follies meetings.
MR. COHEN: And we still have an ice team today that
goes out and inspects the vehicle before every flight.
MR. THOMPSON: I don't know what they're doing today.
It was my understanding -- and you can correct me, Owen. I
was pretty sure we did ultrasonic testing on the tank foam
insulation, looking for any voids. We carefully did visual
inspection. We put together a very comprehensive ice team
that walked up and down the vehicle just before liftoff. We
put the beanie cap on top of the tank to capture the cold
exhaust gas to make sure no frost or ice built up there. We
even talked one time about building a great, big building
around the whole thing and environmental control it, but we
decided that really wasn't probably necessary.
We paid an awful lot of attention to making sure nothing came
off, because we knew if we fractured the carbon-carbon on
the leading edge of the orbiter, it was a lost day. We could
take a fair amount of damage on the silica tiles and still
be all right, but it was a maintenance problem. So we worked
very hard to make sure we did not have any foreign-object
DR. SILVEIRA: You have to understand the exterior of
the vehicle of the orbiter is glass. I mean, the coating on
the tile is a silicate glass, and you have to treat it like
that. So, yeah, impacts are not allowed.
MR. JEFFS: Let me hitchhike on that briefly, too. That
is that it's kind of incongruous, when you look at the overall
picture, the RCC panels are -- the bottom line, for example,
the rear of the panels is not completely true. There's a little
waviness in it which is just due to the way it comes off the
tool and spring-back and so on; but when the tiles are matched
to it, the tiles are delicately matched to mix those interfaces
all the way along. With a graphite epoxy, the coefficients
of expansion are such that you can maintain those shapes just
right. Then we stand back and think, gee, there we go to great
pains to kind of hand-tailor all of this stuff and then all
of a sudden we're hitting it with debris. It just is two different
MR. THOMPSON: Well, let me comment. The silica tiles
that are on the orbiter behind the carbon-carbon, in the damage
testing and the testing we did on that during the program,
in most cases the type of damage you would expect to get on
those is not the kind of damage that kills you. Most of the
time when you hit those tiles hard with something, they were
fragile enough that you knocked the outer layer off but the
inner layer where it's been densified against the two glue
joints and the strain isolation plate, just a portion of the
silica, the two glue joints and the strain isolation plate
gives you enough thermal protection to make an entry. So people
have gotten locked up on the fragile nature of the silica
tiles. The silica tiles are fragile to damage, but they're
actually pretty forgiving. You can take a lot of damage right
there. You cannot take any damage that knocks a hole in the
carbon-carbon leading edges.
MR. JEFFS: Well, let me add one thing to that. That
is that they're a robust system from what their designed to
do, and that's to take the heat loads. They are a little delicate
here and there when it comes to like the coatings because
the coatings are part of the radiating heat transfer. So the
coatings are meant to be there, and it's also pretty critical
on the front edges of that system so that you don't trip the
boundary layer. You certainly don't want to trip the boundary
layer on the front end of that thing.
So as Bob says, those tiles along the interface to the RCCs
are also densified. So they're a higher density than the tiles
further aft. So they're stronger. You do that, taking with
it the higher thermal conductivity through the thing, and
still maintain the bond line temperatures. So they are more
rugged and they will, as he says, give you assurance you're
going to get through even if you have some missing, but you
don't want to do that and you don't want to nick them on that
MR. COHEN: We were concerned early in the program when
you would damage a tile and that tile damage at the bond line
and that the heating then would cause what we call an unzippering
effect where you actually damage the bond line and a lot of
tiles would come off. That would be the case we were concerned
about; but as Bob said, the tile is actually pretty forgiving
with a reasonable type of hits. But you can't take large hits
that really cause you a damage that would destroy the boundary
MR. THOMPSON: Let me take you back on this and tell
one story. We were doing some thermal testing of the silica
tiles in a thermal wind tunnel out at Ames. We heated the
air stream with some carbon heating elements and there was
a test panel with several silica tiles put on it that would
be put downstream and then you would hit it with this heat
pulse in the aerodynamic wind tunnel there. We ran the tests
on the silica tiles. Lockheed, which was the system manager
for the silica tiles, ran these tests out at Ames, and the
heating elements, the copper heating elements in the tunnel
failed and they put a whole bunch of carbon shotgun-like particles
in the air stream. They actually blew off probably 70 percent
of the silica tiles, just like you would shoot it with a shotgun.
They brought that to my office to show me what happened on
that. I said, "Well, okay, that's fine but what happened to
the temperature of the aluminum behind it for the re-entry
They said, "Well, instead of 200 that we were looking for,
it got up to 3 or 4 hundred degrees, but it didn't structurally
I said, "That's the best test I've seen in a long time."
MR. JEFFS: Just a couple of notes on it. When you look
at that wing after flight, it's fascinating to see where the
transitions occur. You can see from the heating patterns under
the bottom wing. You can see how far back that transition
is. So you're laminar a long way back, which is very reassuring.
Even if you had a nick along the front edge locally, it doesn't
necessarily transition the boundary layer throughout the total
wing. It could be just in the local air of the wing, and it
would be probably be survivable. So we weren't really concerned
with the zipper effect. Fletcher was really worried about
that, but we didn't think that would occur.
MR. THOMPSON: Well, you don't want to leave the impression
that if you trip the boundary layer, you would lose the vehicle.
MR. JEFFS: No, but I didn't say that. I said you could
locally trip it and you could have higher heat transfer coefficients
in that region but you're not going to necessarily lose the
MR. COHEN: Let me ask you a question. You may be more
familiar. Have you gone back and looked at Volume 10 now?
Do they have a requirement in there for the size of debris?
GEN. BARRY: Volume 10.
MR. COHEN: Volume 10 would be the design specification
DR. SILVEIRA: That's a Level 2.
MR. COHEN: Do they have a criteria in there?
GEN. BARRY: They do have a criterion, and it's like
.006 foot pounds per hit. It's very, very small. It's almost
minuscule to the point where it can't take hits, just like
Dr. Silveira mentioned. So that's the puzzling aspect because,
in reality, as you trace the hits on the orbiter from the
very beginning, from the very first mission, they've averaged,
you know, as high as 700 on STS 27 to 300 on STS 87 and almost
every orbiter has averaged about 50 to 100 hits. So it's interesting
to see that the design specification really was not to allow
for any hits, although the reality has been it's been pretty
durable for most of that; but the design specification is
contrary to the reality.
MR. JEFFS: Weren't the majority of those coming off
DR. WIDNALL: What runway?
MR. JEFFS: Landing the thing. You get a lot on the
runway. That runway is coarse.
MR. THOMPSON: Here again, Aaron was talking about a
document that was called the 07700 series of documents. Those
are the Level 2 documents that I controlled to put the specs
across the program. Volume 10 was one of those specs, and
that was where the 2058 ring frame came from. In any practical
problem, it would be nice to meet all of your specs. In the
real world, though, you know, I will sit here and let you
shoot at me with a pop gun that's got a little cork in it
that won't come half way over here all you want to; but if
you pick up a .45 and shoot at me, I'm going to get the hell
out of here. So you've got to have some judgment when you're
operating a vehicle of this nature of what you're willing
to live with and what you're not willing to live with. And
that's hard to write in a specific spec and it's hard to live
in an ideal spec world because you run into practical problems
like popcorning of insulation.
MR. JEFFS: Let me say one more thing. I might have
left the wrong impression here, too. That is, you know, first
off with the RCC. We were always concerned about the RCC and
the loads on the RCC. We spent extra money and extra time
to go to the woven cloth, for example. We didn't go to the
single filament stuff to take advantage of the load direction
and all this jazz. We really went overboard to make that as
strong as possible.
We went through the whole litany with McDonald on the problems
they were having on trying to make a graphite tail for the
F-15 or F-18. I don't know which one it was. They had a lot
of problems with it relative to how you weave in the middle
interfacing elements of the carbon-carbon. You can't just
drill holes in carbon-carbon. So you've got to weave in the
interfacing metal elements in order to attach it to the air
frame. So they had special techniques that they had gone to
to wrap it in like you tape-wrap a swollen ankle or something
like that, to really get those pieces in there right. Went
through all that stuff with them. So we really had a rugged
RCC. That RCC, the Q alphas are, I don't know, 900 to 1100
something like that, pounds per foot. So they're taking a
pretty good load up in that front end. So they're not wussies.
MR. THOMPSON: Well, they are strong; but they're still
a ceramic. What you don't do is hit a ceramic with a real
sharp, high-energy low-time blow. Anything going 700 feet
per second, even if it's a soft piece of insulation, if you
look at the force-time curve that we put onto that insulation,
we didn't do a dead-chicken test. We knew well you could knock
it off if you hit it with enough potential energy, kinetic
MR. JEFFS: You guys mentioned the holes have been mentioned
on the RCC. When I looked at the first flight back, up at
Edwards, I was looking at boundary layer transitions pattern
and stuff. I noticed on the underside of the wing that I could
see occasionally a few holes. They looked almost like a circular
hole. Completely circular. Almost like a hole that would be
popped out of your porridge when a steam bubble come up out
of a porridge, you know. I couldn't figure what those things
were. I thought maybe we might have trapped water in the zip
or something and we had gotten over the boiling temperature
of water, which is like 160 or something like that at the
altitude, and that we were building ourselves a little steam
engine there and that might be accounting for the tiles occasionally
popping off, which we couldn't figure out why they would occasionally
come off. But we ran some tests and they ran some tests lately
at Langley and they haven't verified that that's any condition
at all. I noticed you said there some round holes on the RCC,
or somebody was saying that there were some holes. We just
don't know what the nature of those holes are. We had never
seen those before. We had never seen any of those at testing.
GEN. HESS: One of the issues that's often discussed
in the back rooms of the board is this thing about whether
or not the shuttle is an operational vehicle. We wonder if
y'all could share your opinions on that versus being an R&D
MR. JEFFS: I've got a lot of heartburn I can share
with you on that. You know Beggs wanted to declare the shuttle
operational after about five or six flights. That was one
of the reasons for the SPC. It was one of the reasons for
the shuttle processing contract being given at the Cape. Our
arguments or my arguments were that we were still learning
about the machine and we still had a number of things to really
sweat out before we completely understood it and all the characteristics
and, therefore, the development contractor should be maintained
strongly in that act.
MR. THOMPSON: George, you need to ask him what an operational
vehicle is. Define it. A vehicle that flies to earth orbit
will never be operational in a sense a 747 is operational,
if that's your definition of an operational vehicle.
