NASA STS Recordation
Oral History Project
Edited Oral History Transcript
Otto K. Goetz
Interviewed by Jennifer Ross-Nazzal
Huntsville, Alabama – 20 July 2010
Today is July 20th, 2010. This interview is being conducted with Otto
Goetz in Huntsville, Alabama, as part of the STS Recordation Oral
History Project. The interviewer is Jennifer Ross-Nazzal. Mr. Goetz,
would you begin by briefly describing your career with NASA?
I hired into the Marshall Space Flight Center in 1962. At that time
I was assigned to the test laboratory and was a lead engineer for
the testing of the combustion devices and the turbopumps of the H-1,
F-1, and J-2 engines. The F-1 and the J-2 powered the Saturn V [rocket]
to the Moon. I worked on those till 1971-2 timeframe, and at that
time I was assigned to the evaluation board of the Space Shuttle main
engine [SSME]. I participated in the proposal evaluation, and at that
time I transferred from the test laboratory to the propulsion laboratory
and again was the lead engineer for the combustion devices and the
turbomachinery for the Space Shuttle main engine up until the Challenger
accident [STS 51-L].
After the Challenger accident I became the chief engineer of the Space
Shuttle main engine. I remained the chief engineer till ’92
[when] I was assigned the program manager for the SSME which I remained
till ’, and I retired in ’96. I conceived, while I
was the chief engineer, of the Block I and the Block II engines. That
is to improve the Space Shuttle main engine, the reliability. [From
’94 to ’96] I concentrated on [certifying and] flying
the first Block I engine, and then I retired.
It’s quite a long and varied career with the main engines. That’s
what I’d like to focus on today. Would you tell me about the
work that you did with the subsynchronous whirl? I understand that
you did a lot of investigations into that problem.
At the beginning of the SSME program when we fired up the engine,
that is the integrated technology test bed down in Mississippi [NASA
Stennis Space Center], we found that high pressure fuel turbopump
had a vibration problem, a subsynchronous whirl problem. The rotor
vibrates and whirls at a speed lower than the actual rotation of the
speed. That vibration was very violent, and we could not proceed for
almost a year into the testing of the engine at higher power levels.
We had [redline] vibration sensors, accelerometers, on the turbopump
and they cut us off. It was something that was not expected. We got
consultants from universities, even from England we had a consultant,
what to do about the subsynchronous whirl. We struggled for almost
a year in order to solve it.
The solution was multifold, but it was primarily a [bearing stiffness,
damping and] rotor balancing problem. It was a rotor stack problem.
The rotor consisted of multiple parts that got bolted together with
a tie bolt that was hydraulically rammed, and then we ran the nut
down. We had stiffness problems in the bearing loads [path] to the
housing that were too soft, we had to stiffen [them] up.
We also invented, in the process of solving the subsynchronous whirl,
the damping seals. We put damping seals around the carriers of the
bearings and obtained additional damping through those damping seals.
The patent was initiated during that investigation. We had a great
program at Caltech in California [California Institute of Technology,
Pasadena] where we explored the damping seals in regard to their behavior,
in regard to cross-coupling, stiffness, and damping. With those data
that we derived in Caltech we implemented damping seals in the turbopump,
and finally we solved the problem and it went away. [Parallel we developed
bearing models to understand bearing mechanics, heating, etc.]
Later on, since you mention subsynchronous whirl, we also detected—that
was much later—subsynchronous whirl in the high pressure oxidizer
pump. The same solutions were applied. We implemented damping seals
in the oxidizer pump. A little bit different concept, but essentially
the same engineering principle.
Were there other task forces that you served on for the main engine?
Were there other problems that you were looking to solve that you
There were quite a few. Following the subsynchronous whirl, we detected
cracks in the turbine blades in the high pressure fuel pump. That
was the next big challenge that needed to be solved because any turbine
blade failure is catastrophic. We had some catastrophic failures with
the high pressure fuel pump. I brought statistics with me of those
failures, and I can tell you that as soon as 1977 we had turbine blade
failures in the high pressure fuel pump.
