NASA Johnson Space Center
Oral History Project
Edited Oral History Transcript
Stanley A. Bouslog
Interviewed by Jennifer Ross-Nazzal
Houston, Texas – 22 April 2014
Today is April 22, 2014. This interview with Stan Bouslog is being
conducted for the JSC Oral History Project. The interviewer is Jennifer
Ross-Nazzal, assisted by Sandra Johnson. Thanks again for taking some
time out of your very busy week to meet with us and talk about the
Arc Jet. Looking forward to it. Give us an overview of your career
at JSC, and how you came to be associated with the Arc Jet.
first moved down to Houston and started working in the Houston area,
I think it was 1988. I had graduated from school at the University
of Texas [Austin] with a master’s degree, and actually worked
for Tracor Aerospace, up in Austin. I’d worked there for about
five years, and they, like most other places, were doing defense work.
They were filing for bankruptcy, and so I thought, “Well, okay,
maybe it’s time to move on.” My professor at the University
of Texas, Dr. John [J.] Bertin—he had some contacts down here
at NASA Johnson Space Center. I contacted them and asked if they had
openings. They brought me down. It was actually through Lockheed Martin,
who had the engineering support contract for NASA. I was interviewed
and got the job, and it was in the area of aerothermodynamics. One
of my first mentors was Carl [D.] Scott. Carl Scott was at the Johnson
Space Center, and part of the work I was doing for him was associated
with the Arc Jet facility. Most of it was looking at the lifespan
changes of the ceramic tiles that were on the Space Shuttle Orbiter.
You remember, at that time period, we were just starting to get back
to flight after the Challenger accident [STS-51L], which happened
in 1986. This is the end of ’88, when I came in.
Tell us why you were looking at the life cycles of the tile. Hadn’t
that been proven in the seventies?
really. Every time you fly in space and come back, you learn. We were
learning, still, with the Space Shuttle Orbiter; we were still learning
how to operate that vehicle. The intent was to have it fly—each
one of the Orbiters—at least 100 times. We didn’t have
a lot of information on what would happen to all these tiles which
are part of the thermal protection system and how long they would
last, or how they would change over that many uses. Obviously, we
had some history already because we’d gone through a few flights,
but we didn’t know if they’re going to last 100 flights
or are things going to change. We started looking at especially the
surface properties of the tiles. We were concerned about how they
may degrade, over time, and technical things such as emissivity and
the catalycity of the tiles specifically. That was one of my first
tasks when I came here, to start looking at that. It included using
the Arc Jet to expose those tiles to the simulated re-entry environments.
Had you used an Arc Jet facility before coming here?
I had not. I’d done quite a bit of wind tunnel testing, flight
testing, as a matter of fact. I did wind tunnel testing at University
of Texas, and actually at Texas A&M [University, College Station].
I’d use their wind tunnels mostly in support of Tracor Aerospace,
when we were developing airborne counter-measures, but I had never
used an Arc Jet. I’d done re-entry survivability analyses, because
I had a lot of background in aerothermodynamics and hypersonics at
school, so I’d looked at the survivability of objects re-entering
Earth’s atmosphere from space. I’d looked at that analytically,
but as far as an Arc Jet specifically, no, I did not have any experience.
How was it different from using a wind tunnel?
tunnels tend to be better characterized. The flow tends to be more
uniform. There’s a lot of systems that have been set up over
the decades that people have been using wind tunnels to actually understand
the flow field which you’re exposing your test article to. Not
so much with an Arc Jet, and, of course, the energy levels are so
much higher. Wind tunnels, some are subsonic, some of them are supersonic,
some of them are hypersonic, but you go to an Arc Jet and you’re
talking megawatts of power. The power level is much, much higher,
and the flow fields are not well characterized. You’re not quite
sure what you’re exposing your test article to.
That’s quite a change.
it’s a big change. The other part of it, too, is that your options
for determining or characterizing that flow field are much reduced,
just because of the survivability. You cannot stick everything in
the flow field and expect it to survive; it won’t. There’s
many a time when we stuck probes in, and there was not adequate cooling
water or somebody forgot to turn the cooling water on. Of course,
then you lose the probe or whatever very quickly. Water leaks generate
snow cones in the Arc Jet. There’s a vacuum chamber, so you
dump a lot of water in there and generate snow.
You generate snow?
I wouldn’t think that [was possible].
you can get a lot of it real quick.
How does it generate snow?
because it’s a near-vacuum. We don’t get down to vacuum,
but we get very low pressure in the test chamber, and then you dump
water into that and it wants to expand very quickly and cool off the
water. So it just freezes. You get little particles of ice everywhere,
so it sort of looks like snow.
That’s interesting because, when I was listening to everyone
[at the videotaping], I thought you went up to about 3,000 degrees.
when the arc’s on, that’s correct. Of course, that’s
only in the core of the flow field, and the 3,000 degrees Fahrenheit,
that’s not even real high temperatures for an Arc Jet. That’s
actually surface temperatures. The flow field itself is much, much
higher. Unfortunately, we always use inconsistent units, but we usually
talk about 5-10,000 degrees Kelvin in the actual flow field itself,
where the test articles are placed. We shy away from temperatures
after a while because temperature is just a measure of your energy.
It’s very confusing and not very accurate to talk about temperatures
when you get to real high temperatures. The reason being is that a
lot of chemistry goes on. Now, a lot of energy is tied up in the disassociation
or even ionization of the gas, so it’s not real accurate to
talk about temperatures. We shy away from temperatures, except for
surface temperatures, of the test articles. There, yes, we still talk
about it. Even that’s very difficult for test articles that
ablate, that change state, but we still try to get a measure of the
surface temperatures but it’s much more difficult.
Any memorable tests when you were working with Carl Scott?
we did a lot of testing. I think there was some reluctance amongst
some of the folks that operated the Arc Jet to do a lot of diagnostics
of the flow field, to better understand the flow field—part
of it because they just didn’t understand that. So, myself and
Carl Scott really helped moved that along to get all sorts of diagnostics
in there. When I talk about diagnostics, first of all, we looked at
pressure probes and heat flux probes that we would sweep across the
flow field, so that we’d understand what the flow field would
look like. A lot of times, it would have a lot of gradients in it,
the pressure gradients and enthalpy gradients. By sweeping those probes
across, we would have a better understanding of that. I remember specifically
some runs where we had heat flux probes in the flow field, and somebody
forgot to turn the cooling water on. You’d start the test and
insert the probe, and it was gone pretty quick. Everybody would be
looking at each other, going, “Okay, that was not good.”
There was a few times when that happened; those are sort of memorable
What did you learn about the tiles from your tests out at the Arc
was really complicated. We did a series of tests in the Arc Jet, and
we saw some degradation, but it was sort of within the noise. So we
really didn’t come up with conclusive evidence that we were
seeing any degradation of the performance of the tiles with the tests
that we did. There’s another parameter called emissivity that
you also take a look at, and that’s how much radiation comes
off the surface, re-radiated from the surface. You’d get some
evidence that at some of the temperatures we were getting a reduction
in emittance, which could cause your surface temperatures to rise
a little bit, but it was not a big issue for us.
You mentioned you were using different probes—was this when
you designed the mass spectrometer probe tip?
did do that, also. The intent there was actually to try to get measurements
of the species concentrations in the flow field itself. Like I said
earlier, when you expose these gases to an electrical arc, then quite
often, you disassociate the gas. For example, nitrogen and oxygen
molecules get disassociated into nitrogen and oxygen atoms. The problem
is, you don’t know how much, and of course, you don’t
know if nitric oxide—you don’t know how much of that there
is, either. A lot of chemical reactions going on, very high temperatures,
so you wanted to get an idea of the exact quantities of these different
species there was in the flow field.
We did get with a company and started designing a mass spectrometer.
The intent for that was actually to extract some of the flow field.
You can imagine, you’ve got an extremely high temperature gas,
and now you’re trying to stick a probe in there and pull off
some of that flow, some of that gas, and then measure its species
concentration with the spectrometer. We did generate lots of snow.
You can imagine, you’re very susceptible to water leaks, over-temping
various parts of the hardware. That process actually was not working
very well, and that was, I think, one of the lessons learned was that
we never really got good data out of that. We got some, and we really
pretty much abandoned that mass spectrometer. What we did do, then,
was we turned to laser diagnostics, which is much better in the sense
that it’s non-intrusive, which means that you can actually get
a measurement without putting something in the flow field.
