Space Access Update #71, part 1 - 5/4/97
Copyright 1997 by Space Access Society
_______________________________________________________________________
Yes, it's been six months since we put out an Update. We've delayed for
a variety of reasons - we didn't want to get into detail on NASA X-33's
problems while the coming year's funding was still vulnerable, for one.
We were not, for that matter, completely agreed among ourselves on the
nature and severity of the problems until quite recently. And frankly,
we were more than a little burnt out after spending much of the last
nine years working to bring SSTO to respectability. The urge was strong
to tell ourselves everything was fine, we'd succeeded, we could pass the
torch on to a new generation and go back to tending our own gardens.
Alas, it looks as if what we started is now, left alone, as likely to
discredit the whole idea of cheap access via fast-turnaround reusable
rockets as to prove it. And the younger generation seems not yet
totally cognizant of the nuances. So we're baaa-ack! And as long as we
can't retire after all, we intend to have some fun.
We're going to start with a two-part Update dedicated to what's been
happening with X-33 this last year. This part 1 begins with a summary
of our views, then covers X-33 configuration and technical issues. SAU
#71 Part 2 will cover political and organizational aspects of X-33.
_______________________________________________________________________
Disclaimers
It's fifty years since the Cold War started, forty since Eisenhower
warned of a "military industrial complex" threatening to become the tail
wagging the dog, and well over thirty since NASA was founded and given
responsibility for US civilian space exploration. Both NASA and the US
government-aerospace complex have grown large, powerful, and inflexible
in the decades since. Both have accumulated a lot of bad habits.
It's not our job to reform NASA in all its widely distributed diversity.
Nor is it our job to reform the US government-aerospace sector in
general, nor Lockheed-Martin Corporation's particular collection of
dubious practices. Nor for that matter is it our job to fix any of the
other major government aerospace contractors' various failings. Life is
too short.
It is our job to promote radically cheaper, more reliable, widely
available access to space, ASAP. Period. Any reforms we end up
pushing, explicitly or by implication, are purely a means to that end.
Anyone who doesn't want us pointing out places where their pursuit of
organizational self-interest conflicts with the overall public interest
can stop us easily: Clean up your act with regard to government reusable
launch vehicle (RLV) R&D. We don't care what you do elsewhere.
We do not claim to have a great deal of instant clout. What we have is
a fundamentally sound idea, good information, better advisers, and a lot
of persistance. The Administrator of NASA last fall accused us of
"nipping at his heels" after we'd buttonholed him for several minutes
about some of our concerns. Yes, sir, and proud of it - that's what we
do. Occasionally, of course, we convince someone who does have clout to
support us on one point or another... But mostly we persist.
_______________________________________________________________________
X-33: One Year Later
The best way to describe where we are is to go back over how we got
here. It's all considerably clearer in hindsight... What follows is
our analysis of the last year or so of the ongoing NASA X-33 process,
based on information ranging from official published statements through
reliable sources down to plausible rumor. We've spent a good bit of
time working on this - we think we have a pretty accurate fit to the
data. Your mileage may, of course, vary.
Summary
X-33 has serious problems. We think those problems mainly come from:
- Inclusion of too many new technologies in what should have been a
fast-turnaround lean operations demonstrator using mostly ready-to-go
technology. Much of the new tech is having teething troubles.
- Possible lack of commitment to project success (as opposed to bidding
success) by Lockheed-Martin's top-level management, with consequent
imposition of inappropriate project organization and inattention to
adequate project support. There's also a certain amount of unsolicited
inappropriate "help" (and occasional outright sabotage) from various
other parts of NASA.
We think X-33 can still end up being a useful proof-of-technology
X-vehicle. The NASA people involved show some signs of learning from
their experience. The key, in our opinion, is concentrating minds in
Lockheed-Martin's top management on doing what it takes to significantly
improve the odds of project success, while continuing to fend off
extraneous "help" from elsewhere in NASA.
We recommend a two-track policy toward this end. One, NASA HQ should
maintain axe-poised oversight on X-33 cost, schedule, and technical
milestones. The contractor has to be made to understand that they are
in genuine danger of losing the project if they mess up too badly. The
threatened cancellation of the "Clark" science satellite for exceeding
Dan Goldin's new 15% cost overrun limit could help in this regard.
Two, there should be credible and vigorous competition for the project,
in DOD, NASA, or (preferably) both, to ensure that contractor top
management understands that even if they get away with failing
protractedly, they will not buy much extra time for their existing space
launch cash cows. They must understand that their main option for
remaining competitive in space launch past 2000 is to do what it takes
to make X-33 succeed.
Lest anyone take this as mindless attack rather than constructive
criticism, we do support continued funding for X-33, pending the results
of this spring's scheduled Critical Design Reviews, the final step
before freezing the design and committing to construction. (We have
however just heard that the CDR's have been postponed to allow more work
on reducing the current X-33 design's 35% over-weight problem. We await
the eventual CDR schedule and results with considerable interest.)
_______________________________________________________________________
X-33 Technical Description And Current Status
Last year we described Lockheed-Martin's winning X-33 design as the most
"elegant" one submitted, the one that packs the most sophisticated
components into the smallest most closely integrated package. This
sounds wonderful - until you have to either compensate for components
not turning out quite as well as you'd hoped (closely integrated means
lots of interaction between the pieces; lots of other components are
affected) or, once it's assembled, until you have to go back in to fix
something. Small closely integrated packages are a royal pain in the
butt to service.
But NASA's Old Boy Net, bless their ivory-tower souls, think maximum new
technology and "elegant" complexity are just peachy. (Increased
operating complexity? No problem, we'll just pile on more guys with
clipboards and checklists. They're on the payroll already anyway...)