MR. JEFFS: So we were as operational as we ever had
a space machine, I guess, because we had flown it that many
MR. THOMPSON: But it will always be a risky endeavor.
MR. JEFFS: It's a machine that doesn't have the same
wear and tear as an aircraft. I mean, we're not landing it
ten times a day or what have you. It does take heavy loads
on launch. It takes thermal loads on re-entry. So it's different.
It doesn't do much on orbit. It's pretty easy for it on orbit.
But it is not a hard-driven machine from an operational point
of view and it's more like a helicopter.
MR. THOMPSON: You're still hitting it with 4 million
pounds of thrust.
MR. JEFFS: Well, you only do it every once in a while.
You only do it twice a year rather than ten times a day. I
wanted to add one more thing to it, though. That is, further,
it's like a helicopter, and even more so, in that when you
get it to the ground, you can do anything you want to it.
You can re-examine it. You can change, add to the tiles, fix
the tile problems and so on. So you're rebuilding the machine
MR. COHEN: No matter what you say, the hardware, the
process, whatever, needs to take -- you need to have tender,
loving care of it.
MR. THOMPSON: Always will. You need a development mentality
organization managing it.
MR. COHEN: It's a hostile environment you go into and
MR. JEFFS: With all respect to Beggs, though, he wanted
to -- the other side of that argument, the flip side obviously,
is that if you're the development contractor, you're continually
making changes to it. So stop making changes, guys, to make
it better all the time. That's where he was coming from.
MR. THOMPSON: I've heard that all my life: "Don't make
changes." If it's about to break, you better change it.
MR. JEFFS: You've got to have those kind of eyes looking
at it so they can see ahead of time before it's about to break.
ADM. GEHMAN: I'd like to ask Mr. Morris and Mr. Silveira
if you'd comment on this, whether it's an operational or a
MR. MORRIS: Well, I would go back to Bob's question.
How do you define operational? I think, in my experience,
any high-performance aircraft is continually being inspected,
is continually being modified. They're being updated with
glass cockpits and other things that are systems upgrades.
But any high-performance vehicle is continually being modified.
I think the shuttle, although I haven't been involved with
it for many years now, has been modified more than most operational
aircraft, things you call operational; but I don't think there's
a difference in the amount of changes made. I don't think
there's any difference in the philosophy of the way you manage
the program or operate the vehicle. I think a high-performance
vehicle, be it in space or in the air, continues to be something
you are developing and you're learning more about as you operate
MR. THOMPSON: I think it's also somewhat delusionary
to think you can start with a new sheet of paper and build
a new vehicle and it won't have any problems and it will be
easy to operate and it will be cheap to operate and everything
will be fine. That's always what you come out of Phase A with;
but once you build it -- and particularly if it's going to
sit on the surface of the earth and then accelerate to ,000
miles an hour, stand re-entry heating, land on a runway --
you're going to have to give it a lot of attention.
MR. JEFFS: As you say in the aircraft business, it's
operational on condition. It's an on-conditional airplane,
but you've got to have the right eyes looking at it to know
when that on-condition time occurs.
ADM. GEHMAN: Mr. Silveira, any comment?
DR. SILVEIRA: You know, like with any vehicle, you
have to continue to scrutinize the results of every flight.
You know, we had many thousand hours on 737s when we had to
go back and modify the actuator and the rudders because it
didn't really work the way we thought it did on that. I think
that's the thing you have to continually do with any aircraft.
Now, as the aircraft gets more mature, of course, you can
back off some on the scrutiny; but where the shuttles have
actually very, very limited amount of flight time, then you've
really got to pay a lot of attention to it. You say: Are they
operational? To a certain extent, yes, but you still need
an awful lot of engineering scrutiny to examine what the results
were of the last flight.
MR. THOMPSON: You have to also recognize that a rocket
engine, you're essentially building a very hot fire in a cardboard
box; and you have to do it very carefully. If you get a little
bit off on your cooling paths and so forth, you burn up your
MR. JEFFS: We've come a long way. We didn't really
know that much about the regen system with the SSMEs. As a
matter of fact, we had a lot of trouble going through the
gates to get the engine started. The guy I worked for at the
time that ran Rockwell used to say, "How in the world are
you going to get three engines started at the same time if
you can't start one?" That was a very good question. We've
come a long way and we've learned a lot about the engines.
Where we found shortfalls -- or not shortfalls -- but marginal
conditions and we were operating with low margins, those are
things that have been worked on. Changed. Addressed. The pumps
and so on.
MR. THOMPSON: And the digital controller.
MR. JEFFS: And that's the kind of whole process that
should go right along with the evolution of the whole system.
Someday it will be even more on-condition in toto, but it
will still have those things in it that we learn from the
operation of a system like this in space, which is new. We
don't have the aircraft background that we had.
DR. HALLOCK: You mentioned Volume 10. I've had some
many sleepless nights looking at, trying to understand what
was going on, and looking at this evolution over time. You
also mentioned that one of the criteria you had was that you
didn't want to have any strikes, foam strikes, is the way
we were talking about it at that time. But how about the ambient
environment itself? I mean, things like what you might expect
in that when you get up into orbit, such as space debris and
micrometeorites and other types of things that could also
cause damage to the craft?
MR. THOMPSON: I would comment that we did not know
enough about the orbital environment to practically say what
kind of impacts you should take from orbit. So, frankly, we
did not spend a lot of time trying to design the orbiter to
take hits while on orbit from unidentified objects.
MR. COHEN: We did have a criteria -- and I believe
I'm right -- the criteria in the orbiter that you could have
a penetration or an opening of a half an inch or so diameter
and have makeup volume, makeup gas.
MR. THOMPSON: You're talking about the environmental
MR. COHEN: Yeah, environmental control system. So the
crew could get their suits on and do a de-orbit. But that
was not for space debris. That was just for a penetration.
MR. JEFFS: We did have the specs on particle size impingement
on the windows and what have you. So the windows are all designed
MR. THOMPSON: For a certain particle size. But you
could certainly get above that.
MR. COHEN: Sure. As Bob said, I don't recall orbital
debris being discussed very much.
MR. THOMPSON: I don't think you would really know enough
today to put a good spec on a system flying low earth orbit.
MR. JEFFS: We had some data from Apollo that we used.
MR. THOMPSON: It's going to have to be a judgment call
DR. HALLOCK: One of the things you hear a lot of discussions
going on at this point is: Is there someway that one could
make a repair on orbit? Were those kinds of issues addressed
back in those times?
MR. THOMPSON: They were discussed. They were never
addressed in a serious way.
MR. JEFFS: Well, we were pretty serious about trying
to figure out how the heck you might replace a tile. There's
a young lady in the bowels of NASA named Bonnie Dunbar --
or Donnie Bunbar or whatever they called Bonnie -- and she's
a Ph.D. in ceramics. She was right in the middle of the tile
operations. She worked for us a while up at Palmdale. We often
discussed how in the heck if we look at the detailed process
of what the guys had to do just to get a tile on and how you
would do that with gloves, you know, in an EVA situation.
And it's not easy. I'll tell you it's not easy. You know,
you've got to pull-test it and you've got to do lots of things
with it to verify that you've got -- you might take some shortcuts
if you just had to make a repair in orbit, I suppose. I suppose
it's doable, but it's very tough. Now, how you replace an
RCC panel? That's something else.
MR. THOMPSON: First of all, I noticed in the paper
a lot of conversation about looking at the shuttle while on
orbit. We did look at the shuttle while on orbit for the first
shuttle flight, using the Air Force resources. It was more
from a we would just like to know ahead of time whether we've
got some potential problem in front of us, not because we
had any ability to go inside and do very much about it.
MR. COHEN: Those things are documented. I don't recall.
But the real issue is going EVA and trying to get to the various
parts of the vehicle. Even if you had a kit, it's very difficult.
With the space station there, it may be another thing.
MR. THOMPSON: You could do some things like that. It's
a matter of whether that's a good expenditure of your resources
with the probability of what you can really do that's practical.
GEN. HESS: I'm kind of curious if you would characterize
for me the role of the safety organization in the structure
that you had back in the Sixties and Seventies in terms of
how it integrated itself with the system development.
MR. COHEN: Let me say a little bit from the orbiter
point of view on the changes. In our Change Board and my daily
meetings, SR&QA had a person sit in on every one of our meetings;
and I think that was the same thing at Rockwell, also, from
the orbiter point of view. Somebody was there. Again, very
much as the engineer was a check and balance, SR&QA was a
check and balance because in that case I believe Marty Raines
was the head of SR&QA and he reported to Chris Kraft. So again,
if SR&QA had an issue with what we were doing, just as engineering
or operations, there was a check and balance at my level.
MR. THOMPSON: Well, I think I'd comment this way. Within
the program, there was a very active Safety, Reliability &
Quality Assurance presence and activity. We did all the usual
failure mode and effects analysis. We did all the development
of critical items list. I signed off on probably several hundred
critical items, recognizing if that item failed, we'd lose
the vehicle. Safety was spread throughout. Safety, Reliability
& Quality Assurance was spread throughout the entire program.
We looked very carefully at whether we wanted to do what we
called the nines business, whether we wanted to attempt to
do statistical quality assurance kind of things. In looking
at the spectrum across the shuttle systems, the part of the
system where the nines kind of approach made sense in avionics
and things like that was a relatively small part of the overall
system. So we did not go into a formal statistical qualification
program where we could get nines that had some meaning to
tell us which part of the system was relatively good and which
part wasn't. We tried that on Apollo and gave up on it, more
or less. A lot of consideration was given to what we called
the formal or statistical safety and quality analysis, and
we decided it was not worthwhile to try to lay that on the
How you put the statistical number to an O-ring failing is
pretty hard to come by; and if you have a lot of garbage in,
you get a lot of garbage out. So I think you have to be very
careful. If you're building television sets by the thousands
and taking data on this resistor and that resistor and it
tells you which resistor is causing your televisions to quit,
it probably has some value; but when you look at most of the
systems on the shuttle, you cannot do the kind of numerical
numbers of tests to give you, under a properly controlled
condition, any kind of valid input data. And once the people
get those nines, they really maneuver them, whether they have
any real meaning or not.
Owen, you may want to comment on this.