We used a laser-induced fluorescence. Carl Scott and another man,
Dr. Sivaram Arepalli, were very instrumental in starting that at NASA
JSC. I was helping them, mostly on the analytical side, but they were
very, very good about setting up the instrumentation and getting the
hardware, the lasers, et cetera. Again, the intent was to do various
measurements of the flow field. They were getting velocity measurements
very early on, and then they were trying to get atomic oxygen concentrations
measured. Over the years that system had been developed and, up until
just recently, we were getting good data out of that. We called it
a LIF system, the Laser-Induced Fluorescence System. We transitioned
to that system and pretty much gave up on the mass spectrometer.
Is this system going to be used in a different facility, or is it
also going away with the demolition?
much it’s going away. NASA Ames [Research Center, Moffett Field,
California] had had one of these LIF systems, previously. They pretty
much dismantled it. We have requested that they resurrect it, and
so, to our understanding, that is in process. There was a lot of work
done from about 2006 to about 2009, where both JSC and NASA Ames were
co-developing their respective systems in both of their facilities.
We made a lot of progress. I’m not sure why NASA Ames abandoned
theirs, but since that time, since ours is going away, we’ve
asked them to resurrect theirs. I do not know the status of that,
but apparently, they are going to do that, at least in a couple of
their test positions.
You’ve given some examples of what I think are technological
developments as a result of the Arc Jet. Are there other ones that
come to mind?
of the things we often talk about is missions to Mars. Of course,
we’ve sent Rovers there, we’ve sent probes to Mars, but
of course, we also, at some point, would like to do a crewed mission
and let humans set foot on Mars. The problem with that is, of course,
you have to enter the Martian atmosphere, which is carbon dioxide.
We have never really tested our thermal protection systems in carbon
dioxide. We're all pretty much convinced that if we’re going
to send humans to Mars, we better have much more confidence in our
thermal protection systems than we do today. To do that, we would
have to test our thermal protection systems in a carbon dioxide-based
re-entry atmosphere. What was done just a couple of years ago at NASA
JSC is we developed the techniques and the capability to actually
test with carbon dioxide in the Arc Jet.
There were several issues associated with that, a lot of them safety
issues. There’s also a lot of nitrogen there. The Mars atmosphere
is a little bit nitrogen, but mostly CO2. We were running about 10
percent nitrogen and the rest CO2. For example, if there’s various
concentrations of carbon monoxide, if you get that, that can be an
explosive mixture. Obviously, we didn’t want that to happen;
that’s mostly in the heat exchanger, downstream of where the
Arc Jet core is. We had to monitor the species in the diffuser and
the heat exchanger, and make sure we were not getting close to those
The other issue was the formation of cyanide, since, again, you have
carbon dioxide and you’ve got nitrogen. You can form cyanide,
and we were concerned about that depositing on all the hardware in
the test chamber. We had to slowly work our way up. After every run,
you’d call in the safety folks, and they’d go in, in their
bunny suits if you will, and wipe things down and look for traces
of cyanide. In combination with that and with the gas analyzers that
we put in the chamber to look for the carbon monoxide, we slowly developed
that capability to test in carbon dioxide, which to me was one of
the first times we’d done that especially the power levels that
we were testing at. We were testing at 2-3 megawatts of power and
getting heating rates on 4-inch diameter models of a couple hundred
watts per centimeter squared. That was really a technological development
that we did recently that’s sort of gone unpublicized.
Why do you think that’s the case?
sure. I guess part of it’s because our plans for NASA going
to Mars are not really very detailed or set in concrete, and so I
guess maybe it was before its time. If we had been earnestly working
towards a mission to Mars, people may have been more aware of this
Any other technological developments that stand out?
not the ones that I can say that really stand out. There’s all
sorts of minor ones. The people at the JSC Arc Jet facility were very
good about modifying the arc heaters, et cetera, to get them to generate
environments that we were looking for, as test engineers. They really
looked for flexibility in the hardware and trying to get the capability
to change things quickly to produce the test environments we were
looking for. Again, people wouldn’t say it ranks up there real
high as a technological advance, but it was very important to us to
be able to have that flexibility in the test hardware.
That building was open in ’67, and you start working out there
around ’88. Have things changed over time? Do things look different
since you started working out there?
While I’ve been involved in that Arc Jet, they actually did
install a new test chamber. That was a big event. I think it was,
I forgot, exactly, right around 1990. Right around there is when they
installed a new test chamber. That gave us a lot of capability, to
have that bigger chamber, so that made things look a little bit different.
Then, the addition of the laser block house for the LIF system, the
Laser-Induced Fluorescence System, that was added on. Over time, yes,
things changed a little bit. The basic high bay there and, of course,
the one test chamber was the same one that was there when I arrived.
It’s an interesting building, when you walk in and think it’s
been there since ’67.
and of course, the control room’s changed quite a bit. There’s
a few things that are very similar to back in the late eighties, but
a lot of the control systems have been upgraded. You start going to
your LCD [Liquid Crystal Display] monitors, as opposed to the old
CRT [Cathode Ray Tube] monitors, et cetera. A lot of that had been
upgraded over the years, little by little. You notice that, especially
if you’re sitting in the control room during a test.
Did things change in terms of testing, or have they pretty much been
consistent since you’ve been there?
you’re operating a very high-power facility, there are safety
issues. There’s always this balance of taking risks but being
safe, and sometimes the safe part would take over so much, it would
constrain us. I think over the years, we’ve gotten a pretty
good balance of that. There’d be times when I think they would
go a little bit too far because the paperwork, et cetera would slow
us down so much, we couldn’t get tests done quickly, for example.
That’s good. You have to balance that. You want to get your
testing done, and you need to do it in a timely manner, but you also
have to protect people. Protecting the people, I think, was always
very important, was forefront in everybody’s mind.
Sometimes, they also were worried about the hardware, protecting the
hardware, but in these type of tests, hardware is going to get damaged.
You don’t want that to happen very often, but sometimes, you
have to risk that happening. Sometimes, they were a little bit too
worried about hurting the hardware, and people would get upset. They
felt like they had been chastised or whatever because a test article
was destroyed. It may or may not have been through any negligence
of themselves, it just happens. You don’t want to degrade people
because that happens. Test articles were sometimes expensive—$100,000
for a test article, sometimes—especially when we were doing
Space Shuttle Orbiter testing. Sometimes we’d have a failure
in one of the systems during testing, so we’d lose the test
Could you actually re-use a test article that was $100,000?
of them. The one I’m thinking about, no, we couldn’t have.
It was damaged too badly. Some of them were. That’s one good
thing about the Space Shuttle Orbiter tiles, of course, they were
built to be reusable, and so we could quite often get multiple tests
off one test article. Sometimes, we were testing large arrays of tiles,
about 2 foot by 2 foot, so there’s a lot of tiles there. There’s
a lot of instrumentation, and there’s a lot of labor that goes
into building that test article. Sometimes the test articles could
be pretty expensive, and you could damage them.
Wow, I had no idea. I was just thinking they were those little hockey
pucks that we saw.
some like that. We call them stagnation testing, but over in test
position one, we tested arrays of tiles. Some of those were on the
order of 1 foot by 1 foot; others were 2 foot by 2 foot. We did a
lot of testing with that, and actually I helped do a lot of that testing.
That was associated with the Return to Flight after the Columbia accident
[STS-107]. We were looking at damaged tiles and what could they tolerate,
because we all know that we were getting hit by material coming off
of the external tank, or, sometimes, the solid rocket boosters during
ascent. Because the tiles were delicate, and sometimes there was ice,
also, but between the foam or ablators or ice that would come off
during accent, they would impact the tiles and damage them. We were
very concerned about what was the tolerance of damage that the tiles
could handle. The only way to do that was actually to run lots of
Arc Jet tests on all sorts of different types of damages, and then
build a thermal model that could validate through these tests, so
we could evaluate each and every damage that happened. That was a
large part of what we did in Return to Flight, after the Columbia
accident in 2003.
What sort of damage did you simulate? How large were the chunks?
ones were on the order of an inch square. Some of them were probably
almost an entire tile, so that’s a 6 by 6 tile. We would look
at large damages. Also, what we’d look at is the repair techniques
for repairing those tiles, which we developed after the Columbia accident.
Those capabilities were carried on board the Space Shuttle Orbiter
in flight, so in the event that it was needed, we would actually ask
the astronauts to go out on an EVA [Extravehicular Activity] and repair
the tiles. After the accident, we set up what we call a Damage Assessment
Team, that was the DAT team as we call it for short. There were quite
a few people; it was mostly USA [United Space Alliance], Boeing, and
NASA folks that were involved in that. What we did was after launch
and the ascent, during the approach to the Space Station, lots of
photographs of the Orbiter were taken and downloaded to the imaging
folks here at JSC. They would go through with a team of TPS [Thermal
Protection System] folks and look at all those photos to look for
The other thing, of course, we did is put the Arc Jet on standby.