And NASA's Old Boy Net has, we've discovered repeatedly over the last
year, a lock on the NASA source selection process. (More on that some
other time - suffice it to say for now that NASA needs to take a serious
look at how they might find truly impartial people to serve on selection
boards.) (The White House, by the way, also had a hand in skewing the
selection criteria toward excessive new tech, as part of the deal they
made to allow project go-ahead - but it's unclear how much of those
provisions originated there, and how much was whispered in their ear by
the Old Boy Net.)
But, we do have to admit, L-M's "VentureStar" X-33 design is indeed
downright elegant. More important from our point of view, the various
advanced technologies that have to come together to make VentureStar
work - the aerospike engines, the multilobe composite tanks, the
metallic thermal protection - all can be useful to other SSTO
configurations, if as we suspect VentureStar turns out (even at best)
less operationally flexible than optimum for a competitive general-
purpose commercial space cargo ship.
Enough cavilling. On to the design of this X-33 single-stage reusable
space rocket demonstration vehicle.
- Aerodynamics
L-M's X-33 is a "lifting body", a blunt triangular wedge-shaped wingless
vehicle that, when it is in horizontal aerodynamic flight, gets its lift
largely from the airflow around the fuselage. X-33 is designed to
takeoff vertically and fly into space under rocket power, then re-enter
the atmosphere as a relatively high performance (high lift-to-drag ratio
and thus high maneuverability, high "crossrange") hypersonic glider.
Once slowed to subsonic speeds, it is designed to still have good enough
glide characteristics to make an unpowered runway landing with
reasonable reliability and safety.
This combination of good hypersonic re-entry and subsonic glide
performance is one of the keys to making this X-33 design work - L-M
claims to have a proprietary aerodynamic shape that will provide both.
This is one of the first places we come to where X-33 is running into
problems. L-M may well end up meeting their aerodynamic performance
claims - but it seems likely from the significant vehicle shape changes
we've seen that L-M didn't know as much as they claimed back when they
were bidding. The small tip fins of earlier iterations have grown to
small wings, and the overall vehicle shape has changed markedly. More
on this when we talk about the internal structures.
- Engines
This X-33 is powered by a pair of Rocketdyne "linear aerospike" rocket
engines, burning liquid oxygen (LOX) and liquid hydrogen (LH2).
"Aerospike" is an unconventional type of rocket engine that gets thrust
by expanding gases against the surrounding air (if any) and the outside
of the engine, rather than against the inside as with conventional bell-
nozzle rockets. A "linear aerospike" is one where the combustion
chambers are arranged in two straight rows, one along each side of the
wide base of a truncated-wedge aft-facing expansion surface, rather than
in a circle around the base of a truncated-cone expansion surface (an
"annular aerospike".)
L-M chose linear aerospike engines primarily because they integrate well
into the lifting-body vehicle shape chosen - they blend into the tail
better and don't extend as far aft as bell-nozzle engines, reducing the
center-of-gravity problem this sort of vehicle has from engine weight in
the tail. The secondary reason was that aerospike engines provide good
performance from sea-level to vacuum without either going to very high
operating pressures (SSME's work at ~3000 psi, the X-33 aerospikes at
around a third of that - high-pressure pumps tend to be heavy, fragile,
or both) or mechanically changing the expansion nozzle geometry with
altitude.
Aerospike engines have never flown, but they have been built and run on
ground test stands, the best-known occasion being in the early seventies
when Rocketdyne did considerable work on linear aerospike as a potential
Shuttle engine. After that fell through, the project was shelved until
L-M settled on the concept for their X-33 bid.
The X-33 engines are direct descendants of those 70's test-stand
engines, with new combustion chamber feeds but otherwise little changed.
(We hear Rocketdyne has had to track down and hire some of retirees from
that project for their lost expertise.) The propellant pumps are still
taken from the old Saturn 5 J-2 upper stage engine. Unlike the test
stand versions, the plan is that X-33's engines will each produce about
the same thrust as the J-2 bell-nozzle engine their pumps came from,
something over 200,000 lbs thrust per engine. As best we can tell, the
70's tests used only part of the J-2 pumps' capacity.
X-33's main method of steering in powered flight will be "thrust
vectoring" via differential throttling of the engines - no mechanical
gimballing. The main engine combustors are arranged in four rows or
banks - looked at from the rear of the vehicle as it sits horizontally
on a runway, the (horizontal) banks are top left, bottom left (left
engine), top right, bottom right (right engine.) X-33 won't really have
two completely separate engines; there will be side-to-side propellant
cross-connects, both for side-to-side steering and so one set of pumps
can feed both sets of combustors and keep the ship flying if the other
pumpset fails. X-33 would be able to handle an engine-out at up to 90%
propellant load, assuming the original planned vehicle weight and engine
thrust values and a 20% power reserve on the pumps.
X-33 thrust vectoring will be via diverter valves on each side between
top and bottom banks, plus diverter valves between the two sides. This
speeds response time and saves thrust-losses over throttling the pumps.
We have heard Rocketdyne is having a hard time getting sufficient
predicted thrust out of the X-33 engine design so far - we might hazard
a guess this is related to the propellant plumbing in this application
being far more complex than in the J-2 engine the pumps came from.
There was a news item recently that Rocketdyne wants to eliminate the
crossfeed ducting between the engines (and thus the engine-out
capability) but that NASA won't let them - this could be related.
L-M is also committed as part of their X-33 bid to have Rocketdyne
produce and run a test-stand demo version of the 400,000 lb thrust
super-lightweight linear aerospike engine for their proposed Venture
Star commercial SSTO cargo transport.
(Policy note: In large part, Lockheed-Martin won X-33 because their bid
included enough money to develop and demonstrate these new engines.