There again, that was quite an engineering effort to solve these problems.
I don’t know whether you’re familiar with it or not, but
the turbine blades originally were a high strength material that was
cast such that the crystals grew in the direction of the centrifugal
force. But there were individual grains you could see in the turbine
blade. What we did is we went and had a very intense material development
program and converted those directionally solidified turbine blades
into single-crystal blades. One crystal was a whole turbine blade.
That was quite an effort to develop that. We had quite a few materials
people and a contractor involved in that. That was one of the things.
The other thing that we needed to do was we needed to coat the turbine
blades, because in the start sequence of the Space Shuttle main engine,
when the preburner ignited and shot the hot gas into the turbine in
order to get it going, the turbine blade was essentially at ambient
temperature and the hot gas coming out of the preburner was 2,000
degrees. So within two seconds the blade saw a delta temperature of
2,000 degrees. That is, the core of the blade was still at ambient
[temperature] while the surface, the skin, went to 2,000 degrees and
that caused cracking because the delta internal to the material caused
high stresses. One of the other solutions was to coat the turbine
blades. We coated the turbine blades with an insulating material in
order to keep the high delta T for a little while away from the core
of the turbine blade.
We also had to develop dampers. The turbine blade was originally designed
by Rocketdyne with dampers, but the damper was too heavy. When the
turbine rotates, the heavy damper locked up, and therefore there was
no damping. The damping was by rubbing [between damper and blade],
and so we had to design very light dampers in order to effect the
damping, the rubbing.
I don’t know whether you realize, but the fuel turbopump had
a speed of 36,000 rpm [revolutions per minute], which means it has
to rotate 600 times a second. The turbine blade itself, I think there
were 58 turbine blades on it, but the turbine blade itself was [1.3]
ounces. When it rotates at 36,000 rpm, the blade weighs  tons because
of the centrifugal force. That illustrates how a too stiff damper
locks up; you have to make it real light in order for it to effect
the scrubbing between the two blades.
That’s amazing, all of the work that you did to solve those
There were other problems with the fuel turbopump and with the LOX
[liquid oxygen] pump. The thing is that people were afraid the pump
wouldn’t pump or the turbine wouldn’t turbine. And none
of that happened. The problems were with bearings, with seals and
with dampers in the turbine blades. They were all mechanical [in]
nature and materials.
For a long time we struggled. This is not only related to the turbopumps
but also to the whole engine. When we started the SSME we did not
realize how sensitive some of the materials are to pure hydrogen.
We had a hydrogen embrittlement problem for quite a while and they
had to change materials or had to plate material. Like, for example,
the turbine disks—where the blades are mounted on the outer
periphery—they were all gold-plated in order to prevent the
hydrogen from diffusing into the material. If the hydrogen diffused
in that material, the material becomes brittle and cracks. Then you
buy the farm.
Was there ever a point, Mr. Goetz, that you thought you wouldn’t
be able to solve whatever problem or issue you were facing at that
Well, we just addressed one problem at a time, but sometimes we had
more than one on our hands. There were multiple problems. During the
development program we lost 27 engines. Exploded—and sometimes
we found the engine down in the flame bucket after an explosion.
Tell me about the development and design period for the SSME. Then
move into the test program that you laid out for the engine if you
When the development program was structured the approach was to test
every component on its own, develop it, and have a certain maturity
before the component goes onto an engine and then test the engine
system. So we built COCA [test sites]—this is up in Santa Susana
in California, the facility there—in order to be able to test
the preburners on its own, the turbopumps on its own—of course
you need the preburners to fire the turbopumps—the main injector
and the combustion chamber.
I was very much involved. That is a pressure-fed system. The SSME,
the combustion chamber pressure is 3,000 psi [pounds per square inch].