If we had to go test anything, we were ready to go. We had generic
test articles ready to go, and as a matter of fact, in some of the
flights, we had to execute the Arc Jet tests to evaluate damage or
repair. Most notably, we did that in STS-117, when we had a blanket
come up off the OMS [Orbital Maneuvering System] pod, and also during
STS-118, when we had ice damage to a tile on the belly of the Orbiter.
We knew we were in trouble when we first saw the photograph, because
we saw red. Red is the RTV-560, which is the adhesive that bonds the
tiles to the Orbiter. If you look in the damage and you’re seeing
red, that pretty much means the damage goes all the way down to the
bond layer. That’s bad.
We were very concerned about that damage, and so we took out generic
test articles. The crew on board the Space Shuttle actually went out
there with their scanner, a laser scanner was on the arm, and scanned
that damage, so we had good geometric measurements of the damage.
Downloaded that to us, and then we sent that information over to the
machine shop here, and they actually simulated that damage in several
tiles. We went and tested them in the JSC Arc Jet, and then, of course,
we used our analytical tools to assess those to determine whether
they were safe to re-enter without any repair, or if they had to make
a repair to that damage.
Do you think repair would have worked? It’s my understanding
that they never ended up using that capability.
we never did use it. We had enough testing under our belt that we
thought yes, it would work. There was a DTO [Detailed Test Objective],
where they actually had a dispenser that the astronauts would use,
and a little gun to inject—essentially, it’s like an RTV-like
material—into a hole, a cavity of a tile. We did get a DTO on
orbit so they actually had, in the back payload bay of the Orbiter,
a tile array with some real damage. We actually shot some particles
at a tile, and then just bonded it to a plate, so we actually had
some real damage. That was taken up on orbit, and then the astronauts,
during the EVA, took out the repair kit and repaired those tiles.
Then, they cured on orbit. They brought those tiles down, when they
came back from their mission, and we actually tested some of those
tiles in the Arc Jet. That was pretty much the final test. We said,
“Hey, yes, we really think we’re good to go as far as
having the capability to repair those tiles, at least for small damages,
up to the size of an entire tile.”
Were you involved at all in the investigation prior to the Return
that was 2003. I had left this area in 1996. I’d left Lockheed
in ’96, went to Goodrich Aerospace in California and worked
on the X-33 program. Unfortunately, that was canceled in the year
2000, and so I moved back here to Houston, so I returned to my old
job at Lockheed. I was working on the X-38 program, and, unfortunately,
that was not going well either. I remember specifically, just before
the Columbia accident, when things were not looking good. The X-38
looked like it was going to be canceled, and they were trying to cut
back people supporting Space Shuttle Orbiter. Then, the accident happened
on February 1, 2003. I remember that really well because that was
a Saturday, and we immediately got called in, a bunch of us. The first
thing that we were asked to do is to help them locate the debris from
the accident. Then I supported the aerothermodynamics team to help
figure out the failure investigation, about what exactly happened.
The big breakthrough that came for us was when they found the MADS
[Modular Auxiliary Data System] data recorder, almost intact, just
a dent in it. Don’t know how that happened, you just imagine
a data recorder showing up in East Texas in some grassy field. They
found that. They extracted the data from that, and that gave us a
lot of information of what happened. We got those data [points], and
we started piecing together what could have happened. One of the things
we found out was that the spar had been breached with hot gas coming
in, and probably severed some lines, some electrical cables. To test
that theory, we actually fired up the Arc Jet and built some hardware.
The spar on the Orbiter is aluminum, so they had pieces of aluminum
and we would expose them to the Arc Jet and see how long it took to
generate a hole through that. It had the model of the electrical cables
behind that, to see how long it took to melt through all those cables.
We very quickly generated all that hardware and got them tested in
the JSC Arc Jet to support that investigation, to support some of
our theories. That went on quite a while, and we had a lot of informal
meetings with the CAIB, the Columbia Accident Investigation Board
and would brief our results to them. It was a combination of Arc Jet
testing and also analysis. We did a lot of analysis, looked at the
ingestion of the hot gases through the wing leading edge, the various
holes, where it would go, and how it would impinge on the insulation
and in the spar: the whole story of the failure propagated, to the
point where we lost the Orbiter. It was really a serious time. It’s
hard to explain to people. I think all of us look back—we were
all very depressed about the accident, but we all worked very hard
and enjoyed our work and thought it was our responsibility to find
out what happened. It was shortly after the accident investigation
that then I was hired on as a civil servant at NASA JSC.
That’s good news.
was good news for me, yes. It was good news for me, but as I tell
my kids, seven people had to die for me to get a job at NASA. You
think about that. I don’t know. If that accident had not occurred,
would I have gotten a job at NASA? I don’t know. I just don’t
know. I couldn’t tell you yes or no, but I know that my participation
in the accident investigation was one of the reasons that I did get
If I remember correctly, on the first Return to Flight, Eileen’s
[Eileen M. Collins] flight, there was more damage to the tile.
wasn’t so much damage to the tile—there was a gap filler
that was hanging out on STS-114. I forgot how many tiles—there’s
like 20,000 tiles on the Orbiter, something like that—but between
quite a few of those tiles, there’s obviously gaps between the
tiles. In some areas on the vehicle, you don’t need to fill
those gaps, but there’s various reasons when you have to fill
the gaps. During installation, for example, there will be gaps that
are larger than are acceptable, so they have to fill them with something.
They used what they call gap fillers. There’s other regions
of the vehicle where there’s high pressure gradients, so all
the gaps are filled, because they know that those high pressure gradients
can result in the very hot gas getting down in the gaps between the
tiles and over-temping the structure. Over the years, we’d have
lots of gap fillers installed on the vehicles. Unfortunately, we also
had lost lots of gap fillers during flight. Sometimes, even on landing,
they’d find them on a runway.
We knew we were losing them, and I guess the processes for installing
them had not been revisited. We also knew that if they did stick out
or come out during re-entry, they could trip the flow field or the
boundary layer. When you trip the boundary layer, it transitions from
laminar to turbulent, and turbulent boundary layer is much hotter
than a laminar boundary layer, so everything downstream of that trip,
that protuberance sticking out, which is gap filler, can get very
hot. We knew this has been happening for years. My first involvement
with that was on, I think it was STS-50, when we saw what we called
early boundary layer transition. That may have been caused by a gap
filler. In that particular instance, the reason we knew it was that
one side of the Orbiter had transitioned before the other side. It
actually caused a moment on the vehicle, and the pilot had to go in
and correct for that moment and was surprised by it. Nothing crucial,
nothing critical, but people wanted to understand it.
We did start investigating that in the nineties, so we did start looking
at the boundary layer transition. It was just something we ended up
living with. When we got to STS-114, we saw this big gap filler hanging
out when we did our on-orbit inspections of the photography, and the
quick analysis that was done was saying, “Hey, we may be causing
some problems of heating to the wing leading edge.” Everybody
was very sensitive to that because we had just lost a vehicle due
to an impact actually to the wing leading edge. Not that we had an
impact, but we certainly didn’t want to overheat it and over-temp
the hardware. There was a lot of consternation on what to do, so they
did send out an astronaut to go pull that gap filler out. Of course,
you can imagine, that was the first time ever that an astronaut was
sent on an EVA out on the belly of the vehicle, and you’re just
in a sea of tiles, so it can be sort of disorienting. The gap filler
was actually very easy to pull out, and he did. [Stephen K.] Robinson
was the astronaut who actually did that. He pulled that out and brought
it back, a matter of fact. That was the big issue.
There was actually a blanket that was damaged up towards the crew
cabin. They were also concerned about it coming off. It was obviously
by the window, by the crew cabin. The concern there was that that
blanket would detach itself during re-entry, and then go back and
hit, for example, the vertical tail or the split rudders there. We
actually did do some wind tunnel tests with damaged blankets to see
if we were going to lose them. Of course, we also did some impact
tests to say, “Okay, what if it does come off? What’s
going to happen to the rudder?” We convinced ourselves, after
doing all that testing, that we could just leave it alone and we’d
be okay, and we were, even though it got sort of torn up. …
Wow, that’s pretty cool.
there were two things that happened on STS-114.
You guys were running those tests, you mentioned?