These engines are a major reason we still support X-33; they're
applicable to a range of other potential vehicle configurations. We'd
be very unhappy indeed to see either aerospike engine dropped from the
project after they were major factors in L-M's winning the bid.)
- Propellant Tanks
X-33's propellant tanks are another significant new technology required
to make the package work, and in this case we're pleased to report that
from what we know, the tanks are coming along well.
Some background... Generally, the largest single load rocket propellant
tanks have to deal with is internal pressure. Even pump-fed rocket
engines tend to need several tens of pounds of inlet pressure, and the
propellant tanks have to handle that pressure over huge surface areas.
You can keep a pressure tank extremely lightweight, as long as you have
thin high tensile strength tank wall materials, and as long as you then
don't fight a thin-walled tank's natural tendency to assume a round
shape under internal pressure. Build your tank square and you'll need
massive braces to keep it from inflating into a circle anyway when you
pressurize it... So most rocket propellant tanks are "figures of
rotation", shapes that are always circular in cross-section, with some
mix of straight, conic, and spherical sections viewed from the side.
The problem with this is that it limits what shape you can make your
rocket and still keep the tanks light. For a circular cross-section
rocket, no problem. For a squashed-wedge lifting body, well... The
solution is something called a "multi-lobed" tank.
A simple multi-lobed tank might be built up from two identical
cylindrical tanks. Slice one-third off each tank lengthwise, then
attach the pair of sliced tanks together side-to-side, butting together
the openings where you took off the slices. You'll have one tank with
two lobes, with a cross-section like a sideways "8".
Put pressure in this tank, and it'll try to expand into an "O" - you
have to add some sort of tension structure between the halves to hold
the sides of the "8" together. Now you have a stable two-lobe tank. A
lightweight stable two-lobe tank? Only if you can figure out how to
build it without a heavy flange where the two halves and the tension tie
join. These are non-trivial manufacturing problems; multi-lobed (two
lobes is just the simplest case) propellant tanks have stayed on the
wish-list till now. But they would be hugely useful in rocket lifting
bodies and other non-circular vehicles...
L-M has apparently solved the manufacturing problems. X-33 will have a
pair of 4-lobed graphite-epoxy liquid hydrogen tanks (the LOX tank will
be old-fashioned aluminum for now). The plan is to build the tanks in
four sections, "fiber-placed" by machine on forms, with a border of
"green" (unepoxied) fiber left on the mating edges. The edges of the
sections will be "woven-Y" joined along with a centerline tension-tie
truss, then epoxy-impregnated, then the entire tank will be place in a
large autoclave (at least 15'x25'x40', our estimate of the tank
dimensions) that Skunk Works just happens to have lying around, and the
entire tank will be cured into one piece with no heavy flanges.
As of last winter, the techniques had been tested on small sections and
there seemed to be no show-stoppers. Cryogenic insulation and
stiffeners (where required) will be on the outside of the tanks.
- Structure
Much of the elegance of L-M's X-33 design lies in the fact that it has
very little structure per se. Of the four main structural elements,
three are propellant tanks. The liquid oxygen (LOX) tank forms the
ship's nose, the two liquid hydrogen (LH2) tanks are connected to the
LOX tank to form the ship's two sides (with the payload bay in the space
between them), and the aft ends of the LH2 tanks are connected to a
cross-truss that also serves to mount the engines and various
aerosurfaces.
X-33 has no solid outer hull as such - just a relatively light
assemblage of latticework and standoffs that carry the metallic thermal
protection shingles that define the ship's aerodynamic shape.
X-33 has been having serious weight-growth problems, however, and some
of them relate to the structure. One problem is that thermal loads are
turning out to be higher than anticipated, and the TPS shingle attach
structures are getting heavier. Another seems to be that the
aerodynamic shape has changed since the tank shapes were fixed, so
considerably more standoff structure is required to make up the
difference. And another is that the center-of-gravity ended up too far
aft - engine weight growth? aerodynamic changes? weight of the larger
aerosurfaces? - and (in the current design iteration at least) has to be
compensated for by several thousand pounds of lead ballast in the nose.
- Thermal Protection
X-33 TPS is supposed to be advanced lightweight metallic "shingles" on a
lightweight composite standoff structure. The shingles are supposed to
be a thin metallic outer layer, over a honeycomb core for stiffness,
with a bottom layer of ceramic insulation to reduce heat transmission to
the interior of the ship.
We understand there are problems with the TPS so far - details are
sketchy. We mentioned heat loads to the standoff structure being higher
than anticipated previously. We've also been told that inconel is being
substituted for titanium aluminide for the shingle outer skins for cost
reasons - this would account for some weight gain, as inconel is not
light. We would guess that inconel foil outer skins would also have
some durability problems, denting easily under raindrop impacts and
such. We assume that inconel will be just a placeholder for X-33 and
that any commercial followon would require the lighter stiffer TiAl -
which we expect this X-33 program will still develop and test.
- Flight Control Software
X-33 flight control software has some difficult challenges to meet. In
particular, on the ship's first flight the software will have to deal
with keeping the ship stably on course through flight regimes where the
engine efficiency and thrust-vectoring responsiveness won't be known
precisely in advance. Part of the answer to this will be to increase
use of the ship's aerosurfaces for steering while under power in the
atmosphere. Part of the answer, we suspect, will be a lot of muttered
prayers and crossed fingers for the first few minutes of flight #1...
We are told that X-33 flight control software algorithms are being
designed at NASA Marshall, that Allied Signal Corp is then coding them
in C++ offsite, and that the code will then be tested at an Integrated
Test Facility (an "iron bird" ground test rig) at Edwards AFB. This
strikes us as a likely formula for Software Project Manager ulcers, late
code, Ariane 5 style fly-sideways code, or all of the above. Software
may yet end up as the long pole in the X-33 tent. We'll see.