The preburner itself runs at approximately 5,000 psi. In order to
get that pressure with a pressure-fed system without a turbopump,
you have to have a very high supply pressure. We had hydrogen bottles
that didn’t meet the code. We needed 14,000 psi hydrogen so
we had to put the bottles into caves, into the mountain, in order
to be safe if something happens. Then we ran the pipe to the test
stand and pressurized the run tanks, the liquid hydrogen and the oxygen
tank, and then fired the preburner. We successfully developed the
oxidizer preburner. That was a congressional milestone where Congress
looked upon us every day; we had to report to [NASA] Headquarters
[Washington, DC] how far we are on that thing.
Then we started to [test] turbopumps with the LOX pump. We had an
explosion, facility-caused. At that time we realized that this whole
facility, with the extremely high pressures and its facility control
system, is much more complicated than the engine itself. Therefore
we abandoned the approach of testing every component as a component
before it’s moved onto an engine system. Essentially we then
moved to ISTB (the integrated [system] test bed engine [program])
in Mississippi. We continued to do some testing at COCA, however,
only on smaller items, like for example the main combustion chamber
and each one of the preburners [which] had an auxiliary spark igniter,
which is a little burner. That lent itself to being continued to be
tested out there, but we did not test the turbopump or the main combustion
We had two test stands up there, one primarily related to oxygen and
the other one related to hydrogen. Later on we converted one of the
test stands into an engine test facility, and we did some testing
up there. We allowed Rocketdyne to have their own backyard test facility
so they can test up there, and they were responsible for the testing
[in California]. NASA, we concentrated on our testing in Mississippi
on A-1 and A-2 and later on on B-1 [Stennis test stands].
When you were testing components at COCA were you requiring them to
be run for 520 seconds?
Oh no, that is completely impossible with a facility pressure-fed
system, because there was not enough [supply] pressure. With 14,000
psi, for example, the pipe that ran from the underground facility
to the test stand was about 14 inches diameter, but it had a wall
thickness of six inches with a bore of two to three inches because
of extremely high pressure. You had to have these tremendous pipes.
There is not enough capacity of pressurant in order to make 520 seconds.
That’s impossible. We realized that from the beginning that
one full run demonstration on a component basis is impossible.
I need to modify that—because when we developed the Block I
and Block II turbopumps at Pratt & Whitney down at West Palm Beach
[Florida], we wanted to test the turbopumps there too, and we did.
But we did it completely different in the approach as compared to
the COCA facility. We fired the preburner and let the pump rotate,
and then we took the pump discharge to feed the preburner, just like
in the engine. That’s how we circumvented the problem of having
enough pressurant to fire the preburner. We didn’t make 520
down at Pratt & Whitney either, but it was a different concept
in that the turbopump fed the preburner instead of pressurant from
a facility feeding the preburner.
You mentioned the integrated subsystem test program. Can you talk
The ISTB [with a lower expansion ratio nozzle] was primarily a tool
to develop the engine start sequence and the engine shutdown sequence.
That sequence is rather complicated because the SSME was a feedback-controlled
system, thrust and mixture ratio, so the ISTB was used to develop
the start sequence and to develop the shutdown and the safe shutdown.
However, the whole effort was for a long time handicapped because
of the subsynchronous whirl of the fuel pump. We could not get to
100% power level. We were stuck at low power level for quite some
And what is low power level?
The specification originally, when we wrote the request for proposal,
was that the engine be operated from 50% to 109%. So low power level
is 50%, but we never flew at 50% because we had some vibration problem.
As you know even today when the engine flies and the engines are throttled
back during max Q [maximum dynamic pressure], they only go to 65;
you cannot go to 50 because of vibration problems. And it’s
not necessary anymore, so it was overspecified at the beginning in
Could you describe testing with the main propulsion test article?
I was really not involved closely. I was only involved from an engine
standpoint. The testing itself was very structured as I recall. [To
start the program] we ignited one engine first and then two engines
and then three engines. I cannot recall all the details, but it was
a very successful program in that the propulsion system, the feed
system to the engines—that was the main purpose of testing—worked
quite well. We had some problems. One day I got a call, and we flew
down after a test on the article, and [on] one of the engines, the
steerhorn was missing. So we had a major mishap on one of the engines,
but other than that I think it was a good program in order to test
the system between engine inlet and tank outlet.