In that particular test, we really didn’t need the Arc Jet,
even though it was still there. If we needed it, we were going to
use it. After that, with the gap fillers, there was a big push to
go in, first of all, to work on a better method for installing the
gap fillers, and then to pull the old gap fillers and install them
with this new technique so that they wouldn’t come out. There
was regions of the vehicle that we started identifying, of which ones
were critical to get done. Then, of course, we had more than one vehicle
to do that to, so you had to work that within the flow between flights.
We were going in there, and the KSC [Kennedy Space Center, Florida]
folks were going through and pulling the old gap fillers and putting
in new ones, using a new technique. With the new technique, I don’t
think we ever experienced one coming off, not that I remember.
With the Arc Jet, you were testing which gap fillers were most significant?
We did a little bit of testing in the Arc Jet, but not a lot for that.
We still had specifications where we knew we needed to have gap fillers
in certain regions, and so we weren’t going to change those
specifications. It was more developing the technique for installing
the gap fillers, such that there were going to stay there. The problem
was, you can imagine, it’s a thin piece of ceramic material,
and you’re putting some adhesive on the thin edge and then trying
to slide it down a gap, so you’re not getting a really good
bond. A lot of times, what would happen is the RTV-560 adhesive would
just slide off as you were sliding it down the gap. So by the time
you got it down to the surface, to the structure, there was no adhesive
left. You were getting more of the RTV on the sidewall of the tile,
and of course, during re-entry, that would heat up. The adhesive capability
would just go away. They had to develop a different technique so that
they could get a good bond down towards the bond layer, down towards
the structure. They did, but it’s a very painstaking process,
to go in there and start pulling the old gap fillers and putting in
new ones. They did do that, after 114.
How did operations change out at the Arc Jet as a result of the Columbia
thing was that first of all, we got what we called generic test articles
built, and we had a cage where we kept all of those, and a lot of
TPS supplies we kept. We had a lot of hardware that was actually dedicated
to supporting flight. The other thing, of course, right before flight,
we would make sure that everything was operating fine in the Arc Jet.
We wouldn’t suspend testing; we would just not do any risky
testing because we didn’t want to damage the facility. Then,
of course, put people on standby to say, “Hey, you could get
called in,” and that was engineers and technicians. Of course
we did several times, and usually people were put on around the clock,
12-hour shifts, so we had crews supporting the Arc Jet. If there was
an inkling that something was going wrong, then we’d start manning
the Arc Jet facility, just getting preparations.
There’s a few times, I don’t remember the specifics, where
we would actually get some test articles ready, put them in, install
them, and get ready to do calibrations. That’s always the first
thing you do in an Arc Jet test is do a calibration to get the test
conditions you’re looking for, before you actually expose the
real test article. We would do that. We would get the technicians,
et cetera, ready to go; we’d get the test articles ready to
go. Sometimes install the calibration models in the facility, and
maybe do a couple of runs to support a flight, just to make sure that
we were ready to go if needed.
This continued all the way to end, through [STS]-135?
How did you report—you did a test and here is your findings—did
you go to the Mission Management Team [MMT]?
I mentioned earlier, the Damage Assessment Team, the DAT team; all
the Arc Jet testing was coordinated through the DAT team. The results
of the Arc Jet test, then, would be included in the Damage Assessment
Team report and reported out to the MMT.
Did you ever have to go to any meetings?
had to go quite often. The way we’re operating there, Boeing
was actually the lead from engineering for the thermal protection
system. The subsystem manager at Boeing was essentially the head of
the DAT team. He was the person that usually did the briefings to
the MMT. So it was usually a Boeing person or a USA person, and usually,
we were there as backup, the NASA folks. There was a couple of times
that myself and others had to go in there and just support them.
Any memorable discussions that you recall?
of them you probably don’t want to admit.
You don’t want to put those on the recording?
could do some sociological study of it. Groups of people can work
very effectively together, but sometimes they joke around. People
on the outside, if they heard that, they’d think that’s
not appropriate. It’s just the way people work together, and
it didn’t degrade or deter from their capabilities. We would
try to get people to say certain phrases, even during an MMT, just
joking around. It’s not a matter of disrespect, it was just
the way people handle crises sometimes. We did that also; we would
go in there and joke around, and just the way some people handle stress
We had a good team. We had lots of food in our DAT room, [which was]
very close to Mission Evaluation, the MER. We were always notorious
for having lots of food there, and the big bosses, [N.] Wayne Hale
and Bill [William H.] Gerstenmaier would come by and eat some of our
food. We’d talk to them, and they probably also wanted to come
and find out what was going on. That was really, I think, good and
exciting times. It was a good team of folks; they all did their job
very well, even though we joked around.
You had mentioned you worked on the X-38. Were you doing testing out
at the Arc Jet for that?
I was not doing Arc Jet testing for the X-38. I was actually doing
hypersonic wind tunnel testing for the X-38. The one job I had on
X-38 was to develop the aero-heating model for the body flap that
was on the X-38. I actually helped design and conduct a test up at
CUBRC [Calspan-University of Buffalo Research Center, New York], up
at University of Buffalo, and we actually did those tests. It was
a pretty good-sized model of the X-38 with the body flap on it. I
forgot what scale it was. It’s very difficult to do those tests,
so I helped design and conduct that test. I actually spent time up
there doing those tests. We were testing at very high Mach numbers
in this tunnel, the shock tunnel that they have. The model was almost
like 20 inches long, so it’s a good-sized model. We also had
some tests done at NASA Langley [Research Center, Hampton, Virginia],
in their Mach 6 and Mach 10 facility, and there was another set of
test data that was obtained in Europe. I don’t remember the
facility. If you remember at the time, X-38, we were doing a cooperative
effort with ESA [European Space Agency] during that time period. My
job was to take all those data [points] and come up with an aero-heating
model to predict the heating on the body flap during re-entry.
That was what I did do, and that was actually a lot of fun. Unfortunately,
I felt like I really did make—I call it a big contribution to
the technology at that time point—but they wouldn’t allow
us to publish it. It’s one of those things where you just get
satisfaction from knowing what you did, and not getting the, I’ll
call it the public awareness, of the importance of what you did. Nevertheless,
it was a lot of fun. Unfortunately, X-38 was shutting down right before
the accident, and I was still documenting my model and my results
from all the testing. There had been some Arc Jet testing for X-38
but I was not involved. As a matter of fact, for the body flaps, specifically,
the Europeans had developed that and they’d done some Arc Jet
testing and actually did a very good job. It was a carbon SiC [Silicon
Carbide] material that they were using for the body flap. A lot of
that development for the body flap itself was done by the Europeans.
That brings up another question. Over the years, of course, there
are two, I guess, other Arc Jet facilities that NASA has, and other
facilities within the U.S. and around the world. Would you talk about
working with some of those other facilities?
you go back to the Apollo days, boy, there was a lot of Arc Jets.
A lot of companies had their own Arc Jet. After Apollo, it pretty
much got down to three at NASA. There was one at NASA Ames, one at
NASA Langley, and one at NASA JSC. Then, also, of course, there was
the Air Force facility at AEDC, the Arnold Engineering and Development
Center [Arnold Air Force Base, Tennessee]. The AEDC facility was more
set up for ballistic re-entry vehicles, and so, very high pressure,
very low enthalpy, lower energy, but very high pressure, which was
not appropriate for a Space Shuttle Orbiter mission. All the testing
and developing of the Space Shuttle Orbiter TPS was done at the three
NASA facilities: at JSC, at Ames, and at Langley. After the development
phase, the Langley facility was shut down. Some of that hardware was
actually shipped here, to NASA JSC.
However, they did set up a small little Arc Jet we call HYMETS [Hypersonic
Materials Evaluation Test System]. It’s a very low-power facility,
on the order of 400 kilowatts, and it’s essentially for material
screening. You can test articles that are about 1 inch in diameter,
and that’s about it. After that, then, it was essentially the
NASA Ames facility and NASA JSC. Those were the two main facilities
for the rest of the Space Shuttle Orbiter program, the beginning of
the Orion program, and, of course, supporting various other technology
initiatives and commercial entities that wanted to test in there,
too, to support their development. That’s sort of the background
to all of this. The Ames facility, I personally have worked with,
when I went to go work X-33 for example. I was pretty much the architect
of all of the thermal testing for the TPS for X-33.
At the time when I started working it, most of the testing was scheduled
to be done at NASA Ames, but based upon capabilities and scheduling,
et cetera, then some of that testing was shifted over to NASA Johnson
Space Center. We pretty much kept both facilities busy just doing
X-33 testing. We did a lot of testing at NASA Ames. In this case,
it was metallic TPS we did a lot of testing with. We were looking
at doing a lot of large arrays, literally 30 inch by 30 inch test
articles. We did a lot of that. We did some at JSC on those and then
quite a bit at Ames.