- Flight Test Ops
The X-33 Cooperative Agreement calls for 15 flights starting in early
1999, culminating in several flights that will reach Mach 15 (about 60%
of orbital speed) by late 1999. The agreement also calls for two two-
day turnaround demos plus several more seven-day turnarounds.
X-33 flights will launch from a site west of the Edwards dry lake bed,
near the USAF Phillips Labs rocket test area. The X-33 will be returned
after flights on the back of a NASA 747 "Shuttle Carrier Aircraft".
The first two flights will cover ~100 miles, to Silurian Dry Lake Bed,
reaching max speeds near Mach 4 and max altitudes near 116,000 feet.
X-33's VTHL (vertical takeoff, horizontal landing) configuration makes
any less drastic first flight very difficult - the vehicle needs
considerable altitude and airspeed to safely make the transition to
horizontal flight so it can land. Less risky incremental "bunny-hop"
hover tests are right out.
The next series of flights, ten max, will be to Michael Army Airfield at
Dugway Proving Grounds in Utah, and will range from Mach 9-12 at up to
164,000 feet. The final series of up to three flights will be to
Malmstrom AFB in Montana, covering ~1000 miles, reaching ~250,000 feet,
at speeds of up to Mach 15 - if they manage to trim enough weight from
the vehicle to make that performance. The current design iteration is
projected to max out at Mach 13 or so.
- Summing Up, part 1
The preceding isn't an attack on the competence of the working engineers
actually trying to build and fly X-33. More on the engineer-management
divide in part 2... As best we can tell, these problems are a mix of
perfectly normal solvable teething troubles, plus the contractor top
management's skewing the bid toward new technology for its own sake to
win the bid, in turn a result of both the new-technology requirements
the White House imposed before giving the go-ahead, and of NASA's new-
technology uber-alles reflexes. There's also the contractor top
management's possible lack of commitment to ensuring X-33 succeeds now
that they've won the contract - more on that in part 2 also.
Meanwhile, we find it more than a little ironic that, while X-33 still
has a good chance of being a useful X-vehicle technology pathfinder, it
is turning out to be a very poor "Y-vehicle" prototype for Lockheed-
Martin's proposed Venturestar Shuttle replacement. X-33's problems
point out graphically how much trouble Lockheed would have been in if
they'd gotten what they were pushing for two years ago, government
market guarantees for their going straight to developing Venturestar
with no intermediate step. And yes, we told you so at the time, guys.
** continued in Space Access Update #71, part 2 **
Space Access Update #71, part 2 - 5/6/97
Copyright 1997 by Space Access Society
_______________________________________________________________________
Yes, it's been six months since we put out one of these. This is part 2
of an Update dedicated to what's been happening with X-33 this last
year. Look for SAU #72 with broader coverage, RSN - because there are a
whole lot of things happening besides X-33.
_______________________________________________________________________
X-33 Organization and Politics
NASA has problems. As far as they're concerned, X-33 is far from the
largest of these - between Station and Shuttle, they have bigger fish to
fry, with troubles far more obvious and a whole lot more funding at
stake. We take a somewhat different view, but then we would - we think
cheap reliable transport is fundamental. Station in particular we see
as massively transportation-constrained...
X-33 we see as in real danger of failing - failing first flight, or
turning into a "NASP II" technology playpen and never flying at all - in
part because it was poorly-conceived (too many new bleeding-edge
technologies included, too many Shuttle-replacement expectations tacked
on, more a premature operational prototype "Y" vehicle than an
experimental "X" ship) and in part because as best we can tell, the
contractor top management has no urgent incentive to ensure that X-33
succeeds. Make no mistake, we'd like to see X-33 succeed. But if it
does fail, we don't want to be hearing any nonsense about the failure
proving SSTO can't work.
Meanwhile, reforming NASA is not our job - our sole purpose is promoting
affordable reliable access to space for all, ASAP, by whatever means
will still let us sleep at night. But NASA's institutional problems do
have a lot to do with what we see gone wrong with X-33. So, for that
matter, do the institutional tendencies of the US aerospace industry in
general and of Lockheed-Martin Corporation in particular.
So we'll talk about these for a bit.
Background: NASA and the Contractors
Organizations are like people, in that they have histories, tendencies,
habits, quirks - reasons for doing the things they do.
NASA is a functionally and geographically diverse collection of "mature"
government bureaucracies, warring with each other over turf and budget,
reluctantly travelling in loose formation and paying attention to NASA
HQ in Washington whenever they absolutely have to. (With the notable
exception of Johnson Space Center (JSC), NASA's 800-pound manned-space
gorilla, most of them have ended up paying considerably more attention
to HQ since Dan Goldin took over.)
Lockheed-Martin is a major government aerospace contractor, one of the
final survivors of forty years of Cold War followed by eight years of
"defense consolidation", operating in a current business climate where
the stockmarket instantly punishes the slightest lapse of attention to
next quarter's bottom line.
Both organizations are pretty well set in their ways by now. Both do
have a lot of good people doing the best jobs they can. Both also have
some very predictable collective tendencies, tendencies that make sense
in terms of institutional self-preservation but that, uh, aren't always
in the best interests of the US taxpayers who foot the bills.
- The Greying of NASA
Aging bureaucracies are marked by a tendency to divert ever more of
their resources into organizational structure and ever more of their
efforts into defending their bureaucratic turf, to the detriment of
whatever the nominal mission is supposed to be - in NASA's case,
advanced space and aeronautical R&D, plus advanced space exploration.