I think every program that I was involved in—and I’m still
involved a little bit in some programs—there is concern about
what is called the propulsion system, that is the feed lines, the
pressurization system, the control system and what have you. A test
program like that is very much in order. I think in my experience
I would not develop a vehicle without a propulsion system test program.
Did you do any work with the National Research Council and their study
on the SSME?
No, not really. I made at one time a presentation to the National
Academy of Sciences, but only one time. It was following the Challenger
accident. We had to implement quite a bit of correction, even on the
SSME, after the Challenger accident even though the SSME was not involved
in the accident at all.
After the Challenger accident Dr. Richard [P.] Feynman, the Nobel
physics man, came to Marshall. He was interested in the other elements
other than the solid rocket motor that failed during the Challenger.
We had a session with Richard Feynman, and we explained to him the
turbine blade problem, cracking. He was very much interested in that,
but eventually the session evolved into a discussion of reliability
of the Space Shuttle main engine. In the course of the discussion,
the Marshall management had one number of reliability. I recall very
vividly that it was said that the reliability of the engine is one
failure in 10,000 missions, and Feynman leaned back in his chair and
said, “You mean to tell me that you can fly every day [for]
three years and never have a failure? I don’t believe that.”
So he challenged the whole thing.
He also then asked us—and I was [in] engineering at that time,
I was not management. He took a piece of paper and made little pieces
out of it, and he asked the management to leave the room, and he said,
“Write down what you think the reliability is.” Based
on the 27 failures that we had on the engine I had kept my own statistics,
and I wrote down one in 300. If you read the Feynman appendix to the
Rogers Commission Report [on the Challenger accident], in Appendix
F he mentioned the one in 300. From then on we had our task cut out.
I then was transferred and became the chief engineer, and we initiated
redesigns. I think we had over 30, some of them very minor. We initiated
failure mode and effects analysis, the critical items list—we
restructured our criteria as to when is it safe to fly and what factors
apply. For example, the FAA [Federal Aviation Administration] has
a criteria, a factor of two. That factor of two was, before Challenger,
on the SSME a little bit loosely applied. It means that if you have,
for example, a turbopump of a specific design, and you want to fly
the pump five times, you have to have a pump on the test stand with
exactly the same design that made ten equivalent flight durations.
That’s what the factor of two means. Post-Challenger we applied
that very rigorously, while before Challenger we tolerated some deviations
from that criteria.
This was the Phase II engine?
The Phase II engine followed the Challenger accident. This is all
of the history of the development. [Demonstrates] At first we only
flew 100% power level, for the first five flights of the Space Shuttle.
Then we went into Phase I, which allowed the flight to be at 104,
and we flew up to the Challenger accident with this 104. We initiated
the Phase II after Challenger and tried to go to 109% and never succeeded.
We could not make the engine go to 109 safely.
We made 38 changes after the Challenger accident, and then we made
some turbopump improvements, and then we followed with the Block I
and Block II engine. That is what is flying today. We implemented
quite a few changes in order to offload the high pressure system within
the engine. The chamber pressure originally was 3,000 psi. With the
Block I, Block II we went to a combustion chamber with a larger throat,
and that reduced that pressure by about 10% to 2,700. Of course we
made other improvements like the alternate turbopumps from Pratt &
Whitney—the LOX pump and the fuel pump—the two-duct hot
gas manifold, single-coil heat exchanger, and the large throat main
That was quite a development effort in order to do this. At the beginning
of the Shuttle Program most people were mostly afraid of the Space
Shuttle main engine because of the extreme power density that the
engine has and the high pressures. Everybody expected that if there
is a failure of the Shuttle it may be caused by the engine. Thank
goodness the engine never failed on a flight. We had one premature
cutoff on one of the engines during a flight, and that was due to
a sensor failure. It was no big deal, but we made orbit because the
trajectories are designed that from a certain point on you can cut
off one engine and from the next point on you can cut off two and
you still make orbit with one engine. Before that we always had the
thing go to Africa or go to Spain. Have you ever attended a Shuttle
There is a call when it’s safe to go to transatlantic abort,
and then there’s a call abort to orbit. This all is engine-related.