MR. MORRIS: You know, if you take this and go to the
structures, which is really kind of where we're interested
today, we did use fracture mechanics, fracture analysis. We
did have margins in the vehicle; and that's the way, again,
aircraft are designed. Structure has to be qualified to the
level of the margin, and then it has a reliability of 1 in
your nines approach.
MR. JEFFS: Structure is tough, but we also have redundant
load paths. So if we had one failure, we had a second path
in order to take the load.
MR. THOMPSON: In some parts of the system.
MR. JEFFS: Wherever we could.
MR. THOMPSON: For example, we went to safety factor
of 2 on the solid rocket boosters. Typically the Air Force
in their ballistic programs were using either 1 1/4 or 1.4.
We went to a safety factors of 2 on these SRBs in the amount
of insulation we put in, in the structure, design allowables
and so forth, which is relatively high for these kinds of
systems; but we did it because we didn't have a backup for
the SRB. If the SRB failed, you lost a system and we knew
that. We didn't get there by nines; we got there by safety
factors, as best we could.
MR. COHEN: Design philosophy, at least. Margin in the
design, whether it be electronics or it be structures, is
important. Redundancy and margin. I would say margins first
and then redundancy. If the redundancy adds to the margin,
then it's good. If the redundancy doesn't have margin, then
it's not good. So that's what we really looked for was margin
in your design, the deterministic type of analysis rather
than probabilistic analysis.
MR. JEFFS: The tiles in the design was considered for
100 missions with a factor of 4. So a factor of 4 was on top
of that 100 or so. That was considered in the design. The
orbiters were built by MCRs. The MCR is a Master Change Record.
I signed every Master Change Record, and I looked for lots
of things in those MCRs and one of then was safety. But we
had organizations that were tuned and they came out of the
Apollo program. They were looking for the what-ifs. They were
looking for failure modes and how to recover from failure
modes. So therefore, in the design, how do you put something
in when you don't have those failure modes? So we had a very
sensitive organization to that; and that was partially schooled
into them from interfacing with the Mission Control, for example,
in the Apollo stuff, on how to respond and react to in-flight
emergencies. So a lot of that basic background was in the
fundamental design as best we could put.
MR. THOMPSON: We haven't mentioned sneak circuits.
We did all the typical sneak circuit analysis work. We did
all of the kinds of things we had learned to do in the previous
programs to prevent the rocket going off when you hooked the
battery up and that sort of thing.
MR. JEFFS: All the golden chute things and everything.
ADM. GEHMAN: All right. We have a lot more questions
and we're going to go on for at least another 90 minutes,
but we're going to take about a ten-minute break here so we
can all pay attention and be in comfort while we're doing
ADM. GEHMAN: All right. Ladies and gentlemen, we're
ready to resume.
Gentlemen, thank you very much for your very forthcoming answers
to our questions. We appreciate it.
Dr. Widnall, if you're ready, go ahead.
DR. WIDNALL: I'm going to ask an engineering question.
Given that at that period of time that composite materials
were sort of new -- in fact, not to make a pun of it, they
sort of were at the leading edge -- I sort of would like to
understand what kind of testing was done on the RCC panels.
For example, was there a lot of fatigue testing done? Did
you have in-flight unsteady pressure loads data that you could
use for fatigue testing? Did you cycle the panels through
a vibratory environment followed by heating and ultraviolet
or whatever-else-is-up-there environment? Did you rip them
apart? Did you impact them with small pellets? What kind of
testing was done on the RCC? It's clearly an important issue
for the design of the vehicle.
MR. JEFFS: Let me tell you what little I know, and
a lot of things I don't know the details of. First off, the
RCC panels, I'm sure, in the process, were subject to all
the rigors of qualification of everything else on the program;
and that included structural testing of all major elements.
So the RCC panel was certainly a major element. The interface
of the RCC panel to the wing structure itself was kind of
a critical area. The whole issue of water in graphite epoxy
and how it might play in the game. The whole issue of the
specs re salt water, et ceter. Now, whether they vibrated
the panels or not, I don't know and I don't have the documentation
to identify it, but I would be very surprised if there weren't
detailed documentation of the structural testing of those
panels and the load interfaces to the wing. I don't remember
anything in the way of impacting those panels with high-velocity
particles or something like that. I don't remember that, but
the rest of it I do recall that there was.
DR. WIDNALL: What about testing to destruction? I think
one of the issues that we are amused by is that the RCC panels
seem to have broken right along the center line of the leading
edge. So were the panel destruct-tested by putting loads on
them to see where, in fact, they would break?
MR. COHEN: Testing we did on the panels. On the RCC
MR. JEFFS: I'm surprised that it would break in that
DR. WIDNALL: I know. I was surprised. I have no explanation
MR. JEFFS: As I said, that cloth is woven cloth.
DR. WIDNALL: No, right along the leading edge, they
broke. I have no explanation for that, but I wondered whether
structural tests had been done.
DR. SILVEIRA: I don't recall.
DR. WIDNALL: I know they're very expensive panels.
MR. JEFFS: Yeah, what we could test, we tested; and
we tested to know what kind of margins we had. We tested them
certainly up to yield; and whether we went to ultimate on
those panels, I don't know. But I'm sure that the Boeing guys
would have that in their files.
ADM. GEHMAN: Someone else want to make a comment?
DR. SILVEIRA: Don Curry was subsystem manager on the
RCC, would be familiar with what testing we did. As I recall,
we took a number of panels to destruction. I don't remember
seeing a failure like that, at least in the stuff that he
MR. JEFFS: We had material we could work with. You
know, there was a long process that they went through at Vaught
to develop the panels because the panel were pyrolyzed, as
you know, and you build them on this tool that has to go in
the oven with the panel, and then we would get spring-back.
So they went through a lot of steps before they got the right
spring-back in those panels. So they had panels to work with;
and Vaught, in general, did a very good job on those panels
overall. So I'm sure that they tested those.
MR. COHEN: I'll refer to this document.
DR. WIDNALL: Thanks a lot, Aaron.
MR. COHEN: It does talk about -- this is the space
shuttle technical conference and Don Curry --
DR. WIDNALL: I would love to get a copy of that.
MR. COHEN: It does talk about the early design challenges,
the leading edge. Of course, one of the big issues was the
coating, the coating and the degradation of the coating and
how the panels degraded with the degradation of the coating.
Now, it doesn't go into a tremendous amount of detail in here,
but it does give you an overall view. This was written by
Don Curry, and Don Curry is the subsystem manager. I don't
have the data in front of me, but I'm almost sure we did take
the panels to do some structural testing on the panels. I
don't have it here but --
DR. SILVEIRA: The RCC was really a big technical challenge,
as far as building the panels. You know, when we started doing
it, John Yardley made a comment to me one day. He said, "If
I ever hear about delamination, it's going to be your job."
Well, LTV actually did, I think, a superior job in putting
it together. They really did. You had to pack the panels in
carbon retorched to form and the like and there were very,
very few quality problems that we experienced during the development
of the panels.
MR. COHEN: They did Eddy current testing and sonic
testing of the panels in the manufacturing process.
MR. THOMPSON: There was never any thought, though,
that those panels would withstand a 20,000 foot pound kinetic
energy strike. They were not designed for that. The whole
intent was to not let it happen. You could not set out and
design -- I wouldn't know how to design the leading edge of
that wing to take a 20,000 foot pound kinetic energy strike.
DR. SILVEIRA: Not many airplanes are designed that
MR. THOMPSON: I think we may have had to abandon the
program, had that been a requirement.
GEN. BARRY: I'd like to address the issue of the design
of the space shuttle itself insofar as life span is concerned.
Right now in our readings, of course, the original design
was to fly 100 times in ten years. So that's ten times a year
per shuttle. Here we are at 2003. We know the Columbia was
on its 28th flight, not 100, and certainly not within ten
years. So we've entered an era that the board has pretty well
identified as an era of reusable vehicles in an aging space
platform in an R&D or development based environment. So let's
say aging spacecraft in an R&D environment, for practical
purposes. I'd like to get your perspective on how long you
anticipated in the original design on how long the shuttle
would last, in light of the fact that NASA has announced now
that the shuttle will fly until 2020. Can I get a perspective
on life span for the space shuttle?
MR. THOMPSON: Let me comment. Then I'd like to have
some of the other people talk. We debated a lot about what
kind of a number to put in the spec for that. Frankly, we
could never find very much that was sensitive to that number
in the kind of application we were talking about for shuttle.
You know, 100 times would be a minor load for an airplane
or airplane structure or fuselage and so forth. We put it
in there to help ferret out any problems that people might
come back and say, "Hey, it won't go 100 times." I don't remember
anyone coming back and saying that was a constraint for anything.
I would think, with reasonable attention and oversight and
proper upgrading of subsystems and replacement of systems
as appropriate, I don't see any reason why the shuttle couldn't
last many, many years. You know we have B-52s out there flying
after 30 or 40 years. We've got some T-38s at Ellington that
have got how many years on them. So that 100 number we put
in there was never much of a driver to us on the program.
We didn't quite understand what we were trying to control
with it in the first place very thoroughly, and it was more
put in there to see if it drive anything out. And I don't
ever remember anyone coming and asking for an option on the
Owen, you may want to add more to this.
MR. MORRIS: I don't think, in my memory at least, that
we ever really addressed any issue that said we have to have
5 more pounds or we have to do something to be able to reach
100 missions. I keep going back to aircraft; but, again, if
you look at T-38s, yeah, they're still flying. They're flying
okay. Now, they've had some wing problems. There have been
cracks. The cracks are carefully monitored on a per-flight
basis or every ten flights, whatever the spec is on that,
and you continue to operate. You know, I think you can do
the shuttle the same way.
MR. JEFFS: Let me say a couple of things about it.
What we did on both Apollo and shuttle, we did have age life
critical item identification. So we identified all the items
that we knew about in the system that were age life critical.
For example, all the rings, the N204 and all those seals were
on that age life list. There are all the pyros. The pyros
were also bootstrapped so that you fire pyros every six years
from the same lot to see that, in fact, you still had life
in that pyro which could change.
I think the specs for the review of the orbiter after every
so many years, there are certain items called out to look
at specifically in those; and some of those were kind of age
related in the thinking when they went into that review spec.
It's kind of like the 3,000-hour turbine engine or something
like that. They're in that overhaul spec requirement.
I think the rest of it, as you say, it was a development item.