The limiting case for bureaucratic "maturity" is when output drops to
the point where the bureaucracy is in danger of losing its funding. In
NASA's defense, they're far from the worst mature federal bureaucracy in
the US - the nature of NASA's missions are such that they attract a lot
of talented people who get useful work done despite all the obstacles,
and NASA's job is too high-profile for them to ever get away with zero
(or even negative) output. Unlike some... But nevertheless, there's a
lot of friction to overcome anytime something needs doing in NASA. Much
of the agency, alas, is mainly concerned with making sure the paperwork
proves nothing was anyone's fault, while marking time till retirement
rolls around - totally averse to allowing anything risky (like flying
real X-vehicles) anywhere nearby.
- Turf Defense
Reusable Launch Vehicle (RLV) work at NASA treads on all sorts of
existing bureaucratic turf - Marshall and Stennis on engines, Michoud on
tankage, Langley on vehicle configuration and aerodynamics... Most of
these overlaps have been resolved, generally by farming out part of X-33
to the affected center - this has had its effect on the shape of the
program, in general spreading it thinner and increasing costs. Shrug.
This was likely inevitable once SSTO was assigned to NASA.
The main conflict that hasn't been (and likely won't be) resolved is
X-33's overlap with NASA's 800-pound manned-space gorilla's toes.
Johnson Space Center in Houston is the center for manned space in NASA.
Between Shuttle and Station, JSC at this point controls around a third
of NASA's total budget. X-33 the way we originally envisioned it, as
a harbinger of radically cheaper more frequent more widely available
spaceflight, endangers JSC's entire way of life, built as it is around
scarce, expensive, exclusive space access. The JSC manned-space mafia
(by far the most powerful faction within NASA's Old Boy net) has reacted
predictably, with a two-pronged effort to either capture X-33 and fit it
into their existing structure, or to render it ineffectual.
NASA HQ thus far has resisted outright JSC capture of X-33 - it
continues to be run out of HQ and NASA Marshall, with JSC involved only
in an advisory capacity. In particular, JSC's "man-rating" bureaucracy
has been kept away from X-33, lest they triple the time and increase
costs ten-fold - "man-rating" is the 1960-vintage process of ex post
facto inspecting in quality to bring inherently 90% reliable artillery
rockets up to 99% reliability so astronauts can ride them. A major
point of X-33 is to bring inherent, by-design reliability up to several
nines beyond 99%, rendering the whole "man-rating" process obsolete.
The manned-space Mafia has had its effect, though, notably in the
insertion of Shuttle-replacement prototype requirements into the X-33
CAN, trying to force-fit X-33 into their world. They're largely
responsible too for the erroneous public impression that X-33 will lead
directly to a Shuttle II, via repeated statements to that effect in the
media, despite Administrator Goldin's repeated assurances to the
contrary.
There have also been a number of instances of what look like outright
sabotage attempts against X-33 - statements to the media that flying
X-33 over land is far too risky and the program should be converted to a
series of ground technology demonstrations, press releases faxed (with a
JSC fax number still listed on top, tsk tsk) to towns where X-33
environmental impact hearings were due, telling those towns X-33 would
surely rain flaming death down on them... Naughty naughty, boys.
- On to the Contractors
US government aerospace contractors, meanwhile, have spent fifty years
getting far too used to catering to a single customer with finicky
tastes and bottomless pockets. Guessing what this customer really
wants, promising it in spades regardless of actual current capabilities,
then spending whatever it takes in money time and talent to deliver
something more or less resembling what was promised - this has been a
way of life for generations of executives and engineers. The corporate
culture that has resulted is not well attuned to anything resembling an
open commercial market. (It's not at all clear that any of the current
major aerospace outfits will end up being major players in the 21st
century spaceliner market. How many buggy manufacturers survived the
jump to automobiles?)
We'll digress for a moment to express our opinion of "defense
consolidation". The US government, post-Cold War, has been pursuing the
appallingly stupid policy of not only waiving antitrust, but actually
twisting arms and paying cash to the former half-dozen or so major
aerospace contractors, to encourage them to merge into two mega-
contractors, on the dubious theory that this would encourage efficiency
and save the government money.
The practical result is that the US will very shortly be reduced to a
grand total of two design bureaus capable of dealing with large complex
aerospace systems. This is more massively monopolistically inefficient
than the Soviets at their worst, and we're already reaping the harvest,
with one of the soon-to-be-two remaining majors declining to bid on a
multi-billion dollar NASA space operations contract. Rotsa ruck driving
a hard bargain with the other one, guys.
Where were we... Right. Meanwhile, Lockheed-Martin is, like most such
organizations, sharply stratified into two layers - more than a little
schizophrenic. At the top is the political-management level, VP's and
up, of necessity obsessively concerned with "stockholder value", in turn
closely linked to the coming quarter's profits. The carrot is the
market value of their stock-options; the stick is that the big
institutional stockholders, the huge pension and mutual fund managers,
will turf their butts out if they don't deliver strong stock prices.
The legendary Norm Augustine said a couple years back that if (then)
Martin-Marietta could get a better return for investing in gravel pits
than in space-launch vehicles (Martin had a construction materials
subsidiary) he'd damn well invest in gravel pits. People at this level
have no choice but to be "stockholder value" nuts - that's just the way
it is in US business these days. Rocket nuts need not apply.
The other, far larger, corporate level is the grunts, the troops, the
hire-and-fire interchangeable cannon-fodder - the people who spend a lot
of their lives doing viewgraph design studies for, say, rockets, who
eventually, if they're very good and very lucky, get funding, get some
semblance of a chance to actually build and fly something.