How is JSC’s facility different than the one at Ames?
all sorts of different things. For example, at Ames, it’s a
little bit older facility; they have more power, more air capacity
or vacuum capacity, so they can blow more gas at higher power than
we could at JSC. They have two facilities. One is the workhouse, which
is what they called the IHF [Interactive Heating Facility]. It’s
rated about 60 megawatts. Then, they have another one, which is called
the Panel Test Facility, which is smaller test articles. It’s
just for panels on the order of 20 inches by 20 inches you can go
test. Then, they have the AHF, which the Aerodynamic Heating Facility.
It’s about 10-20 megawatt facility. Each of those are essentially
test legs off of a bigger system of boilers and power grid. They have
a footprint that’s much larger, if you will. At NASA JSC, there
was usually only just two test positions, and we were rated at about
10 megawatts, so a little bit lower power. Both of those test positions
fed off of the same steam injector system, the same power supply,
et cetera, so we just switched between the two.
The infrastructure was a little bit different. Obviously, much less
infrastructure at JSC. Over the years, the testing folks had developed
different arc heater systems. Ames tended to go one way with their
design for arc heaters, and JSC tended to go a different way with
their arc heaters. There’s benefits, pluses and minuses, to
both. It was just different, so that led to different test capabilities.
Part of the problem with the Ames facility was a lot of their hardware
was very hard-wired, if you will. Plumbing connections, et cetera,
were just pipes bolted together, right? If you wanted to make a change,
you got to un-bolt all those pipes and bolt them back up, which takes
a lot of time. You’re always worried about leak checks, et cetera,
where at JSC, for example, they had flexible tubing for their water-cooling.
You could move the heater, for example, without having to un-bolt
a lot of pipes, so we were more flexible in making nozzle changes,
Also, I mentioned the arc heaters are different. At Ames, it’s
a little bit more amenable to higher power, which is good if you want
to test at those higher power conditions. To protect the electrodes,
for example, what they had to do is they had to feed in argon very
close to the surface of the electrodes, and that contaminates the
gas—potentially, it can—which you’re testing in.
At NASA JSC, we mixed nitrogen and oxygen; in our electrodes we’d
dump in nitrogen and then mix it, so we came up with the right concentration
of nitrogen and oxygen. We didn’t have any argon. We used argon
to start the arc but then quickly switched over to a combination of
nitrogen and oxygen. There was capability there that we had that was
The other problem that the Ames facilities had was since they were
trying to actually split the electrical arc between different electrodes,
they had to actually control the resistance, the resistors. To get
certain test conditions, you had to pre-set those resistors ahead
of time to get the electrical arc to split and attach to different
electrodes. That was not done at JSC. It was essentially a cathode
and an anode, and yes, they were wear parts, and you had to replace
them after so many runs, but we didn’t have to pre-set any resistors.
It was much easier to make a run, and then actually vary the power.
During a run, we could go from pretty much bottom of our range to
the top of our range in power within seconds, so you could actually
do a simulation of a flight heating profile in the JSC facility, but
just couldn’t do that at Ames. You had to set up and say, “Okay,
I’m going to go test these conditions, and that’s it.
If I want to test these other conditions, then I’ve got to shut
down and reset everything, and then run again.
Again there was pluses and minuses to both approaches, but I think
the way that happens—which is always good—is that you
get different groups working to solve the same problem, right? We
had this problem. You’re trying to simulate re-entry, and some
people are going to approach it one way, and another group of people
are going to approach it a different way, so they come up with different
solutions. Sometimes, they result in similar responses or similar
results, or complementary. I think it ended up being pretty complementary.
There was things that Ames could do that we could not, especially
in the high power region. We couldn’t hit real high heat fluxes
or real high pressures. On the other hand, we could hit the real low
heat fluxes and low pressures very easily. Also, of course, we had
more flexibility in our testing.
There was some overlap of test conditions, which is good because then
you get a comparison between the results in two different facilities,
and you have more validity. If you get the similar results in two
different facilities, you’re more convinced you’ve got
the right answer. Then, of course, there were capabilities that were
different between the two facilities, and so that’s good, too.
A lot of it had to do with that approach. Ames folks tend to go and
design an arc heater a certain way, and the JSC folks went a different
In your opinion, was there ever any competition between the three
there’d be a lot of competition. I think at the working level,
it was not so much of an issue. It was more at the higher levels.
Never really understood that. Of course, when there’s not a
lot of resources, everybody’s competing for resources, so that
can force people that may be even our friends to be enemies. Yes,
there was a lot of conflict. In my tenure here, there’s been
a lot of conflict between JSC and NASA Ames over this, but again,
not so much at the working level, it’s been at the higher levels.
Sometimes it’s gotten really nasty, which is sad. We shouldn’t
be doing that. We had a program here that was working Constellation,
and it was called ADP, the Advanced Development Program. We were looking
at doing pre-development, if you will, of the heat shield for the
We did have a fair amount of money, so we actually developed a lot
of capabilities in Arc Jet testing at JSC and at Ames, and we worked
very well together on developing those. It was very complementary
and helpful for both sides. Once that was over and the resources started
tightening up again, then the competition comes back and people are
very wary of the other party, let’s put it that way.
You mentioned Orion and I did want to talk about that because you’re
the subsystem heat shield manager, correct?
Talk about some of that testing that was done. How did you determine
which material you were going to use, and were there challenges in
making that selection?
there were. First of all, [in] the very early days, what we did is
we put out a request for proposal [RFP], just surveying industry.
“Hey, what would you propose to use?” I can’t remember
how many materials we got, but five or six, so we got a pretty good
response. We finished that phase and did some early on testing, and
of course, you can imagine, there was problems and the materials were
not behaving well. Some were behaving very well. Then the next phase
was to go take that a step further and say, “Okay, now, we’re
going to put out an RFP.” Now you’ve got not just a material,
it’s a system, so you’ve got to show that you can actually
build this to some scale that’s applicable to the Orion capsule,
which is big. It’s 16 feet in diameter.
We put that out, and we got one response. It’s like, “Uh-oh.”
That was not good, right? You don’t want to have a competition
with one group and that’s it. Boeing had responded, but they
were actually using PICA [Phenolic Impregnated Carbon Ablator]. After
the fact, we realized the schedules were such that we were giving
penalties in the contract for not meeting schedule. Some of the companies
just were unwilling to risk that. Other companies just didn’t
have the capability, especially the smaller companies, to build something
that big. I’ll say it’s good or bad—we had a schedule
slip in Constellation, and because of the slip, that allowed us to
submit another RFP, and then we got a couple more responses. We ended
up with three companies that we were working with for the development
of the heat shield. This was actually during the Advanced Development
Boeing was the one that was the early one with the PICA system, and
PICA stands for the Phenolic Impregnated Carbon Ablator. Another Boeing
proposal, but it was with their own material, which was the BPA, the
Boeing Phenolic Ablator. Then, of course, the third was Textron, with
the heritage Avcoat material that was used on Apollo. We worked all
three of those, and each one of them separately. Obviously, we had
a head start on the PICA system because we’d gotten the contract
to them earlier, and the other two systems were lagging. The Boeing
Phenolic Ablator, we did some testing of that and really were liking
its performance. It was really good. Unfortunately, when they started
scaling up their production and looking at building bigger test articles,
they were experiencing a lot of cracking. They couldn’t resolve
it in the timeframe that we had, and so that contract was terminated.
At that point, it pretty much left Textron with Avcoat and Boeing
with the PICA system.
The Avcoat had some problems at first, but then it looked like we
got those resolved, so then we actually had both Boeing and Textron
build what we called manufacturing demonstration units, where we had
full-scale, stainless steel structure, in which I think it was about
quarter of a pie of their material to that, showing that they could
actually build that hardware. It was on top of all the Arc Jet testing
and mechanical testing we were doing of the material itself. After
that, we were getting data from both systems. A big problem with the
PICA system, of course, is it’s what we call a tiled system.
You can only build blocks of it or tiles of that PICA that are so
big, and then you bond them down to a structure, but what do you do
with the gaps in between? As a matter of fact, I helped lead up a
trade study on different what we call gap fillers for that system
and really never could come up with a candidate that solved all of
We looked at all sorts of things. As a matter of fact, I think one
thing somebody submitted a patent for, we called it PICA on edge,
where they actually pre-crushed the PICA so it was sort of spongy
and bonded that into the gaps between the PICA tiles. That had structural
issues, it was cracking. Performed very well thermally, but we were
experiencing cracking during the testing, as a matter of fact, both
mechanical testing and Arc Jet testing. We were concerned with that.