We didn't know everything, too, that might have some characteristics
re aging. So a lot of that is as required as we go through
and look at the spacecraft. Certainly, you know, I think about
this ofttimes at night because I own and fly helicopters a
long way and what I do in those helicopters is far less than
what we do on that shuttle in the way of looking at it very
carefully to see what is aging as we go through the process,
particularly on the thermal protection system.
MR. COHEN: The real issue on extending the life would
be the obsolescence of the subsystems, the replacement of
parts and the computers and this type of thing. Of course,
we did upgrade the cockpit; and really obsolescence of hardware
and replacement of hardware is probably one of the biggest
issues, I would think.
MR. JEFFS: Let me say another thing. One thing that
worried me was the screed. The screed worried me on the wing.
I was worried about screed from the point of view of were
we introducing something here that could, in fact, be sort
of a zipper kind of effect. So I specifically went after that
through the years; and the guys convinced me that there was
no aging identifiable, that we had a true, solid bond in the
screed on that wing. So that's one of the kinds of things
you look at from an aging point of view.
ADM. GEHMAN: If I could follow up on that, some things
age by how many times they've been used, like cycling an aircraft,
but then there's also some things that chronologically age.
Carbon-reinforced panels and things like that age by stress,
but they also age chronologically. If you had an RCC panel
and you left it out in the breezes of the Atlantic Ocean and
you never flew it, it would deteriorate. But wiring ages and
wiring insulation ages. And you mentioned seals and things
like that. They obviously age. But there are a number of critical
items on the shuttle which, when you get to the 20th anniversary
and you're thinking about flying it another 20 years, even
if they've been properly maintained, it does occur to us that
there are a number of critical systems that have to be looked
at very, very carefully. Wiring comes to my mind. Wiring insulation.
MR. THOMPSON: Then again, you still have to ask yourself
am I safer to continue to do that or do I embark on building
a new vehicle, which one puts me into more risk. Frankly,
the vehicle you have experience on, if you're looking at it
at that level and watching those kinds of things, you may
be safer sticking with the B-52.
MR. JEFFS: Let me say something about wiring. After
the Apollo fire, we redesigned the Apollo; and the wiring
in that Apollo was superb. I mean, it's better than any airplane
I've ever seen, by far. That same wiring, all those wiring
specs and so on, were carried over into the orbiter. So it's
not just a matter of redundancy in the wiring and separate
routing of the wiring; it's the detailed quality of the wiring
itself and the combing of the wiring and the ties of the wiring
and the curvatures and everything else that are all carried
over directly into that shuttle. So there may be wiring problems
there in the insulation, for example, in certain areas and
it should be looked at, but in general you're starting out
with a wiring set that is far superior to most of those that
you're normally familiar with.
ADM. GEHMAN: Let me ask a question.
MR. JEFFS: May I say one more thing there?
ADM. GEHMAN: Absolutely.
MR. JEFFS: On the panels, the RCC panels. We were always
worried about water in the RCC panels because, you know, graphite
epoxy is sensitive to water. You get water in it and you're
going to lose properties of the graphite epoxy -- and it is
graphite epoxy, after all. So it always worried me that we
should take a special look at those panels, and I think the
guys were doing that. For example, in the Columbia I think
those had just gone through a recycling back at the plant,
as I understood it. I was always worried in that hashed-up
field that we've got between those bodies that we might get
some occasional buffeting on those panels and might be working
the RCC panels at the interface to the structure itself. I
don't know whether that's true or not. There's no way to tell,
you know; but it is one of those kind of things that would
contribute to aging in that you get a lot of cycles on that
ADM. GEHMAN: That's a line that we're curious about.
For example, the RCC is a pretty tough piece of structure
but one wonders, after it's been heated to 2000 degrees two
dozen times or three dozen times, what are the changes in
its properties. That's one of the things we would like to
MR. JEFFS: You've got some RCC panels back, didn't
ADM. GEHMAN: Oh, yes.
MR. JEFFS: They went through kind of an unusual environment,
but you might get some information along those lines.
ADM. GEHMAN: We're going to do things like shoot foam
at them and things like that at 00 feet per second.
Let me change the subject here a little bit and go back to
the original design here again, the Seventies again, and talk
about weight. Weight was one of the issues that you all wrestled
with in order that you could get enough payload up to make
it worthwhile, and the history of the program shows a lot
of concern about weight -- the weight of the vehicle, the
weight of the payload, and a number of steps which were taken
to lighten the vehicle and to thereby increase what it could
Certainly, as a layman, one of the things that struck my attention
was the decision to stop painting the ET because you could
save 375 pounds worth of paint. So you get the impression
that the concerns about the weight of the vehicle as it developed
and the weight of the payload it could deliver into orbit
was always on your mind as you were watching weight at all
times. Could you describe the history of that process and,
am I correct, was this a big concern that you were watching
all the time?
MR. THOMPSON: Well, let me comment on that. Anyone
who designs a vehicle to go to orbit will have to be careful
about weight. Getting 99 percent of the weight to orbit isn't
acceptable. So one of the things we struggled with was how
to, first of all, select the weight targets and how to allocate
the weight among the elements, what kind of weight to hold
in reserve at the Level 2 or the program manager's level and
how to manage weight over the lifetime of the program like
As we got underway in the development program, we intentionally
phased the startup of different elements based on several
considerations; but weight affected some of this. We started
the rocket engines for the orbiter first because we felt that
was the most difficult development cycle. Several months or
almost a year later, we started the orbiter development; and,
of course, all during that time we were doing the systems
engineering level things, doing the wind tunnel tests of the
total system, doing the overall early design things that begins
to see how much a design, as it matures, might meet the weight
target you put in it to start with.
We deliberately delayed the start of the external tank until
we were pretty far along on the engine and the orbiter so
that we could then size the tank, because the amount of propellant
and the ISP of the propellant tells you what you can take
to orbit. We then started the SRBs last, and we actually left
some growth. If you look at the SRBs today, unless someone's
done something I haven't heard about, there's about 2 feet
on the front end of the SRBs where you could add more SRB
propellant if you really had to. Now, you only get a 1-for-8
gain on the SRBs; but there was still that kind of consideration
as we got into weight.
Now, once you have gotten into the program well enough to
where you then can have pretty good confidence on your allocations
to the different project elements, you still keep a certain
amount of weight reserve at Level 2. Then if one of the element
managers begins to complain that he's got a problem he'd like
to fix but there's a weight constraint -- I can remember in
one of our ice follies tests the tank project manager wanted
me to give him relief from ice forming on the LOX line because
it was going to take too much weight to fix it and a little
bit of ice isn't going to hurt you. I said, "No, you cannot
have any ice on the LOX line and I'll give you 500 pounds
to go fix it." And he went and fixed it.
Now, did weight make us do anything dumb? I don't think so.
Did we have to manage weight from Day 1? Absolutely. The 65,000
pounds, 100 nautical miles due east, when we got to the point
where we had to trade a little bit off late in the development
program, we did; but then we got it back. Fairly early in
the program, we went to the fusion-bonded titanium thrust
structure in the orbiter because we picked up a good block
of weight and we thought it was a good thing to do, not because
we were in so much trouble we had to do it. But we had to
do it -- I mean, we did it to pick up that weight.
As far as I know, they quit painting the tank after I left
the program. Painting the tank gives you a little bit of advantage
to the external surface, but the number that I remembered
was 700 pounds of paint on the tank. As far as I know, they
quit painting the tank more to save money and it wasn't really
necessary rather than that they were in any kind of critical
We put moderately tight but reasonable weight targets, and
I cannot excuse a single dumb thing we did on weight.
Owen, you maybe want to comment at a systems level.
MR. MORRIS: Actually I think you're right, Bob. We
did have a weight margin all the way through. As I remember,
the tank decision to take the paint off the tank -- and this
was after I left, but I was associated with it peripherally
a little bit -- I think at the same time we quit machining
the tank after we sprayed it. Initially there was a machine
job; you actually machined the foam. This left a much more
porous surface. At the time that it was decided not to machine
it anymore, you then had a hard finish on the outside of the
foam; and the paint was no longer needed. And the tank guys
at that time, I think, had some weight problem and that was
a good trade-off to trade that.
MR. THOMPSON: I do remember one time in discussing
with J. Bob Thompson, the engine program manager, some concerns
he was having. I asked him specifically. I said, "J. R., if
I give you another 1,000 pounds of weight, is there anything
you want to do differently?"
He said, "No, I don't want another 1,000 pounds of weight.
I don't need it. I don't want it."
MR. JEFFS: Let me add a couple of things. One of the
reasons that the aircraft falls through as far it does on
landing is the short forward landing gear. One of the reasons
for that is to make sure that the weight was minimum of that
landing gear. So we looked for saving weight everyplace we
could on this machine. It's characteristic of all the space
programs, as Bob said. On the MCRs that I talked about, which
are thousands of them, every one of them has a place on it
for how much weight this change adds to the system and which
drawings carry them. So it was pervasive, and it was designed
that way to be sensitive of the weight.
MR. COHEN: From Day 1 in the orbiter project, we were
concerned about weight and we had a weight problem, but as
Bob said and George said, I don't recall doing anything that
was irresponsible because of weight.
Of course, that heritage came from the Apollo program. You
talk about a weight program. Owen was the aluminum module
program manager, and we didn't get off the lunar surface unless
we get to some real fancy footwork on reducing the weight
of the lunar module. On the command module we had to take
weight out because of the parachute hang weights. So we had
weight problems on every program, but I don't think it caused
us to do anything that was irresponsible.
MR. JEFFS: As far as Bob's comment on the weight side,
the element of the system that has worried a lot of us from
the beginning the most is, of course, the engines, the SSMEs.
We're always been concerned that that was probably the place
that if we ever had any problems, that's where we might have
them. Of course, we had years of development of engines at
the bottom of flame pits and so on, as we went through that
development, to understand how sensitive and how critical
that element was.
One day Sam Phillips and I were sitting together at a meeting
at Rocketdyne and they were talking about the weights on every
individual component of the engine. We thought that was the
right thing to do as far as the requirements were concerned;
but we thought, gosh, if we had to allocate the weights, we
would probably add a little bit more to the engine side somewhere
here, guys. But that's the only area of weight allocation
that I could see. We didn't have any problems with embracing
that concept on the orbiter itself.
ADM. GEHMAN: Thank you. Another design parameter that
historians have written about is the requirement for reusability.
For example, as you are well aware, re-entry vehicles prior
to this had had, for example, ablator-type coatings on them
which were, of course, gone when they came back but --
MR. JEFFS: Not true. They weren't gone. Some of it
ADM. GEHMAN: They were used.