Grunts, of course, have to work with whatever resources the political-
management types will give them plus whatever they can scrounge - grunts
who assume that management will as a matter of course back them with
whatever they need tend to aquire ulcers and permanent puzzled
expressions. Management has its own priorities, up to and including
assigning grunts to projects that management in its heart does not give
a damn about the success or failure of. Grunts have even been known to
be fired for succeeding at projects management wanted to fail... But a
true grunt says, screw management, flying rockets is what really counts,
and runs for daylight if ever he is fortunate enough to glimpse it.
The X-33 CAN
We bitched about various things as NASA went though drafts of its
proposed X-33 "Cooperative Agreement Notice". Some they acknowledged,
some they didn't. In 20-20 hindsight, the most important points we made
were these: Throwing in all the Shuttle II prototype requirements would
drive up costs and risks; it was too soon for a prototype - what was
needed was an X-vehicle. Insisting bidders cough up a share of the
expenses for what was nominally an X-vehicle would inevitably drive them
to figure out how to earn a near-term return from the project, given the
profits-now orientation of today's corporate culture.
And earning a near-term return is largely incompatible with the basic
concept of an X-vehicle. The point of an X-vehicle project is to gather
data as quickly and cheaply as possible on what happens when a given set
of technologies are pushed to new limits. X-vehicles should have no
mission but expanding the envelope and no payload but pilots and
instruments. X-vehicles are essentially disposable; you build three
because you expect to break one or two as you fly them and learn from
them. AFTER you've built and flown an X-vehicle, THEN you know how to
build and fly a prototype of something that will make money.
But that takes too long for the market to wait for, typically three
years for the X-vehicle and another three to five for the operational
"Y-vehicle" prototype to follow. That's why we think the government has
a legitimate role building X-vehicles - they're an investment that
benefits the whole country, but the payoff takes too long for commercial
financing in this impatient age. After X-vehicles have shown the way,
then the commercial sector knows what it takes to build commercial
ships.
(And in case you're still wondering, no, X-33 is not a genuine X
vehicle. It's a bastardized X-Y hybrid with a lot of marginally related
ground technology projects grafted on. We hope a useful X-vehicle plus
some useful new technology developments can still be salvaged.)
NASA needs to keep this X-Y distinction carefully in mind in the future,
in order among other reasons to avoid competing with private commercial
efforts. A good rule of thumb: If it can be flying missions and making
money in three years, it probably isn't "X" and NASA probably shouldn't
be doing it.
- X-33 Proprietary Rights: Results for One
Meanwhile, NASA seems to have handed Lockheed-Martin far more in the way
of X-33 proprietary rights than they should have, given that X-projects
are supposed to benefit US industry in general and that we taxpayers are
covering 80% of the tab. We hear conflicting reports on this, but the
ones we trust most say that Lockheed-Martin has exclusive rights to much
of the X-33 technology for several years after project completion. It's
hard to say for sure though, since out of seventy or so pages in the
actual signed X-33 cooperative agreement, NASA has only released about
twenty. The rest is allegedly "proprietary" and hasn't been released.
Including the progress payment and tech milestones schedule... (FOIA,
anyone?)
This makes it rather difficult for Congress to oversee the project, for
one thing. For another, it seems likely that even if L-M succeeds with
X-33, they could sit on the data until the government offers them a
really favorable deal to build a followon.
- The Contractor Contribution CAN-Can
As we alluded to above, there was a provision of the X-33 CAN
(Cooperative Agreement Notice, the document outlining competition
requirements) that in hindsight is a classic example of The Law Of
Unintended Consequences. We groused at the time about a major part of
the competition being how deep each bidder would dig into their own
pocket to build X-33 - we assumed that in the current US stockholder-
value-uber-alles business climate, this would drive bidders to figure
out how to show a near-term return on their X-33 investment, lest their
boards fire them for incompetence.
We lacked imagination - we assumed the bidders would do this by grafting
a payload bay onto X-33 and use it, post-test, for popup launch of upper
stage plus smallsat payloads. We worried that superimposing this
operational requirement would complicate the program, increasing cost
and time. And in fact all three X-33 bids did have some level of
provision for a small payload bay...
- Investing In The Future or Protecting The Present?
What we overlooked was that X-33 was both a medium-term threat to
existing corporate space-launch cashflows via its notional commercial
followon, and also a big enough project that it was very likely to (and
in fact did) soak up almost all available government RLV research money
for an indefinite period. (We also assumed that the X-33 technology
would be made available to US industry in general. The degree of
proprietary rights L-M seems to have negotiated for their 20% of the
cost astonishes us.)
In other words, a clever bidder CEO could show his board a relatively
near-term payoff from investing in X-33 even if it never flew a
commercial payload, indeed even if that X-33 crashed on the first flight
and never led to a commercial RLV followon. The near-term payoff for
winning X-33? Protecting the winner's existing space-launch business
against any low-cost RLV competition for a decade or more, by preventing
any competitor taking X-33 and leveraging it into an eventual successful
commercial RLV.
The X-33 bidder contributions proposed were in fact roughly proportional
to each bidder's existing annual space-launch cashflow. Lockheed-Martin
has half of the United Space Alliance Shuttle consortium, plus Titan 4,
Commercial Atlas, LMLV, their Russian Proton marketing partnership -
something over three billion dollars of annual cashflow. Lockheed-
Martin put up about $250 million toward X-33. Rockwell had the other
half of Shuttle plus Rocketdyne's expendable engine business, something
over one billion a year, about a third of Lockheed-Martin's launch
cashflow, and we understand that Rockwell bid a bit over one-third what
Lockheed-Martin did as their proposed share of X-33.