When you bond down these blocks, it’s very hard to verify that
you’ve got a good bond.
Going back to our history on Space Shuttle Orbiter, most of the tiles,
we did what we call a bond verification run. We actually take and
bond the tiles on, and then you actually use a vacuum chuck and pull
that tile to make sure it’s well bonded to the structure. The
PICA blocks were so big, you couldn’t do that, plus the PICA
material is very porous. Yes, we could paint it, and we did. We put
a paint on it and tried to do that to some extent, but it was not
working very well. Then, of course, the material itself, PICA, is
relatively weak. It’s not much stronger than the higher density
tiles, so we were very worried about it structurally failing.
One failure mode is for an in-plane crack. In-plane cracks are always,
I’ll call it the nightmare of a TPS person, because then you’re
losing a whole chunk of material, almost all the way down to the bond
line. If you have a through the thickness crack, we’re less
concerned about that because most of your material’s still there.
You can get a little bit of hot gas in the crack, but you’re
not going to overheat your structure very readily. That was the weak
link, if you will, of PICA, was those two things: its strength, especially
its in-plane strength, and then of course, also the gap fillers to
go in between there.
When we went to the down-select, we actually had an official down-select,
I think it was 2009. We looked at Avcoat and PICA, and at that time
Avcoat was selected. It is a pretty laborious system. The good thing
is that you really get a good bond to the structure. It’s a
honeycomb material that’s actually bonded down, and we are able
to do pull tests to make sure that honeycomb is well bonded to the
structure before you fill the honeycomb with the ablator material.
Since then, we have built a heat shield with Avcoat. It’s very
laborious. You’ve got to bond in the honeycomb, you’ve
got to fill all the cells, and it takes a lot of time and a lot of
labor to do that. The other thing is that we know that we have a risk
of cracking, but it’s through the thickness cracks, but nevertheless,
it’s still cracking. In our experience with our EFT [Exploration
Flight Test]-1, that can happen during processing. We didn’t
understand that. We knew Apollo had had similar problems, but we didn’t
know the details; it’s not well documented. Even today, we are
predicting the potential for cracks, through thickness cracks, for
Avcoat. Every TPS material, there’s issues with, so it’s
a real challenge to deal with them. There’s no easy solution.
Would you talk some about the Arc Jet testing of these materials?
the initial screening, you’re mostly doing the small pucks that
are about 4 inches in diameter. That’s just to get some idea
of the material response. We look for two things during that: we look
for recession, and we also look for the in-depth thermal response.
We actually put what we call thermocouple plugs in the material so
that we have thermocouples that are at certain depths inside the material.
We take those data [points] to validate our ablation models. Obviously,
we want to do that over a large range of conditions, so we build a
lot of those and expose them to the Arc Jet in different heat fluxes,
in different pressures, and different enthalpies. You’re trying
to define your flight space saying, “Hey, here’s my flight
envelope, and now choose a bunch of points over that whole envelope
that I can test on to validate my model and make sure it’s working
We did that for both Avcoat and PICA. When you’re going through
and saying, “Okay, it’s not only got to perform thermally,
but it’s got to perform structurally, too,” there was
various things we had to go do. For example, we ended up having to
change the honeycomb vendor. The one honeycomb vendor we had was not
performing very well; it was costing us an arm and a leg, so we decided
to change to a different honeycomb. Of course, then you’ve got
to go back and retest because there’s different material. It’s
very similar, but nevertheless. The other thing is that when you put
the ablator material inside the honeycomb, you have to prime it, so
we changed the priming material. We had to go back and re-test that.
That’s all been done, now, but you can see that that caused
a lot of additional testing over this flight envelope that you’re
looking at, just to get the thermal performance down.
On the PICA side, the PICA material itself is actually easier to model,
but of course, we still had to do that. It’s essentially a phenolic
and a carbon; it’s a little bit easier to generate the ablation
model. The Avcoat has glass in it, actually glass fibers for strengthening
up the char, and, of course, that glass complicates the ablation performance
and the modeling of the material itself. On the PICA side, we still
had to test the material over a large range of simulated flight conditions,
but then also we had to deal with the gap fillers. That was our big
challenge, so we started off with 4-inch diameter pucks, and pretty
much split them down the middle and put different types of materials
in between, to evaluate the recession of those materials, and then
also the thermal performance. That’s where we started having
lots of problems, especially the PICA system, since it’s an
array of tiles. Then we actually had to do other testing, too, so
we had to test in wedges, et cetera, and we also tested the Avcoat
in wedges. That gives you sort of a shear flow on the material, as
opposed to just a stagnation. The flow is impinging directly on the
So, we did a lot of testing, probably more with the PICA than with
the Avcoat, in wedges. We tested them out at AEDC, at JSC, at Ames.
AEDC, we went there because we knew that for some of the re-entries
we had pretty high pressures and shear on the material—aerodynamic
shear. They could hit those conditions at AEDC better than we could
at our NASA facilities. We did a lot of the testing at Ames, also,
on that. We’re still doing testing on Avcoat. Since Constellation
was canceled, and we moved to the MPCV program, with the Multi-Purpose
Crew Vehicle, we’re concentrating on the EFT [Exploration Flight
Test]-1 flight, which is the flight that hopefully is going to get
off by the end of this year. We’re building that flight hardware
right now. We concentrate all of our testing on that flight envelope.
Now we also have to move on to the next mission, which is a lunar-centric
mission, so we’re talking about going around the Moon and coming
back from the Moon, or close to the Moon. Different set of flight
conditions—re-entry environments are different, so now we’ll
have to expand our flight envelope and test additional models at that
those conditions to further validate our ablation model.
We don’t always get good results from our Arc Jet testing. It’s
always a learning experience. We’ve tested some materials at
Ames, Avcoat, in particular, and PICA, and gotten different results,
when we thought they should be the same. We don’t understand
that, sometimes why we get different results. We also know that we
don’t characterize the flow field real well in either facility.
It’s always a challenge. We know that enthalpy is a very important
parameter, and the enthalpy is essentially the energy in the gas.
Knowing that is very important and has a very large impact on your
recession rates and predicting your recession. Probably heat flux
and enthalpy are the two biggest drivers in how much the material
recesses as a function of time. We would get different results from
different facilities. We didn’t know if that’s because
various test conditions were different, or just because our knowledge
of the test conditions were wrong. That’s why we’ve continued
to encourage folks, “Hey, you need to spend money and time to
understand the flow fields,” and that includes the laser-induced
fluorescence, and other systems, other probe systems, to back out
The enthalpy in the flow field is not a directly measurable quantity,
so you have to go through alternate ways to back that out. Heat flux,
on the other hand, is. You can get a heat flux gauge, expose it to
the flow field, and get those measurements and pressure also. We did
a lot of testing for those, for PICA and Avcoat, and we’re going
to continue to have to do it for Avcoat for the EM [Exploration] Missions.
That’s interesting that you said you get these different results,
after more than 50 years of using these facilities.
don’t have a very good understanding of even the flow fields
that we’re exposing it to, and especially the interaction of
those flow fields with the materials. In a lot of ways, it was neglected
after Apollo. There was not many people working that. NASA had moved
on to the Space Shuttle Orbiter. We had reusable thermal protection
system materials, so the ablators, which are necessary for the real
high-energy re-entries, that development just stopped. NASA didn’t
need it. They still used it for some planetary probes, but that was
very empirical, probably because people were willing to take risks
because there was no crew on board, or they were not mass-critical.
They’d go in there and say, “Okay, well, based on our
experience, here’s how much material you need. Why don’t
we just double it and go fly?”
The basic understanding of the material behavior, when it’s
exposed to the re-entry environments, was not progressing much at
all. If you go back today and look at what a lot of those people did
in Apollo, you’re just like, “Boy, they were really smart.”
It’s like we’re trying to play catch-up. We’re trying
to do what they already did in the late sixties, and in some ways
we’re still chasing them. We’re not quite there. That’s
in testing and in analysis. I have the utmost respect for the folks
that worked Apollo.
Have you brought in some of those graybeards?
some of them we have, and we’ve worked very closely with them.