MR. JEFFS: They were used. I spent a lot of time trying
to convince NASA to shave off those ablators to fly again.
They were over-thick.
ADM. GEHMAN: They were well used when they came back.
But the reusability parameters drove a number of things. Well,
I'll let you describe for me what kinds of things it drove,
but the history tells us that it drove such things as TPS
systems which could be taken apart in little sections so you
only had to rework little sections at a time and things like
that. I don't know if that was driven by reusability or not.
You can correct me on that. Again, going back in your experience,
how was the reusability requirement characterized in your
decision-making and your engineering design work?
MR. THOMPSON: Well, again, let me start off. At the
systems level when we got into the early Phase A part of the
program, full reusability was leveled on the program as a
program requirement, under a perception that that would make
it a more cost-effective program, particularly in the cost-per-flight
regime. Of course, that was coming into a space business where
staging and expendability had been a fundamental part of flying
to space. One of the reasons the early system could go to
space was because you would stage. You'd go part of the way
and throw off weight. That even helped explore the South Pole
when they went down there.
So we accepted reusability during Phase A and came up, as
I talked earlier, with the two-stage fully-reusable vehicle;
but as we got into Phase B and particularly began to look
at the details, when you've quit cartooning and gotten down
to the specifics of designing and building and basing your
reputation on something, then you begin to ask the question,
does it really make sense to do it that way. I used to make
a kind of simplistic argument that if expendability didn't
make sense, there wouldn't be any Dixie cups around. You know,
everyone would wash their cups and reuse them.
So there are systems that are more cost effective if you throw
part of the system away. Particularly as we looked at putting
the cryogenic propellants inside these vehicles and you had
to think about insulating those tanks, making a good thermos
bottle inside that tank and accommodating a minus 30-degree
liquid that's going to shrink that tank. I've got to shrink
that tank 6 or 8 inches and it's still part of my structure.
Putting cryogenic tankage within the aerodynamic envelope
of the vehicle is an extremely difficult job. I don't think
we've even done it to this day. So it began to make a lot
of sense, at least to me and lots of others when we got into
Phase B, to look at throwing part of the system away. The
first thing we did was take the LOX out of the orbiter and
then we took the hydrogen out of the orbiter and then we looked
at, well, if we did that, we got the orbiter down to a size
where we didn't need this kind of booster and this booster
had a hell of a lot of complexity to it and maybe if we want
to meet the national funding level, this is a better way to
go than that way and might even be better if we had all the
money in the world.
So reusability had a significant impact at the broad systems
level and the fact that we put the propellant in an external
tank and threw it away, in my opinion, was probably the best
-- and I would even defend today -- the best overall systems
level decision we made. I think even if you were starting
a system today with today's technology, you might come to
the same conclusion.
Now, reusability, once we decided to partially reuse the boosters
by fishing them out of the ocean and cleaning them out and
so forth, brought some concerns to us, particularly as it
affected the gimbaling of the nozzles on the SRBs. You have
to worry about the APU and the gimbaling systems and so forth
after you parachute them into the ocean. So that reusability
was concerned; but the fact that you got them and looked at
the O-rings and things of that nature were some pluses.
Reusability on the orbiter? I never remember the fact that
we were going to use the orbiter over and over gave us any
unique set of problems that we could have avoided by throwing
something away. Throwing the tank away, I think, was a great
thing. Partially reusing the SRBs made a lot of sense; and
reusing the orbiter, particularly with the three expensive
engines in the back end, made a hell of a lot of sense.
MR. COHEN: Well, you question is, if we didn't have
reusability on the orbiter whether we could come up with a
different thermal protection system. I think that's where
you were going with it. I don't know the answer to that, but
I do know that if you had tried to use something like an ablator,
it would be very, very heavy. You know, just to give you an
example, if I recall correctly, the Apollo ablator was something
like 100 pounds per cubic foot and the tile is something like
9 pounds per cubic foot, 20 pounds per cubic foot. So if you
tried to use an ablator on the orbiter, although we have ablators
now that are much lighter, you would probably never get off
the pad. But I don't think that you would have come up with
a different thermal protection system.
MR. JEFFS: The whole beauty of the system is the reusability.
I mean, you get the spacecraft back. That's the first time
we got a spacecraft back really to speak of, unless you got
some pieces of it back on parachute or something for other
reasons. It's the first time we got the engines back. Usually
the engines guys bury their sins in the Atlantic Ocean out
there. That's what ELVs are. We don't do that; we get it back.
If you try to minimize cost to orbit, you get your airplane
back, get your hardware back. So these guys got as much of
the hardware back as they possibly could; and the orbiter,
bless its heart, is the most beautiful example of reusability.
That whole reusability was facilitated by that radiated heat
shield to get back. And getting the engine back was an added
bonus. So you want to get your avionics back which are expensive,
your engines back which are expensive --
MR. COHEN: Fuel cells.
MR. JEFFS: -- your air frame back. And the heat shield
makes that possible.
MR. THOMPSON: But had we made you put all of that cryogenic
propellant internal to the orbiter, you'd have had a hell
of a bunch of ablator.
MR. JEFFS: Much more difficult.
ADM. GEHMAN: Thank you. But tell me something. I mean,
I understand what you're saying, the fact that we have this
wonderful reusable machine is a work of art and a work of
engineering. It's an engineering feat. But you are trading
some things. For example, you are lifting three 8,000-pound
engines into orbit for no good reason other than reusability.
MR. THOMPSON: You've got to go to orbit with three
8,000-pound engines, no matter what you do. You can't get
there without those engines. Now, you can throw them away
or you can bring them back. Now, the orbiter has to have some
capability to bring ,000-pound engines that wouldn't be there;
but you've got to go to orbit with those engines.
ADM. GEHMAN: Just as you have to have the ET to supply
the engines with fuel.
MR. COHEN: Right.
ADM. GEHMAN: The ET doesn't go to orbit.
MR. THOMPSON: Well, it goes, for all practical purposes,
within a feet per second to orbit. Then you use the OMS to
kick it on into it. We did that so we could put it in the
Indian Ocean where it didn't bother people.
ADM. GEHMAN: I'm not in any way diminishing the engineering
feat of building the orbiter, but there are design trades
that were made in here. For example, if you decided you wanted
to reuse the engines or for some reason it was a requirement
of the system that the engines be part of the reusable cycle,
you now are in the position of having to lift the engines
and bring the engines back. It makes the mass of the orbiter
higher on re-entry by 10 percent or something like that.
MR. JEFFS: That's the price of a two-way airplane.
ADM. GEHMAN: That's correct. I assumed that this was
all debated and there were people that had positions on both
MR. THOMPSON: It' still being debated.
DR. SILVEIRA: I think involved in that, of course,
was the operational cost of the shuttle in itself and then
what you want to do is to return the high-dollar cost components
like the engine and the avionics and the like. So as a result,
you place the main engines in the orbiter. You know, no doubt
reusability shaped the thermal protection system because the
two that we really gave serious thought to were high-temperature
metals as well as surface insulator. Surface insulator, we
thought, was a considerable weight saving.
When we started the program, we actually took on three major
developments. One was the main engine, which was the only
thing that made shuttle possible. The other thing was a TPS,
which was a major development. You know, we ended up with
6-inch tiles because the guys kept coming to me after tests
and said, "Milt, the 12-inch ones keep cracking in half,"
and I said, "Well, why don't we make them 6 inches." That's
what we settled on. I mean, simple as that. Then, of course,
the other was the integrated avionics which, you know, is
very complicated because, again, when you decided to take
the engines to orbit, this gave an airplane with a very aft
CG and as a result you had to go to a control-configured system
to be able to fly it back.
MR. THOMPSON: Well, you would have had to do that anyway,
DR. SILVEIRA: Not necessarily. I think you could have
flown it back without it if you had a proper CG on the airplane.
GEN. BARRY: I'd like to address another topic, if I
may. Another topic would be managing risk, if I could get
your perspective on this. We have clearly a system of systems
integration element here with the STS. We are trying to address,
as a board, providing substantive recommendations that might
allow the shuttle system program to be strengthened. So in
light of the way you managed risk at the beginning of the
program, I'd like to maybe call on that knowledge base to
just comment on a few things.
I know from the readings -- and, of course, my experience
at NASA during the Challenger when Milt and I were there at
NASA headquarters -- that with the CIL listing, you clearly
had a focus -- and you've already brought it up a number of
times -- that a certain was with the SSME. Then we have a
failure on a simpler, less-complex part of the shuttle; and
that is, of course, the O-ring on the solid rocket booster.
Now, we jump 17 years later and you look at the CIL list again
and, lo and behold, at the top of the CIL list is a clear
focus on the SSME and we have a problem with, of course, the
tragedy on Columbia and it is part of the simpler part of
this system of systems. It's foam on the external tank as
the leading candidate, as the board has been working here
and trying to determine what the cause.
So the question that we have really got is: How do you manage
risk in a system of systems, complex environment that certainly
we have here, when you clearly have a good focus on some of
the complex elements -- and the SSME is a case in point --
but we miss listening to the materiel that is talking to us,
insofar as an O-ring in one case and maybe some foam in this
MR. THOMPSON: Let me start with that and then y'all
jump in. What you say certainly was the emphasis on -- if
you had asked me when we started this program what would be
the first thing that would fail that would cause us to lose
a system, I would have probably talked to you about a failure
in the liquid engines in the orbiter, No. 1. I might have
talked to you about some failure on the thermal protection
system. I would have been a long time probably before I got
down to an O-ring on the SRB; but independent of that, any
flight anomaly should be put on a PRACA, Problem Report And
Corrective Action list. And the discipline in the system ought
to be such that that PRACA is properly evaluated, in the sense
that it's very clear whether it's a life-threatening issue
or is not a life-threatening issue and who can sign off on
Now, the O-ring, I could argue whether that would be something
that the SRB project could handle alone because you could
argue that's internal; but when it's squirting hot gas toward
the tank, it's not internal. It's a Level 2 PRACA. Both of
those items should have been entered on a Problem Report And
Corrective Action. It should have been listed as something
that could destroy the system and it should have come to the
Level 2 program manager for full discussion and full disposition
and full willingness to accept it on the next flight. And
at the Flight Readiness Review, the program manager should
have signed off on both of those PRACAs, saying, "I understand
what the failure is, I understand the consequences of it,
and I'm willing to fly." Now, if the system's working, that's
the way you manage risk; and you should manage it whether
it's an O-ring or TPS or a turbine blade in a main engine.