- Mac-Dac Folds
McDonnell-Douglas meanwhile had about 3/4 billion a year cashflow from
their Delta II operation, but put up essentially nothing in their X-33
bid. This looked like overconfidence at the time - they did have the
best proposal technically, operationally, and in terms of development
team experience - but in hindsight it was probably another symptom of
Mac-Dac top management's since-apparent fixation on cashing out their
company ASAP at the highest possible price. This also explains our (and
others) ongoing frustration with Mac-Dac top management: It was clear to
us that with DC-X/Delta Clipper they had the inside track on becoming
the Boeing of next decade's commercial spaceliner business - but they
repeatedly ignored opportunities to strengthen their position, and
ultimately blew themselves out of the race. Apparently they simply
didn't care about the company's long term future.
- And the Winner Is...
Anyway, one of the things this extra available money let Lockheed-Martin
do was include significantly more new technology work in their bid than
their competitors, notably including flightworthy aerospike engines for
their X-33, plus a promise of ground tests of a boilerplate version of
the larger engine for their proposed Venture Star commercial RLV. NASA
loves new advanced technology development, and they're inclined to see
more as better even in a project that doesn't really call for it. We'd
thought X-33 was supposed to be a reusable rocket fast-turnaround
operations demonstrator, not a new technology driver... Oh well. At
least we're supposed to get useful new engines out of this.
Other factors contributed to Lockheed-Martin's win - they played the
NASA selection process like a violin, with a bid carefully tailored to
match NASA's Shuttle-shaped vision of the notional future X-33 derived
RLV, with CAN details fortuitously changing to better match their bid,
with a corporate head of RLV who just happened to be the former Shuttle
mission commander of NASA's RLV boss. Lockheed-Martin won 67% of the
government projects they bid on in 1996, and we're told over 80% by
contract value. They're *good* at the bidding game. Maybe not so good
at delivering the goods on cost and on time afterwards - F-22, THAAD,
LMLV.... but they do win bids.
But the largest single factor in their winning X-33, in our estimate,
was that they could justify to their directors bringing more money to
the table in order to protect their existing space-launch cashflow -
purely an unintended consequence of X-33's bidder-contribution
requirement and its sole-major-RLV-project monopoly status.
Live and learn - and we hope NASA does learn. "Future X" should be set
up so there's never just one project eating most of the funding. Or if
that's completely unavoidable, at least set up the bidding so track
record and technical preparation count for more than ready cash.
- ...Lockheed-Martin, By A Split Decision
Meanwhile, last July, Lockheed-Martin had just been announced as the
surprise winner of X-33. The surprise winner, for a couple of reasons:
One, Lockheed-Martin's configuration was probably the least suitable of
the three for future fast-turnaround flexible-basing commercial
operations, what we'd thought X-33 was supposed to demonstrate.
Rockwell demonstrated considerably more effort toward minimizing the
ground establishment required for their (much simpler) VTHL vehicle,
while Mac-Dac led the field, having concentrated on these qualities from
the start.
L-M's however was best suited of the three to drop into the current
Shuttle operation with minimal layoffs, in our cynical opinion. It was
heavily sold that way; the promo graphics always showed it docking with
Station, for instance - but then if NASA allowed that to affect their
selection process, it's NASA's fault, not Lockheed-Martin's. (We
strongly recommend NASA take a serious look at recruiting source
selection board members from outside the NASA-Academia Old Boy Net.)
- Not Ready For Prime Time
Two, Lockheed-Martin was as best we can tell the least prepared of the
three bidders to go ahead and actually build an X-33. You'll recall
that by our analysis Lockheed-Martin's main reason to put significant
funds into X-33 was to protect their existing launch business - we have
strong indications Dan Tellep and Norm Augustine were actually thinking
along those lines, by the way - and thus could reach a major (if not the
major) goal just by winning the bid then sitting on it.
If true, this just might adversely affect the quantity, quality, and
timeliness of corporate resources available to the actual X-33 team
within Lockheed-Martin - the company has plenty of other projects, most
where they're not already locked into a win-win setup and many where
they can make actual profits. Why give X-33 one bit more access to
finite corporate resources than required to keep NASA from firing
Lockheed-Martin and starting over fresh?
Numerous indications we've gotten since the X-33 downselect support this
hypothesis. Both before and after the downselect, Lockheed-Martin seems
to have been allocating the bare minimum resources necessary to win and
then to keep the project.
One example: The LASRE SR-71 borne 1/10th scale transonic-airflow
aerospike-rocket efficiency tests. This project was touted as part of
L-M's bid, as providing essential performance data for aerospike engines
operating at low supersonic airspeeds, data not currently known to any
great precision. LASRE was originally supposed to fly before the
downselect. Last fall, NASA informed Lockheed-Martin that A: NASA still
expected LASRE results, but B: not one more cent of NASA money was going
into it. A year late now, LASRE is still ticking over, still not due to
fly till months from now - if ever.
As best we can tell what happened to LASRE, L-M assigned people to it
who had to learn on the job how to build a complex rocket combustor test
rig reliable enough to bolt onto the back of an irreplaceable Mach 3
aircraft, and apparently has since put the minimum necessary resources
into LASRE to avoid the embarrassment of formally shutting it down.
Mind, there's nothing wrong with on-the-job training for engineers - but
it shouldn't be misrepresented as a tight-schedule sure thing.
Another example is flight control software. McDonnell-Douglas as we
understand it had 80% of their X-33 software already running, with a
proven fast and reliable development setup already in place, using a
high-level language the flight control engineers could work with
directly - a result of their DC-X experience. Rockwell showed us what
they described as a prototype of their flight control software hooked
into a mission simulator at their X-33 bid open house last spring.
Lockheed-Martin seems to have started hiring on programmers after
winning the bid last July... We heard stories of programmers looking to
bail back out of the project shortly thereafter, for what it's worth -
we have no confirmation on that story. But it is true that top
programmers these days can afford to be picky about what projects they
stay involved in.