If you go back then, you’ve got to put yourself in perspective,
remember that people were calculating things with slide rules, not
with calculators or high-powered computers. They were making plots
by hand, not Excel or not some other software on your PC. Nobody had
PCs [Personal Computers]. The end result of that is when you pull
in people from Apollo, it’s based upon their memory, and the
documentation is just not there. Even the plots, sometimes we’ll
find plots that are hard to read. That’s difficult. It’s
not like today, where it’s so much easier to document things
and save it. Unfortunately, a lot of the technical documentation was
destroyed after Apollo, for one reason or another. Going on memories
is not, sometimes, very good, but we’ve done that.
I wanted to shift gears a little bit and ask about day to day operations
out at Building 222. Can you share some recollections of how things
my role is more of what I would call a customer, so I was not involved
in the day-to-day operations of the facility. There is, for example,
a facility manager there, that’s a NASA person. There are also
what they call the test directors. The test directors are NASA people,
and the rest of the crew are contractors. The contractors include
what they call test conductors, and that’s the orchestra director.
He’s the person that’s actually turning the knobs, et
cetera, trying to control the facility. Then, he has his support folks.
They have a power operator, who actually controls the power, essentially
it’s the current, and then you’ve got quality folks that
are just monitoring, making sure everybody’s doing their job
correctly on the contractor side. Of course, you’ve got a boiler
operator that’s out controlling the boiler, that generates the
steam to generate a vacuum in the test chamber. You’ve got a
data operator, which is obtaining all the data that’s generated
during the test. You have technicians that are maintaining all the
facility hardware, prepping test articles, doing measurements on test
articles, et cetera.
The test director, which is a NASA person, would be monitoring what
the contractors are doing, would have to sign off on all the paperwork
that would go through, and obviously, the contractors who are generating
the paperwork, and that was work orders to provide instructions to
the technicians on what to do. Whether it be prepping a test article,
or repairing or maintaining some part of the facility, that would
go on on a daily basis, as they would assess the facility, what needed
to be done to keep the facility running safely and producing the environments
we want. There was maintenance or parts that were broken, and you’re
a very high-power system, so you know you’re going to have parts
break. You have anodes and cathodes that degrade over time, they have
to be replaced; somebody has to check them. Usually, it’s a
visual check. You have to look for potential water leaks all the time.
They’re doing that on a daily basis, and also they are preparing
test articles for test. They go in there and log in a test article,
give it a number, photograph it, do measurements of it, weigh it,
and then store it, using bonded storage. Then, of course, enter all
that into a data system, so it’s all stored. Some test articles
require preparation before a test,; some of them have to be integrated
into test fixtures, so the technicians would have to do that. For
all those steps, work orders would be generated, instructions. The
NASA person in charge down there would have to sign off on those and
say, “Yes, that’s the right thing to go do.” Once
a test article is prepped for testing, then, of course, they would
have to install them in the test chamber, get them oriented correctly,
positioned correctly, make all those type of measurements. It was
sort of tedious. After that, you’re ready to go test, but then
their job’s still not over because post-test, you have to pull
the test article out of the test chamber. Pull it off of the fixture,
photograph it, weigh it, quite often do geometric measurements of
it, especially if we’re talking about recession, to see how
much it recessed. The data operator’s pulling off all the data
and organizing it.
Data can be facility data—for example, mass flow rates, water
temperatures, current, pressures at different places in the arc heater—and
then thermocouples on or pressure measurements on the test article
itself. Also, they have optical pyrometers. Those are focused on the
test article, and they use those to get surface temperature measurements.
Got very high resolution cameras that they have to deal with. It takes
a lot of set-up work, and of course, once you get the videos back
from those, then you’ve got to process the videos. Some of those,
we have what we call a photogrammetry system, where essentially, it’s
this pair of cameras, very high resolution cameras, that are synchronized
so that between those two cameras, you can actually calculate the
recession as a function of time during the run. It took a lot of post-processing
to be able to do that.
There’s all that work that goes on, on a daily basis. Once you
get the test articles, then you have to store them after testing,
after you’ve done all your measurements on them. Then, you have
to track them for us, so there’s a lot that goes into the daily
operation in the Arc Jet facility.
As a customer, what sort of things would you have to do prior to and
after an Arc Jet test?
big thing, of course, that we were responsible for is generating what
they call a customer-based test plan. That’s pretty much a description
of what test conditions you want the test articles exposed to, any
special handling that you want [of] those test articles. For example,
some of them may have to be kept in bags with desiccant. If they pulled
them out of that bag for attaching thermocouple connectors or taking
photographs, they’d have to actually log how long it was out
of the bag because we’re worried about moisture absorption into
the test articles. Things like that, you have to generate in the plan.
We would write that plan and provide that to the facility. They would
take it and then write what they call their detailed test procedures.
They would also take that and send it over to the safety folks. They
could provide an integrated hazard analysis. Just to be aware, for
example, most of these materials off-gas quite a bit, so we usually
put in there that after a test, you would keep the test chamber evacuated
for 30 minutes, just to take all that off-gas and get rid of it before
anybody opened the doors. Sometimes, we’d repress and then re-evacuate
a chamber, just to make sure we’re getting all that stuff out
of there. All of that information, the customer has to provide. “Hey,
here’s the type of material we’re testing, here’s
how we want to test it, how to handle it.” That all goes into
the customer-based test plan. Then, we provide the test articles.
It was often a negotiation, too, with the facility folks saying, “Hey,
this is what we’d like to do, so what kind of test fixtures
do you have?” We would work back and forth because the facility
folks would actually take our test articles and integrate it into
the test fixture, and so we had to know what they were doing and they
had to know what we wanted. We would work together on developing that
After we got the test plan, after we got the test articles manufactured
and ready to go, then, as a customer, mostly what you’re doing
is just making sure that the facility is doing what you wanted them
to do. There’s times when you realize, “Oh, I forgot to
tell them to do such-and-such,” right, so you’re having
to correct them. Or things change. Every test, it seems like something
changes. Some of it’s because you test and you find something
else out, so you want to change your test, and then you have to do
a deviation. Even during a test program, you may have a dozen test
articles, and you’d say, “Hey, okay, we got these first
four done, but now I need to change how we’re going to test
the remaining.” You have to generate a deviation to the original
test plan and get that approved before they can continue testing.
Post-test, usually, you want to go in there and take a look at the
test articles themselves. At that point, you want your data. Of course,
the facility folks are usually busy trying to get another test program,
too, the next one in line. It’s always sort of prodding them
to get the data that you are looking for, and they’re busy trying
to get their next test moving, too. It was always a challenge for
them to satisfy everybody. That’s a lot of work.
There’s also test article design. There was a lot of interfacing
with the facility folks. To this day, I can’t say we know always
how to design test articles. We get anomalies that we don’t
understand, and sometimes we think it is because of the way we are
testing. For example, most of these materials are porous, and so,
obviously, you put high pressure on them, there’s going to be
some type of flow through the material. That can affect our results
because since we’re not testing the exact same geometry as flight,
then that can affect our results, and not be realistic compared to
what we’d expect in flight.
We’ve tried to account for that sometimes and gotten results
we certainly, to this day, don’t understand. We keep on revisiting
our test article design. For example, in wedges, quite often we would
see what I would call local gouging in the test articles. We convinced
ourselves that that was probably because you’ve got a water-cooled
wedge, so it’s cold. The flow field, it’s going across
this water-cooled surface, all of a sudden, it sees a hot surface,
which is your test article. There’s a lot of chemistry that
goes on right at that point and that’s not realistic because
the flight vehicle doesn’t have a water-cooled surface upstream
of your material, right? We would get strange results, and so we’d
have to put in graphite transition pieces, for example. All of that,
we have to negotiate with the test facility because they’re
going to integrate the test article into the test fixture, so that
takes a lot of back and forth between the facility folks, the technicians
who are going to do the work, and the customer.
I noticed that you’ve authored a number of articles about tests
or work that you’ve done in the Arc Jet facility. Have there
been any breakthroughs that you’ve found as a result of Arc
Jet testing, that you’ve let the military or aerospace industry
don’t know if I’d call them breakthroughs. It seems like
it’s all incremental. We continually learn, but I guess none
of them I would classify as a big breakthrough. It’s sort of
a continual learning basis, and unfortunately, since there’s
not a lot of continuity, we’ve tried to get that going, but
nobody seems to want to fund that. You come up, you learn the stuff,
and you publish it, and then that need’s gone away for at least
the time being. Five, ten years later, you go, “Oh, yes, I think
that back then, I did such-and-such,” and you’re trying
to go back to it and figure out where you were, and then continue
on. Those of us working at NASA, unfortunately, that’s sort
of how we have to operate. It’s a little bit different than
academia, who a lot of times, they will choose a technical problem
and then pursue it for years, maybe decades. It’s the same problem—maybe
different grad students working it, but the professors slowly developing
that capability over time.