It should be no difference.
MR. JEFFS: Let me make a suggestion here. I spent some
time on this broad area of management review operation with
Sheila and others on the Deltas. I think it gets down to the
depth of what was stated here by Bob, and that's attention
to detail and to every last detail. Every last detail. It's
hard to just wrap your arms around something and corral that
One thing that I have found useful in the past and suggest
on big programs to look at where some of these details need
further scrutiny are the MRs. The MRs are Material Reviews.
They are identifying little places that you should listen
to. In the space business or in airplanes or anything, you've
got to listen to the little voices because that may be the
last thing you hear.
MR. THOMPSON: And you have to hear the little voices.
MR. JEFFS: Yeah. You've got to hear them, and you've
got to do something about them. What I suggested doing with
the MRs is what I call -- it's kind of a parallel to what
Krantz and NASA and others have done down here on the what-if
processes pre-flight -- and that's to review each MR. If I
have an accident, I'm going to go look at the MRs among other
things, first thing anyhow. So look at the MRs and do a pre-accident
investigation. Just like it was an accident. Go through all
those MRs. They are at least an identifier of where some of
those voices are listening to be heard.
So how to answer your question any further than that, I don't
know. It's get to the details and get to the right details,
and that means you have to look at all the details.
MR. THOMPSON: But these two items that have caused
the accidents in shuttle are clearly Problem Reporting And
Corrective Action items. Clearly. And if the PRACA system
is working, if they're properly identified and they're brought
to the right level and the right people discuss it and they
make a decision, right or wrong, that's the way the system
works. You've got to get them discussed with the right information
and the right people and make the right decisions.
ADM. GEHMAN: Let me follow up on that. I think we all
kind of agree with that. Some management arrangements migrate
over the years. For example, the experience base of you and
your team having wrestled with Gemini and Apollo issues, when
you had to make engineering decisions or engineering evaluations
in the shuttle program, you all came with a rich history of
being able to sense when you were operating too near the edge
of margins and you had the dirty-fingernail basis for understanding
that you really did have to give that guy a 500-pound budget,
you had to increase his weight budget and he really did need
that 500 pounds to do that.
Over the years, management styles have changed. Management
organizations have changed. A number of things have happened.
For example, the role of the U.S. government person has migrated
up and been filled in behind by contractors such that we don't
have government people -- not that they're any better than
contractors, but they have a different reward system. The
experience level of these managers didn't get the same experience
that you had because they didn't have all of these projects
to experiment on and grow up in and they just don't have this
rich background that you all have. They're just as smart and
just as dedicated, but they just don't have the same background
that you all have.
You have such managerial twists as this Max Faget and his
engineering department has been morphed over the years now
to where the programs have to pay his bills or he loses his
employees. In other words, he's not independently funded anymore.
That's a gross exaggeration; they are, but not to the extent
that they were independent back in your days. There are a
whole lot of managerial trends that have taken place, driven
by style and budgets and things like that.
So now we get to this meeting in which we're going to properly
process an IFA or properly process a waiver or properly process
some kind of a PRACA or something like that, but the machinery
has changed now. The mechanisms have all changed. Based on
good principles, based on first principles that you all have
indicated, how do we balance this thing so that these good,
proper sign-offs can be made by people who are qualified and
understand the system, when the things are not the same as
they were in your day and they can't be made the same? I mean,
we can't go back and find people with the same kind of experience
you had. It's not possible because NASA doesn't have, you
know, four or five different space exploration projects going
on in sequence in which to build the people with your experience.
So somehow we've got to replace that.
What I've heard from you and what I've written down are what
I would call first principles, and the first principles are
you have to have knowledgeable people with experience and
they have to have the authority and they have to have the
richness of engineering horsepower behind them in order to
make this case. And there has to be some checks and balances.
Three or four of you have indicated checks and balances, not
single-point failures in the management system.
Could you give me your views on today how you accomplish the
things that you've said, when the dynamics of the management
system have changed so much?
MR. THOMPSON: Well, you've asked kind of a complicated
question for some discussion there. Let me comment this way.
I think clearly, over whatever period of time you want to
talk about, you have to maintain the internal procedural disciplines.
You have to maintain the PRACA system and you have to maintain
the forcing function that that puts in the program because
that's a discipline that makes you look at anything that's
off nominal whether it's in the worrisome engine or in the
not-so-worrisome SRB. So you have to deal with PRACA. You
have to deal with it in a formalized way through a Flight
Readiness Review or whatever technique you want to use. So
you have to maintain those systems.
Then you have to maintain enough high-quality well-trained
people to make good judgments with those decisions. Neither
one of these accidents that we've had on shuttle require Ph.D.s
in physics to understand. In fact, they barely exceed high
school physics to understand. Erosion rates on an O-ring when
there should be no erosion is an obvious thing. Kinetic energies
of a 2 1/2 or 3-pound hunk of tile when it's traveling 700
feet per second, that's high school physics. There should
not be anyone in a key management position in a shuttle program
who doesn't understand those things in considerably more depth
than it would take to make a good decision on them.
Now, why those things didn't happen is the kernel of your
question. It appears to me that the agency needs to, No. 1,
make sure that the procedures that bring the Problem Report
And Corrective Action to the right discussion forum and then
the right people are dealing with them in a timely manner.
Now, having said all that, there may still be some actions
that occur in the shuttle that those systems don't catch;
but there's certainly no excuse not to have those systems
in place and have reasonably good people deal with them.
MR. COHEN: I think George Jeffs probably said it the
best and the simplest. I think the people involved need to
pay attention to detail, need to bring issues forward, that
they need to pay attention to detail.
MR. THOMPSON: And they need to understand them. It's
one thing to pay attention; it's something else to know what's
MR. COHEN: I'll tell you a story, if I may. We were
getting ready to go to the moon on Apollo 11. I remember this.
The initial measurement unit on the lunar module was no drift
rate. All of a sudden it started drifting high but not out
of spec. We, the Draper Labs or the MIT instrumentation lab
and the subsystem managers, all went to George Low and told
him he did not have to change the IMU on the lunar module.
Very risky. The lunar module was made out of Reynolds wrap
almost. And George Low looked at us. He said, "You may be
right, but I'm going to change it out." It was telling a message.
It was telling a message that it was drifting -- not out of
spec but it started doing something different. I'll remember
that as long as I live as a thing that you need to think about.
MR. JEFFS: Well, you've got to make sure that you get
people in the right places that qualify in three categories.
One, they've got to be intelligent. They've got to be dynamic,
and they've got to care. They've got to care. If you lose
any one of those three, you've got a miss. So you've got to
make sure at least the leadership has those qualities. That's
for the near term.
For the longer term, though, it's a bigger problem because
we in industry are losing our capabilities in these areas
and our backgrounds; and you in government are doing the same
darn thing. I don't know what the answer to it is. Apollo
was a stretch. Apollo stretched us technically, and it brought
to bear a lot of interest and a lot of people in science and
engineering. In the broader sense, we probably need something
like that in the future to be able to attract our young people
to science and engineering.
DR. LOGSDON: This is really kind of a follow-on to
the discussion we were just having. I mean, the five of you
represent the first generation of people that learned how
to do things in space in this country. As Bob Thompson has
said, putting people in orbit and getting them back safely
is one of the hardest things that humans do. Most difficult.
Most challenging. You are all here under the auspices of the
NASA Alumni League, which should indicate that you have continued
some involvement with the agency. Are you willing to give
us your impressions of the NASA of 2003 as an organization?
Is it up to the job that faces it? If not, what sort of things
you've suggested in the past few minutes are needed to fix
MR. THOMPSON: John, I would personally dodge that question
because I left NASA 20 years ago. I do not think that manned
space flight is beyond the technical capability of this nation
by any stretch of the imagination. I think the young generation,
in many respects, is smarter than we are by far, better trained.
So I think that what we're talking about here is easily achievable.
There's no reason the NASA of today can't function well and
operate the shuttle safely, whatever that means, and take
on whatever future things you want to do in manned space flight.
So I haven't lost faith in the agency.
Now, I do think you have to be extremely careful when you
draw the interface between government and industry. I've been
on both sides of those fences. The people on both sides are
just as honest, just as dedicated; but they're driven by different
things. If you're in industry, you've got a different set
of constraints on you if you run the program than you are
when you're in the government. I think the NASA of today ought
to be very careful in drawing back so far and saying that
contractor's responsible. When he really doesn't have the
ability to be responsible if he doesn't control the subs or
doesn't control the associates or he's not in a position to
make all the right kind of balance judgments, don't put the
muscle on him. I mean, don't put the monkey on his back if
he doesn't have the muscle. So my only comment is I don't
believe NASA is serving itself well if it pulls back too far
in feeling an overall technical management responsibility
for ongoing programs.
MR. COHEN: I'm not going to answer your question directly
either because I've been away from years. But I have had the
opportunity since I've been gone to teach at Texas A&M. Seniors.
I can guarantee you that those young men and women that are
coming through the class, I would hate to compete with them.
They are truly outstanding. Many of them, whether they get
their advanced degrees and go to MIT or whether they go to
Purdue or whatever, most of them want to go to work for NASA
or their contractors.
So good students are very interested in the space program
and a lot of my students did come to work at the Johnson Space
Center and other space centers. So, you know, I think the
people are there and the people are good. I mean, the students
today, as you know, are just outstanding.
DR. SILVEIRA: John, if I may. You know, there's no
doubt in my mind that the kids today are better educated than
we are. I have two kids that work in the program, and they're
both smarter than I am. The thing I get paid for, at least,
is to try to go out and find out what's going on in industry
that we don't get the product we used to get out of them.
I think some of it comes about because we have started to
train a lot of paper engineers rather than hardware engineers.
Kids are not looking at the hardware enough to really understand
what's going on and, anytime there's a little discrepancy
in it, really get to understand what is happening. The hardware's
trying to tell us something, and we don't carry it to a point
where we really go and understand it and fix it.
You know, recently we had a PDR of one of our programs, you
know, and the contractor was proud: "We have spent 3,000 man-years
on documentation." I can't imagine a program demanding that
kind of paper to keep it going. I think the thing we need
to do is to get kids out from behind the computers and get
them to go out and walk the factory floor and really see what
hardware's all about.