You mentioned that the development effort for the Block I and the
Block II engines was huge. Can you tell me about that effort? Can
you give some examples of how the Block I and Block II was developed
and then tested?
Originally we thought that we could do the whole thing together. Due
to some fiscal constraints we did it a little bit in sequence, and
that’s why there is a Block I and there is a Block II. The major
effort—well, all of it was a major effort—was to replace
the turbopumps, because the turbopumps had to be removed after every
flight from the vehicle. In order to do that the engine had to be
removed from the Shuttle. That was very costly, and we wanted to get
higher reliability; we wanted to get longer time between overhaul.
That’s why we conceived the Block I, Block II.
The turbopumps were developed down by Pratt & Whitney, designed
and partially tested at the facility that I mentioned earlier. The
main development was on an engine system at Mississippi, both LOX
pumps and fuel pumps. The original SSME had a hot gas manifold we
call it, where the two preburner and then the two pumps were mounted.
Then from the turbine discharge, the hot gas goes into the main injector.
It’s mixed with liquid oxygen and burned. That’s how you
The design of the hot gas manifold was such that the feedback from
it into the turbine, especially on the fuel side, was such that there
was a mal-distribution of pressure going into the turbine, and there
was a mal-distribution of temperature. Both of them affected, for
example, the turbine blade life. The prime objective for redesign
was in order to correct the distribution of the temperature and the
pressure going into the hot gas manifold that affected the turbine.
Of course originally when we designed the SSME in the early ’70s,
we didn’t have the computer programs, and we didn’t have
all the sophisticated analysis techniques that we have today.
In the redesign for Block I and Block II we really made use of these
sophisticated tools, the models, calculated pressure distributions,
flow distributions, and temperature distributions and let some of
those models guide us in the redesign of the flow system from [preburner
to] turbine [and] into the main injector. That was very successful.
The use of these modern tools was a key factor in the successful development.
We did not have many problems on the test stand when we tried to test
the redesigns for the first time. Too bad we didn’t have these
tools at the beginning. It really helps. Those tools are being used
today also on the J-2X and the other engine development that are being
Did these tools shorten the design, development, and testing program
for the Block I and Block II engines?
I don’t know whether these tools shortened the design phase.
I am tempted to say yes, but you still have to design. These tools
were very effective to do it right the first time, and you didn’t
have to go around and around and redesign here and redesign there.
So the tools were, in my judgment, very effective in shortening the
development program in that it cut out a lot of trial and error. The
SSME at the very beginning relied mostly on testing to prove is the
design correct or not, and in quite a few instances it was not. While
in the Block I and Block II, thanks to the modern tools, those iterations
were not necessary.
It was quite helpful to have these tools. Of course I need to say
that some of these tools you cannot just take what the computer spits
out at face value. Some of the tools had to be verified, and at Marshall
we had a hot gas test facility where we essentially calibrated the
tools. Not necessarily at the same condition as the SSME, but the
tools were calibrated, and you gained confidence that yes indeed,
what the computer tells you is correct. So we insisted on some calibration
effort of those tools. It was not just taking them at face value,
because you could be misled by computer outputs, but I have to say
that these tools helped a lot in Block I and Block II.
Would you tell me about the flight acceptance of the various engines
as they evolved?
Each flight engine went through a flight acceptance test down in Mississippi.
We had a prescribed program of a series of tests that the engine had
to go through in order to be able to be flown. For example, one of
the test programs was a stair step test where you go from 65% power
level in increments to 104 in order to show that there is no vibratory
or any other problem hidden in the engine. Some people think that
from a vibration standpoint the worst case is the highest power level.