Actually, in thermal protection system and in aerothermodynamics,
it’s sort of that way. As I mentioned earlier, we didn’t
deal with ablators much for probably several decades at NASA. We sort
of lost our capability, and so now in the last five, six years, we’ve
been regenerating that capability. I feel the same thing for a lot
of stuff that I’ve worked on, it’s been that way. It’s
been jumpstarts. We’d work on it for a few years and then drop
it and go work something else. Luckily, we’re a little bit better
at documentation now, so it’s easier to come back up to speed
than maybe it was in the Apollo days.
Were you there at the last test out at the Arc Jet facility?
was there right beforehand. I had to leave just before they actually
started doing the testing because I had some conflicts and some meetings
I had to be at. We call him the guru out at the Arc Jet, Jim [James]
Milhoan, he’s been a mentor for a lot of us. He was actually
a NASA person who helped develop the Arc Jet capability at JSC. I
forgot what year it was, but he retired. He’s come back as a
contractor and been supporting us for years. When he was there that
last day, they asked him to talk, and he just said, “I can’t
because I’ll just cry.” I could see that, too. There was
sort of a desire not to be there, in a sense. It was just something
you didn’t want to cope with because it was sad, which is our
human nature, I guess.
What do you think the Arc Jet here at JSC has meant to you and to
NASA, over the years?
me, I sincerely believe that it was an Agency asset. That probably
is not well recognized. I think over time, people will understand
that, but it’s going to take for it to be gone, for people understand
that. You have to be very embedded in this area of technology, in
aerothermodynamics and in thermal protection systems. Of course, there’s
not that many people involved in those areas. Luckily, some of the
universities are actually building up that capability, but there’s
some universities, like my alma mater, University of Texas in Austin,
they just don’t deal with that much at all anymore. I don’t
know why—hypersonics in general. Some universities are building
up that capability, so there’s sort of a lack of recognition
of how important that is. Even though right now, for example, in Orion,
they say the heat shield is one of the top risks for the program,
so there’s an understanding that it is very risky.
I think that the JSC facility, obviously those who’ve tested
in it over the years, have a lot of respect for it. Let’s just
put it this way: we were too busy working hard on it than to publicize
its benefits. Literally, we did not spend time doing that. We were
too busy getting the job done. We didn’t walk the halls of [NASA]
Headquarters [Washington, DC] or tout our capabilities to everybody
and their brother; we were too busy just doing the job. Maybe it just
went unrecognized a little bit. It’s sad that we don’t
have that test capability. The other thing, too, is the JSC facility,
they’re a pretty lean, mean group, so we didn’t cost a
lot of money. NASA Ames has got, like I mentioned before, a lot of
infrastructure, a much larger test team, so they’ve got to support
all that, and that takes lots of money.
Now, there’s no alternative. You’ve got to go to Ames,
and you’ve got to pay lots of money. Time and time again, I
just hear people saying, “Well, I can’t afford that.”
That can be professors, it can be companies, it can be technology
folks. They just say, “I can’t afford that much money
to go test.” I’m real worried about how the progression
of this technology is going to happen in the future because I just
don’t feel like the amount of testing is going to be there.
You really do learn by testing, over and over again.
What would you say have been the benefits of the JSC facility?
think just being a workhorse, first of all. Just getting lots of testing
done. That’s one of the big benefits and of course, working
together with the Ames folks, et cetera, getting some of the diagnostics
together on the laser-induced fluorescence, a lot of that work was
done at JSC. Some of it was done at Ames. I think that test capability,
its progression and technology development, a lot of it was because
of what we did here at JSC. I think that was real important to the
Agency, and to other people, too—professors, et cetera—they
understand that. We don’t deal too much with the DoD [Department
of Defense] folks. Occasionally, seems about once a year, we talk
to them and exchange some notes here and there. It just depends on
where they’re headed to.
I think that there’s capabilities that the JSC facility has
that they could have used more. Sometimes we had to turn them away,
especially during the Shuttle program, since our top priority was
supporting Shuttle with the JSC Arc Jet. I know that I personally
was involved in planning some tests for the Air Force in our facility,
and they never happened because we were just too busy supporting Orbiter.
The Air Force just couldn’t get it done then and said, “Okay,
you can’t help me; I’ve got to go someplace else or not
What impact is the closure and the demolition going to have on the
say we don’t know. Ames, they did take some of our hardware
and tried to get some of the capabilities that Ames did not have but
we had. They tried to get that up and running. They were supposed
to have that all done here, by the beginning of this year, and they’re
still having problems. There are schedule delays, so there’s
testing right now that Orion is supposed to be doing, and it’s
been delayed again and again and again. Part of it’s just a
learning curve you go through. The folks at Ames, their technicians
and engineers, are not familiar with this type of arc heater system.
Obviously, if you’re not familiar with it, you’re going
to have problems with getting it going. I’m sure there’s
going to be delays in testing—already is. How that’s going
to impact us, it’s hard to say. At some point, it won’t
be pretty, it’ll be us technical folks saying, “Yes, but
you need to get these tests done,” and you get program management
saying, “I can’t delay my schedule; you guys need to solve
It’s like, what do you do, right? At what point do you say,
“Okay, I’m going to delay a flight because I can’t
get my testing done?” You have everybody and their brother on
you at that point. I don’t know where it’s headed, and
that could be pretty ugly. I also expect us to have to accept more
risk because we’ll be able to test less. Again, that’s
not a good position to be in because it’s where us on the technical
side are willing to draw our line in the sand and say, “I’m
not going to approve us flying unless we get these tests done.”
You can imagine, that’s a very tense and stressful situation
to be in. It just depends on who’s making the decision, whether
they’re going to cave or not, because of the tremendous program
pressure to move on. They don’t understand.
That’s one thing that dismays me about NASA today, compared
to Apollo days, is that if you go back to look at those folks who
are heading up the programs and heading up NASA, every one of them
was technically very, very competent. You could go have a technical
discussion with them in detail, and they knew what they were talking
about. Today, that’s less so. Managers tend to be managers and
not so technically competent. They themselves are not able to make
those technical judgment calls. They’re having to rely on their
lower-level technical folks. That causes a problem because if they
don’t understand, then they’re more willing to just take
the risk without really understanding the risk. Then it depends on
the lower-level technical folks to be more adamant to stick their
necks out and yell and scream to say, “Hey, you can’t
do this, you really do need to test more,” for example. That’s
sort of my crystal ball of what’s going to happen.
Just to backtrack a little bit, too, I’ll give you an example
of the technical competence of our leaders. I’m not sure if
everybody knows, but the first Center Director for, I guess it was
the Manned Space[craft] Center at the time, was [Robert R.] Gilruth.
He actually owns a patent for Arc Jets, so he himself, obviously,
was instrumental in getting the Arc Jet capability going here at JSC.
I don’t think he’d be very happy with the situation today.
You go look at a lot of these folks—Max [Maxime A.] Faget and
Chris [Christopher C.] Kraft and all those—and how competent
they were, technically, and understood the re-entry problem; different
It’s hard to say what the future’s going to hold, but
two things, I think, to summarize: I think we’re going to have
more schedule delays because we can’t get our testing done,
and then also, we’re going to have to accept more technical
risk. Hopefully, it’s not too much. We don’t want to lose
I think everyone would agree with that. We’re almost at 11:00.
I had one more question, but I wanted to see if there was anything
else you wanted to talk about, maybe we haven’t discussed about
the Arc Jet? Trying to be thorough and think about different aspects
of working out there.
I can’t think of anything. I’ve been talking here for
a while. Nothing’s jumping into my mind at the moment.
I like to ask people, because I’m guessing in the future, people
are going to wonder, what were things like in terms of camaraderie
out at the Arc Jet? Anything you can talk about and share?
probably like every office. There’s good times and bad times,
right? There’s different levels of competence. Some people are
very competent at what they do and some are so-so. Learning how to
deal with that is always difficult, but still being respectful to
the people. I think there was a lot of camaraderie, especially in
the Arc Jet Team, there’s people who’ve been there for
years and years and years, and so you get to know them and respect
them. Some of the new folks that came in performed well, some of them
didn’t, but I would say yes, there was a lot of camaraderie
amongst those folks. I always had a lot of respect for them. At least,
I think they respected me and they usually treated me well when I
walked in the door. I really liked working with those people. I think
in general, they were really top notch.
I think that’s a good note to end on. Thank you very much for
coming in today. Appreciate it, enjoyed it.
[End of interview]
to JSC Oral History Website