MR. JEFFS: I'll say three things from the industry
side. I won't try and reorganize the NASA. That takes a little
longer. But I think that, as Bob mentioned, we march to different
drummers, in a way; but when I ran the space and energy operations
for Rockwell, I was also a corporate vice-president of Rockwell.
So I had a lot of pressure that didn't have a thing to do
with the space program, but it didn't keep me from applying
the right kind of people on the problems at the right time
in the right way. And I think these guys will all attest that
they didn't see anything in the results of what happened with
the industry on their hardware that was influenced in any
negative way by profit motives or otherwise in getting those
No. 2, there are a lot of smart people out there in industry.
They can be assigned. There are talents available to the people
that run these companies. I think it takes their focus also
to get the right kind of people in the right place at the
right time on the space program and to look at their priorities.
The third thing is that one of the things that made Apollo
and shuttle happen was an excellent working relationship between
industry and government. That working relationship was criticized
in many ways by being too close and what have you; but I assure
you, when it came to solving the technical problems, it wasn't.
I also assure you when it came to getting any money out of
these guys, it also didn't manifest itself in the way of excess
profit. So I think that encouraging the good working relationship
on mutual utilization of each other's capabilities is an excellent
additive to making these big programs happen properly and
MR. MORRIS: I'd like to follow up on that just a little
bit. I think one of the things that over the last 10, 20 years
has happened in this process of NASA going up and being backed
by contractors is a lack of sufficient check and balance.
The one thing we had in the Apollo program, in the shuttle
program, during the design phase, was parallelism between
the government and the contractor. Both were very good, but
they also were checks and balances. When you turn all the
responsibility either to the government or all the responsibility
to the contractor, you lose some of that check and balance.
I think the process that you have to look at things like the
O-ring or like the foam, you need to make sure the process
you have asks the second question, not what did that cause
on the last flight but what else could it affect. I think
in both cases the second question was not asked properly.
I think that's the thing that can be fixed with a system.
The system that assures the right checks and balances and
the right questions are asked.
DR. WIDNALL: Not including the space program, what
are the other major scientific and technical challenges faced
by our nation that have the power to motivate our young people?
MR. THOMPSON: I think, frankly, the Defense Department
is one of the greatest motivators of our young people. I think
maintaining a very strong and very active military or defense
capability or offense capability, either way you want to talk
about it, is a very important contribution to our society.
We in NASA often take a lot of credit for technology advancement.
I'm not so sure in the same number of years the technology
advancement wasn't stimulated more by the Defense Department
than NASA. The fact that you have to solve the kinds of problems
that the military solves on a routine basis drives technology
certainly as much as the space program. Obviously medical
research. So I could list eight or ten things, but certainly
we benefitted to a great extent in the NASA space program
by what was going on in the Defense Department in similar
activities -- be it rocket science, be it structures, be it
flight control systems.
For example, at the same time we were putting the control-configured
flight control system on the shuttle, DOD was doing the same
thing with the F-16. And we visited their research laboratories
and they visited ours. We took some things, learned from them.
They took some things and learned from us. Both systems are
working today, 35 years later, quite well. So I would like
to see us maintain an extremely strong national defense capability,
if for no other reason, to drive the kind of thing you're
MR. COHEN: I think in my observation, being in academia
for a while, that there is a lack of funds for students that
want to get their advanced degrees, to go on to get their
Master's degrees and Ph.D.s. I think that could be a big stimulus
to producing more graduate students and actually enhance our
engineering capability in the country.
MR. JEFFS: They had a session not too long ago that
George Abby pulled together at Rice that addressed the subject
in part; and it seemed to me that to attract the young people,
it's going to have to take something that has duration long
time. Most of the military programs, albeit some of them are
changing now, are lesser duration. It needs something that
people can address and assign their life to, youngsters, and
enthusiastically do that. I think that the NASA has that within
its grasp if they better structured and articulated the total
space program, the unmanned systems and the manned systems.
And I think manned systems have to be an element because they
have the aura. They have the thing that brings the young people
into it more than the unmanned programs do. But the unmanned
programs and the manned programs go together. So a better
articulation of the total program. The targeting of something
like a Mars stretch or something such as that, like the Rumsfeld
approach, get out in front of the pitch, go out --
DR. WIDNALL: George, I specifically ruled out the space
MR. JEFFS: Oh, you did.
DR. WIDNALL: Yeah, I did. I really wanted to talk more
comprehensively about our whole society, science and technology
and our young people. I think obviously I think we all understand
the power of space.
DR. SILVEIRA: As you know, the President has charged
Missile Defense Agency with a deployment capability into '04,
beginning of '05. That's a pretty big technical challenge.
ADM. GEHMAN: Let me ask a question that I think is
related. Once again, going back to your experience in Gemini
and Apollo and Spacelab. These programs were not exactly heel-and-toe
programs. There was a little overlap among those programs
and people migrated and people learned and worked their way
up through the process.
In your judgment, what's a generation in a space vehicle?
In other words, how long do you think that we should stay
with a space vehicle and how big a leap do you need to make
to have its replacement come along? Is 20 years, 25 years,
40 years a generation, and should we have a replacement program
already have been started? What's the time frame here and
what are the indications or the characteristics of when it's
time to say that's a generation? You've all heard of Moore's
law that a generation in computing power is 18 months. Well,
what's a generation in a space vehicle?
MR. THOMPSON: Let me make a jump at that because I've
thought about this a little bit in my own career. In my working
career, I spent the first 11 years in basic research at a
research laboratory and, frankly, I was beginning to not get
burned out but I was ready for change. The space program came
along. I got in the space program; and we did Mercury in about
four years, as I recall, from the time we started talking
about it until we had finished it. Before we finished that,
we took on Gemini; and we finished that in maybe five. Let
me just pick a number. Five or six years. Before we finished
that, we had Apollo. We did Apollo in ten years. We then bootstrapped
Skylab in there for three or four years, using the residual
Apollo hardware. So during that 20 years, you know, I never
spent more than ten years in any one focused area -- sometimes
as few as four, sometimes as many as ten.
When we took on the shuttle, Skylab and Apollo/Soyuz were
the only things in town, and we had a gap of activity of three
or four years, five years where we didn't fly anything from
Soyuz until we flew the shuttle. But that ten years was a
very strong development cycle. So for people at least like
myself, there was an interesting activity every four to ten
years that lasted anywhere from four to ten years. So you
could jump from one to the other and grow as you jumped.
Now, if the country does not take on those kind of programs
and you say stick with the shuttle for 50 years, then you
have to find some way, internal to that, to keep people excited.
Maybe you do it somewhat like the military does, by rotating
them every three years or rotating them every --
MR. JEFFS: Two months.
MR. THOMPSON: Again, the military found out in the
R&D program it didn't want to rotate them as much because
they lost the technical competence. So if it's not possible
for the nation to throw an exciting new program out there
every five years, then you have to look for some other motivation
below there. I would say ten years in any one kind of an assignment
is probably enough for most people and they need to go do
something either more complex or something different. But
that's just a wild guess.
MR. JEFFS: These programs cost a lot of money; and
therefore when you start them, you better darn well make sure
you've figured out what you want to do with them and what
you're trying to do with the programs. That's kind of Item
The other thing is that these programs are often paced not
by money and talent but they're also paced by technology.
So there's no point in taking off on a single stage to orbit
if you don't have an engine that can perform that kind of
mission. So we go charging off and we all get together and
say, "Let's go single stage to orbit." Then say, "Well, that's
great but how do we get there? Oars."
So therefore you've got to look at the technology base as
it permits you to make decisions for the next generation.
So I think, like Bob, it seems like it's five years, Gemini;
10, 15 on Apollo; 15, 20, maybe 25 on shuttle. The next one
is going to be longer than that. But it's going to have the
technology behind it that enables you to commit that kind
of funding and that duration of lifetime of people to it.
MR. COHEN: I think there are things you can do. In
fact, things have been thought of that you can do is to in
some way combine the talents of the human exploration program
and the robotic program for Mars exploration and bring the
human element of the program involved in that. I think those
are things I think you could do.
I mean, one time we looked at a Mars sample return mission,
JSC working hand in hand with JPL to do a Mars sample return.
It never did come to fruition, but I think things like that
would really create the interest and keep the people sharp
and keep people very interested.
DR. SILVEIRA: When you consider that the shuttle is
a first-generation vehicle, first of its kind, you would think
-- and I know a lot of the mistakes we made in the design
initially that we have found out as a result of flying the
vehicle. You would think within a 20-year time period that
we would be coming up with a better design, seeing it's going
to take another ten years to build a vehicle. I think it's
far overdue that we should be into a second-generation vehicle
similar to shuttle.
MR. JEFFS: If you know what you want to do with it.
ADM. GEHMAN: My question was more along the programmatic
and technology angle than it is the human resource angle.
I appreciate what you say, and I agree with what you say.
You've got to challenge people if want to keep good people
working on these things, but it does seem to me that a generation
in space vehicles -- I mean, I can't put a number on it but
I can tell you that it's not zero and I can also tell you
that it's not 40. A generation is someplace in between there;
and if it's some number less than 40 and it takes seven, eight,
nine, ten years to produce this thing, I'm wondering how urgent
it is that we get on with this.
MR. JEFFS: You know, I would like to add one thing
to the previous statement. There are lots of opportunities
that can be identified; and some of them have some very interesting
possibilities, I think. I would commend the agencies and others
from initiating the nuclear engine programs. I think this
is a whole new avenue that's going to open up a lot of possibilities.
I think that the idea of coming up with some engine that will
essentially be unto itself, a turbojet or engine, a rocket
and the whole schmeer in one swoop is an excellent kind of
focus if there's feasibility basis behind it.
Those are the kinds of things that will offer the opportunity
to identify these kinds of program. If I were going to try
and build a new orbiter today I would do a few things differently,
but I don't think the machine would be a heck of a lot different
than before. It might have titanium in it instead of aluminum,
for example. It might have a more rugged tile system, even
though the one we've got is adequate. There might be a lot
of things that we could do with it that would make it a better
racehorse, but it would be in the thoroughbreds instead of
the claimers or something. You know, it's not going to be
that big step forward. But those other kinds of things like
the engines and so on, nuclear engines and so on, those are
the things that are going to offer the opportunities for us.