That’s not necessarily the case. You can have vibration problems
or coupling at low power levels while the high power levels are okay,
and that’s why the engine had to go through these steps, [to]
demonstrate that there is no hidden vibration problems.
Then you may have heard about the fleet leader concept. I mentioned
that earlier with the factor of two. Any flight engine had to have
a test engine demonstrated by a factor of two, the duration and the
power levels that the flight engine was supposed to see. Originally
if we had a failure on something we asked ourselves, “What’s
the effect on the flight engine?” Usually the failure was addressed,
and then we applied a factor. Depending on the failure it was either
a factor of two or even a factor of four sometimes, depending on the
failure analysis, [our knowledge] and what we found [in the hardware]
in regard to the failure.
For example Feynman, when we discussed with him the blade cracking,
we got into an argument. “What is a failure? Is the failure
when the blade flies off? Or is the failure when the crack starts?”
It was his opinion that the failure is when the crack starts. We took
that to heart and examined quite a few turbopumps, and then after
the Challenger accident structured the limit of the fuel pump exposure
to that criteria.
Later on we modified it somewhat because the crack can grow and not
grow to a catastrophic failure, which is fracture mechanics. Of course,
in order to make a case with fracture mechanics you have to have very
solid material properties, crack growth rates, and understand the
environment exactly before you can say, “I let a crack grow
in flight.” That was quite an engineering effort in order to
arrive [at the conclusion], “Can we or can we not tolerate crack
growth?” On some of the locations in the SSME we tolerated crack
growth, and some we did not.
What areas did you allow for crack growth, and what areas didn’t
For example on the main injector, there is an [interpropellant] plate
that separates the hot gases from the liquid oxygen. There are cracks,
and those cracks grow. You have to have a very good analysis, understanding
the material what have you, but the cracks grow till they reach a
neutral fiber where the stress field transitions from tension to compression.
Where you have compression, essentially the crack does not grow. Where
you have tension, the crack grows. You have to [have an] exact good
understanding as to where the neutral fiber is. If you cut apart a
few of those items and demonstrate that yes indeed, the crack has
stopped by inspection, by cross section of the material and by analysis,
you can tolerate a crack up to that point. While in other cases where
there is no neutral fiber, a crack cannot be tolerated, period.
How did the initial engine perform on the first flight, on STS-1?
Perfect, on STS-1.
Did you make any changes as a result?
Well, that’s a long time ago now, in ’81 in April, but
I do not recall any major problem on STS-1. If it would have been
a major problem I would remember it. It flew at 100%, not at 104,
as I indicated earlier.
Same with the return to flight and the Phase II engine performed perfectly?
Yes. But as I said, during the whole flight program we had one premature
cutoff. [STS] 51-F was the flight I think.
I believe you’re right, because it was an abort to orbit.
It was rather interesting, that flight, because a temp [temperature]
sensor failed and on the second engine another temp sensor failed.
We inhibited the cutoff and flew with two engines, and it turned out
that was a benign failure of the sensors. If you predict, from a reliability
standpoint, that two sensors fail in the same flight on two different
engines—it was very extremely remote, yet it happened.
Did Mission Control call you at that point? Or were you in the Launch
Control Center to advise?
I imagine things were a little tense at that point.
Oh yes. Everybody was sitting on [pins and needles; it] was quite
Going back to the transitioning from Phase II to Block I. The discharge
volutes of the turbopumps and the inlet on the turbopumps, the manifolds
on the main combustion chamber, the outlet and the inlet for the coolant
for the main combustion chamber—they were all originally built
by welding together forged segments. You had little pieces of forged
segments that fit together but that had to be welded. These welds
caused a lot of problems. They had to be X-rayed, dye-penetrant inspected,
and what have you. Cost a lot of money, was very very expensive and
time-consuming. In the Block I and II we specified to get rid of these
welds, and we developed with a contractor a fine-grain casting of
For example, we went from more than 500 welds on one of the turbopumps
to zero welds in Block I and Block II by eliminating the technique
of putting together, almost like a puzzle, these single pieces and
welding them together to volutes and to manifolds. This development
of the fine-grain casting is, in my judgment, a major achievement
in not only reducing manufacturing time, saving cost, and what have
you, but also in increasing the reliability of the engine. Because
we never knew exactly, did we or did we not get everything by X-raying
these welds, and are there left over cracks or not? With fine-grain
casting all those concerns were essentially eliminated, and in my
judgment that was one of the major achievements.