More serious (if true) are the recent rumors coming out of L-M that the
flight control software could be a year or more late. Again, we don't
have hard confirmation of this - but it fits with what we know about
software development in general and Lockheed-Martin's X-33 flight
control software setup in particular.
- Dat Organization, Dis Organization
We pointed out last year that Lockheed-Martin was touting their X-33 bid
as a "Skunk Works" project, but had meanwhile spread the project out all
over the map, to gain support for the bid within the newly-merged
Lockheed and Martin corporations, to gain support within NASA, and also
presumably to get closer to the "contractor in every congressional
district" ideal for a high-profile federally funded program.
Mac-Dac and Rockwell both seemed to be operating closer to the old all-
key-people-under-one-roof Skunkworks paradigm. Mac-Dac in particular
had already proven they had a tightly integrated fast-moving
"skunkworks" design shop via the DC-X and DC-XA efforts.
We've been assured that all the various X-33 contractor divisions and
NASA shops are working together in one harmonious "skunky" whole. We
remain skeptical, given the hints that has come out of the program so
far. Ultimately, of course, results are what count. We'll see.
- Summing up
We're reasonably sure Lockheed-Martin wasn't all that well prepared when
they won X-33. They showed signs of scrambling to hire enough pegs to
fill the holes right through last fall. Their project organization
still looks more widely dispersed than optimal. And their design shows
signs of not having been as refined as it should have been at the time
of the downselect.
In our experience, two things can happen to a project with this sort of
start. The confusion can settle down into order and converge on a
workable design. Or it can bog down and diverge into chaotic thrashing,
with results at best a crude and unwieldy approximation of the original
objective.
We thing the thing to do about X-33 for now is to wait on the results of
the CDR's, the Critical Design Reviews. The project is looking for a
little extra time to get their act together; within reason they should
get it. We don't expect X-33 on time or on budget at this point anyway.
We do expect that X-33 will fly in 1999. We also expect that if it runs
over budget, the overage will come out of Lockheed-Martin's collective
pocket, in cash, in reversion of proprietary rights to the government,
or both. We also expect that Lockheed-Martin will deliver on the
various new technologies they promised to win the bid, or pay for the
difference.
If the CDR's don't clearly show convergence on a workable design, if the
project runs significantly over budget and Lockheed-Martin refuses to
pay in cash or kind, if Lockheed-Martin tries to renege on any of the
major elements of their bid - we say kill the project and start it over
as a genuine X-vehicle project. We've spent a long time working on
this. A few more years to get things right won't kill us.
Meanwhile we recommend axe-poised NASA oversight plus strong external
competition (more on that in SAU #72) to concentrate minds in Lockheed-
Martin's top management on ensuring X-33 success, by removing their win-
by-failing option.
X-33 at best will still be more an overpriced premature prototype than a
genuine affordable-risk X-vehicle, but it's currently the only project
we've got, and it could still be pretty useful. The industry as a whole
could still get major chunks of useful new technology, and Lockheed-
Martin could end up with a design team that's learned how to do it right
next time. Despite the current "management" fad for treating the
techies as disposable interchangeable parts, a proven experienced design
team is a valuable commodity, a whole greater than the sum of its parts.
Spending what it takes to make X-33 fly could yet turn out to be the
best investment Lockheed-Martin could make for the coming century.
-----------------------(SAS Policy Boilerplate)------------------------
Space Access Update is Space Access Society's when-there's-news
publication. Space Access Society's goal is to promote affordable access
to space for all, period. We believe in concentrating our resources at
whatever point looks like yielding maximum progress toward this goal.
Right now, we think this means working our tails off trying to get the
government to build and fly multiple quick-and-dirty high-speed reusable
"X-rocket" demonstrators in the next three years, in order to quickly
build up both experience with and confidence in reusable Single-Stage To
Orbit (SSTO) technology. The idea is to reduce SSTO technical
uncertainty (and thus development risk and cost) while at the same time
increasing investor confidence, to the point where SSTO will make sense
as a private commercial investment. We're not far from that point.
With luck and hard work, we should see fully-reusable rocket testbeds
flying into space before the end of this decade, with practical
radically cheaper orbital transports following soon after.
Space Access Society won't accept donations from government launch
developers or contractors - it would limit our freedom to do what's
needed. We survive on member dues and contributions, plus what we make
selling tapes and running our annual conference.
Join us, and help us make it happen.
Henry Vanderbilt, Executive Director, Space Access Society
To join Space Access Society or buy the SSTO/DC-X V 3.1 video we have
for sale (Two hours, includes all twelve DC-X/XA flights, X-33 bidder
animations, X-33, DC-X and SSTO backgrounders, aerospike engine test-
stand footage, plus White Sands Missile Range DC-X ops site footage)
mail a check to: SAS, 4855 E Warner Rd #24-150, Phoenix AZ 85044. SAS
membership with direct email of Space Access Updates is $30 US per year;
the SSTO V 3.0 video is $25, $5 off for SAS members, $8 extra for
shipping outside North America, US standard VHS NTSC only.
__________________________________________________________________________
Space Access Society "Reach low orbit and you're halfway to anywhere
4855 E Warner Rd #24-150 in the Solar System."
Phoenix AZ 85044 - Robert A. Heinlein
602 431-9283 voice/fax
www.space-access.org "You can't get there from here."
space.access@space-access.org - Anonymous
- Permission granted to redistribute the full and unaltered text of this -
- piece, including the copyright and this notice. All other rights -
- reserved. In other words, crossposting, emailing, or printing this -
- whole and passing it on to interested parties is strongly encouraged. -