ADM. GEHMAN: Thank you for that. Assuming that if we
could cast off to the side, for example -- this is argumentative,
so you just have to make an assumption with me here -- if
we could cast off to the side that the next step that we make
in space has to be a leap -- I mean, why can't it be a tiny
step? You know, aircraft developed by evolution. We didn't
go from the Wright flyer to the 747. We went in many, many,
many evolutionary steps.
So I hear this all the time that, well, you've got to stay
with the shuttle because the next giant leap is not there
in front of us. I don't find that to be completely compelling.
The President has already said that man is going to continue
his journey in and out of space. Is there any reason why we
can't do that journey in an evolutionary way, that we have
to have some big, giant leap in technology to do it?
MR. JEFFS: No, but it has to be enticing enough for
the new generation of people coming along to want to dedicate
their lives to it. We're already losing our capabilities now
on the one we've got. It's not sexy enough. It's not exciting
MR. THOMPSON: Well, let me argue with that a little
bit. I tried to allude to this before. When Nixon made the
decision, the so-called low earth orbital infrastructure decision
that I spoke about earlier, there was no big national-level
discussion of it or national-level announcement of it or national-level
description of it. So a lot of attention was not drawn to
it. Part of the reason, politically you were proposing to
do something that was considerably less expenditure, less
effort, less glamorous than the Apollo program. So compared
to what Kennedy did with the Apollo program, announcing a
low earth orbital infrastructure wasn't nearly that sexy,
so to speak. Plus, the personality of the man, he wasn't that
interested in space. So he didn't make a big to-do about it.
There is plenty about what we're doing today and what he will
do in the next 10, 15 years that should excite a lot of capable
people to work on it, even though it's not exploring Mars.
I frankly think it will be a long time before you can convince
any Congress to spend the money to embark on a properly thought-out
Mars exploration mission because it's going to be extremely
costly and there's going to a hell of an argument about whether
it's worth that cost as compared to putting the cost somewhere
So I think what is needed is a little more attention to explaining.
For example, the space station, I think, is a very exciting
program. The thought somewhere in the future of direct solar
conversion to electrical energy with a solar power station
in orbit. The kinds of things you can do in a low earth orbit
with shuttle and space station type vehicles could be made
into a very exciting program.
Part of the problem is that people want to throw that aside
and go to Mars for some reason, and we've got to put the defense
in that because I think where the nation's going to spend
its money for the next several years in manned space flight
is going to be in low earth orbit and we'd better start explaining
the beauty of it and I don't think you're going to have any
trouble getting plenty of people to work on it, good people,
if you'll talk about it and explain it properly.
MR. JEFFS: The only addition to that is that Apollo
dragged with it a lot of technology. A lot of technology came
with Apollo. A lot of new businesses came out of Apollo. It
was a stretch and it was an exciting kind of thing. And if
you don't have a stretch, you're not good to drag the technology.
And I think that dragging the technology, forcing it into
the forefront is the thing that's best not only for the space
program but for the nation.
MR. COHEN: In order to do what you say, though, I think
some group or some body, some body of people need to establish
the need for doing it, what is the need, what are you really
trying to accomplish, before you can really move forward to
the next step, I think.
ADM. GEHMAN: Let me close this by asking the last question,
which is a complicated one. My understanding of the glorious
history of space exploration in which you all play an important
role is that over the years the role of the NASA engineer
has migrated in a sense. You read in popular literature that
in the original program that Werner von Braun was accused
of wasting money because when he received components from
contractors, he had his engineers take them apart and put
them back together again. I don't even know if that's true
or not. In any case, those engineers, even though they didn't
build this thing, they now got dirty-fingernails experience;
and as you went through the Gemini program and Apollo program,
a lot of that was in-house work. There was a certain amount
of basic research and basic engineering that was done in house
and some of it was done by contractors and some of it that
was done by contractors was checked by in-house engineers.
Then as we migrate away, more and more of this work is being
done by contractors and less and less of this work is being
done by NASA employees.
So my two-part question. Is this management by subs -- let
me get to my bottom line. Then I'll ask the question.
One of the possible outcomes of this board's work may be some
comment about some kind of a system qualification or a system
recertification that if you were to really fly these orbiters
from one decade, two decades, into their third decade that,
just like a 747 or something, if you're going to extend the
service life of it, you ought to do some kind of a system
qualification or system certification. Well, if there's nobody
at NASA that has that hands-on engineering experience, then
you've got to have contractors do it.
Now, does that get us into a boxed canyon here? Does that
trouble you, or would you think that the style that you all
grew up on in which NASA engineers also had hands-on engineering
experience by some way is either critical or not critical?
A lot of people have said it's not necessary to do that. How
do you feel about that? Particularly in light of a possible
outcome where it's possible that we might have to in some
way formally recertify the three remaining orbiters or requalify,
do we have to do it system by system and who does it?
MR. COHEN: Well, I know when I was center director
of the Johnson Space Center I always liked to have at least
one or two projects, in-house projects where the engineering
talent at the Johnson Space Center was doing the work. I think
that was carried on. I think they went pretty far with one
of the crew rescue vehicles they were designing here at the
Johnson Space Center. They went pretty far with that. So I
think in-house NASA projects or in-house projects at NASA
that they can actually, as Milt said, get their hands dirty
on is very worthwhile; and I think it does teach them an awful
lot. Now, that takes money, it takes emphasis, but I think
some type of steady, continuing having of in-house projects,
I think, is very important. That would answer, I think, part
of your question.
MR. JEFFS: I'd like to make sure the picture is not
painted in some strange fashion here. The NASA guys are the
guys that set the requirements and check the product as it
meets the requirements. Industry is one that puts the product
together. The drawings are all prepared by industry and all
the specs are prepared, all the list of materials. Everything
is built and tested. All the tools are made by industry. Industry
does the job.
Now, if you're going to recertify the vehicle, industry, with
NASA's overview, would be the one that puts together the details
of what that recertification process should constitute and
consist of. So it's not like NASA is doing all the job. NASA
is a supervisor and an overviewer. Industry is the one that
does the job.
I'd also like to say that you made some comment earlier about
testing and checking. On occasion we've had to check NASA
tests. Every once in a while NASA runs some pretty strange
tests, too. So we've had to straighten that out. So it's both
ADM. GEHMAN: It is both sides, but it is healthy.
MR. THOMPSON: You do, though, need to have -- what
George says is exactly correct. Nowhere in our manned space
flight experience, except extremely early in the Mercury program,
did NASA sit down and do the drawings and build in NASA shops
a spacecraft. The first spacecraft we flew in Mercury, we
actually designed with civil servants in the Langley Research
Center. We built it in the Langley Research Center shops with
civil service people and we took it down with the support
of the Air Force and launched it on an Air Force rocket at
the Cape and got our early Mercury data off of a thing called
Big Joe. From that point onward, the people who do the drawings,
the people who do the detailed internal stress analysis, the
people who do the certification, formal certifications at
all level, that is industry's job. That's what you contract
My point I would like to make is you need to contract with
them in such a way that they can bring their talents to the
program effectively, but you have to leave the government
in a proper control mode in that contracting format. If you
contract in such a way that it isolates the government from
some feeling of responsibility or some feeling to need what's
going on or some reason to make critical decisions, then you've
backed the government out too far. For example, if you take
all of the contractors working on shuttle and assign them
under one integration contractor and give him all those contracts
to run, that's fine; but you haven't gone down to one contractor.
You've gone from 0 to 81 contractors, and you then have to
back the government off to let that contractor assume a certain
level. Otherwise, you might as well stick with the government
and 80 contractors if you're going to still penetrate to where
you are. But you also have to set up the contracting channels
properly and the responsibilities properly.
I personally favor something much more like we had in shuttle
where, for example, no contractor in shuttle had the leverage
over the other contractors. Rockwell could not go tell Martin
to do anything from the orbiter to the tank. It had to come
through a government channel to get something done, and the
government then was in a very knowledgeable and in a very
controlled position to do it that way. It puts a responsibility
upon the government that you've got to be prepared to fulfill,
but I think it keeps you involved in a much more meaningful
Typically, in my judgment, in the earlier years, NASA penetrated
the program probably a notch lower than the military DOD typically
penetrated their programs. The NASA that I knew did not need
the aerospace support to the same level that the Air Force
needed aerospace support on the ballistic missile program.
Either way, you can make it work; but you ought to decide
which way you're doing it and make sure you make it perfectly
clear. And I would very much like to see NASA retain a capability
to penetrate the programs relatively deeply.
MR. JEFFS: I'd like to make a comment on one other
statement that you made. That was about hands-on. I think
hands-on is a fundamental need for the engineers on both sides
of the fence, both the NASA and industry. One of the classic
examples was to take thermodynamics people down and show them
the hardware that they were actually influencing, changing,
and controlling the configuration of. It's a revelation to
those. You find the aerodynamics guys, thermo guys and so
on tend to get remote from the program and work with just
paper. Get them out and show them the hardware and it gives
you a better project, a better person that's working on it
engineering-wise, and he has greater accountability and responsibility
for it. So that's true on both sides.
MR. MORRIS: I'd like to build on that a little bit,
if I could. I think NASA in particular needs to be very careful
that they retain smart management. I think, to do that, they
have to come up through the ranks with a few dirty fingernails,
maybe even greasy fingers. One of the things that really upset
me was the cancellation of the X-38 project, the recovery
vehicle that Aaron was talking about. This was a chance for
the people working for NASA to actually understand how you
go make something happen. By doing that, they then become
much smarter managers.
I think at the time NASA pulled away from management in detail
-- and there were a lot of good reasons to do that -- there
was then at the same time a promise made that research and
development internal would be increased, and increased materially.
I don't think that's happened. Therefore I think the NASA
personnel have lost out both ways over a period of time. They
no longer are managing in detail and they are not backing
up, in research and prototype development, the experience
level within the organization that they really need.
ADM. GEHMAN: Well, thank you very much, Mr. Silveira,
Mr. Morris, Mr. Jeffs, Mr. Thompson, Mr. Cohen. We thank you
very much for joining us here today. We thank you very much
for your open and candid discussions of all these issues.
As you can see, the board has a fairly wide aperture about
what we are going to write in our report. They include such
matters as you have discussed with us today; and your background
knowledge is still valuable, still of great benefit to the
nation. I thank you very much for agreeing to contribute it
here in such an open forum. We really appreciate it very much,
and we wish you all the best of luck. Thanks very much.
We will reconvene at 1:00 o'clock.
(Luncheon recess, 12:14 p.m.)
Back to April 23 Transcripts