In addition to that, for example, I mentioned earlier that at the
beginning when we had the turbine blade cracking problem that we went
to coating the blades. The Block I and Block II turbine blades don’t
have any coating. We went to a different material, and that material
does not need any coating. There’s all kinds of disciplines
involved in improving the engine, stress dynamics, thermodynamics,
materials. There’s a lot of inputs and a lot of coordination
required between the various disciplines in order to achieve something
Do you have any idea of how much these improvements changed maintenance
time or overhauling of engines? Does anyone keep those type of figures?
Our goal was that we fly ten times without removing the turbopumps.
I really don’t know now exactly what the criteria is today,
how many times they can fly. The latest I had an input was that they
were flying six times. I don’t know if you are familiar with
the Cape [Canaveral, Florida], but when they have to remove the engine
on the orbiter in the OPF [Orbiter Processing Facility], it’s
quite a thing to remove engines and then remove turbopumps, put them
in the engine facility, and inspect the turbopumps. Quite time-consuming,
costs money. The latest I have was a factor of six.
How long would it take to process an engine once you took it out of
the OPF and the orbiter itself until you put it back into the orbiter?
I have been down there quite a few times, but that’s hard to
answer. It depends how many engines they have to work on. If there
are no problems and you just inspect and put everything back together
the way it was, I would think that it could be done in four weeks,
but I’m not sure. That’s a guess.
How long might a Space Shuttle main engine last? Did it vary depending
upon the original engine, Phase II, Block I, Block II?
Right now I think the engine can make ten flights, and then you have
to—not refurbish everything, like the main combustion chamber.
At the beginning of the program the temperature distribution out of
the main injector was not ideal. We had what is called blanching on
the main combustion chamber. I don’t know if you are familiar
with it or not, but there’s a copper liner where on one side
you have hot gas of 3,000 degrees and on the other side you have liquid
hydrogen flowing through channels at -420 degrees. The separation
between the two, the 3,000 and the -400, is about  mils.
If that web in the main combustion chamber would break through or
fail then you lose the hydrogen going into the main injector, going
into the preburner, and you essentially lose the engine. First there’s
a performance degradation, and then because [of] the leaking hydrogen.
It had to be that thin because of the heat transfer. You cannot put
a big barrier between the hydrogen coolant and the 3,000 degrees hot
gas in order for it to cool and not melt the copper alloy. You had
to make it as thin as possible, what the structure would allow.
At the beginning we had some blanching. Then we had to take sandpaper
and prepolish and be careful not to get it too thin. That was one
of the limitations that we had to take out. This is not the case anymore
with the Block I and Block II. The temperature distribution is such
that there’s no problem anymore.
Would you tell me what your role was in terms of mission support over
Post-Challenger, when I was chief engineer, I was down at the Cape
during the launches sitting in the firing room with Rocketdyne people
and making decisions. In some cases it was rather stressful because
[if] there was an anomaly, “What do we do?” And then can
you lift off or should you try a hold? In order to cancel a flight
you had to answer a lot of questions. It turned out that only once
did we have to cancel a flight.
Would you tell me about that decision?
There were other problems whereby flights were canceled that were
not related to SSME. For example, hydrogen in the aft compartment.
The hazardous gas was too much, and things like that.
The hydrogen in the low pressure fuel pump—while the propellants
are loaded, that pump is full of hydrogen. We had the pump insulated
but it was apparently not completely adequate. Because of the very
low temperature, like the -400 degrees, the surface was accumulating
[liquid N2 (nitrogen)] from out of the aft compartment atmosphere.
The [LN2 (liquid nitrogen)] was dripping on a sensor that was critical
for liftoff, and the sensor went out of whack so it was inadequate.
Later on we improved the insulation, but that was one of the incidents
where we did not lift off because that was a flight-critical sensor,
and it was completely off range. But only because the dripping of
the extremely cold fell onto the sensor.
Looking back over your years with the main engine, what do you think
was your most significant accomplishment for the program?
I am tempted to say there were a lot of accomplishments, significant
ones. From an emotional standpoint, the most significant affecting
me was the first flight after the Challenger accident [STS-26], even
though it was no problems. We all were very very excited and very
apprehensive because of all, like I said earlier, the 38 modifications
that we had made. We had quite a few meetings down at Kennedy [Space
Center, Florida], [especially with Frederick H. “Rick”
Hauck, the commander] and then the flight readiness review where we
had to go through all we did and how safe it is. I think from a personal
emotional standpoint the reflight after Challenger was to me more
exciting than STS-1. STS-1 was exciting too, but I felt the reflight
Did you work with the return to flight crew at all?
Oh yes. I knew quite a few astronauts, yes. The astronaut corps always
had one guy assigned to the SSME. He participated in all the meetings
either here at Marshall or out at Rocketdyne. People like Dick Richards—Richard
[N.] Richards—John [E.] Blaha and Brewster [H.] Shaw. Those
were all people that I had to interface with. A lot of others, like
[Daniel C.] Brandenstein. Yes, they were very much involved.
One of the things, we put a briefing together with movies and charts
of all the [major] incidents that we had on the engine. You see in
the movie how an engine explodes. Bob [Robert L.] Crippen insisted
that we show and make this briefing to the younger astronauts. I think
we scared some of them.
Was this before [return to flight]?
That was after Challenger, yes. But he insisted that we do that, just
to make them aware how catastrophic a failure can be.
What do you think was the biggest challenge working on the SSME? I
know you have lots, but is there one in particular?
It was a great challenge. It was a great [personal and] professional
challenge to work on the SSME. My background a little bit. I came
here to the country—I worked in Switzerland before I came, and
I worked on pumps and turbines for big power plants. So from an engineering
standpoint I had quite a background on pumps and turbines, but I had
to learn a lot about the other systems on the engine, for example
the control system. I was pretty much a greenhorn on the electrical
systems, and I had to learn a lot. There were quite a few challenges.
Going back to my background, in Switzerland after I graduated from
the university and hired into that company, I had an introduction
by the chief of the company. He told me, among other things, that
it is the policy of the company to design everything for a lifetime
of 25 years. From a fatigue standpoint, design everything for 25 years.
When I transferred over to NASA after I came here, and I attended
a meeting on the F-1 engine. I was a greenhorn a little bit on rocket
engines at that time; I knew about pumps and turbines. There was a
problem with a discharge duct on the turbine. It was a live problem.
I asked the question, “What is the design life of this engine?”
At that time the program manager turned around and looked at me like
“you dummy,” and he said, “It’s 150 seconds.”
Based on 25 years I couldn’t believe it. That’s how much
I had to learn. It’s quite a transition from stationary power
plants to a rocket engine, which is also a power plant.
What brought you to the United States?
Oh, that’s a long story. I wanted to go back to Switzerland
and make more money with US experience, but I never went back. We
had some interfaces with US companies when I worked in Switzerland.
For example, we designed the turbines for the Grand Coulee on the
Columbia River, the big dam there [in Washington state]. That’s
That’s pretty good reliability.
Quite a difference.
Yes. Is there anything we haven’t discussed that you were hoping
we might talk about today about the main engines? I think we’ve
hit pretty much all of the questions that I had prepared.
I think that we touched almost on everything.
All right. Well, thank you very much for your time today.