Transterrestrial Musings

Comments

For one, the computer power needed to carry out rocket and engine fabrication, including computational fluid design, is now affordable for small firms.

This is a very interesting point, the first really plausible reason I've heard for why small newspace firms might really be able to bring down the cost of orbital flight substantially. I know, I know, there's the theory that NASA is run by troglodytes who spend 500 times more than necessary for every screw. But I've never really believed that, much. I'll give the bureaucrats a factor of 2-5, which is shameful enough, but swallowing a doofus factor of 100 or more has always been tough for me.

But if it's true that you can drastically reduce the cost of design if you can model stuff on the computer accurately, then I can believe there's a revolution coming, since without doubt you're starting to be able to do on $1000 hardware what twenty years ago you could only do on $1,000,000 hardware serviced by a team of $50,000 a year men.

Posted by Carl Pham at April 23, 2007 11:33 PM

if you can model stuff on the computer accurately
that completely does not explain how Armadillo has gotten through countless iterations of totally different rocket hardware for $3M so far. they probably have most varied current experience ATM with different types of rockets, and AFAIK they dont do much simulation or modelling.
DC-X used to have similar capabilities that Armadillo's Pixel has now, and it cost at least a factor of ten more ( and that was considered cheap and innovative by government standards at the time ) and i dont know how easily could the team rebuild and improve the design, as it was never tried.

Posted by kert at April 24, 2007 12:03 AM

XCOR, Armadillo et al are the best hope for an "Accelerando" into the solar system in the next fifty years, but there are so few of those firms and so many pitfalls between here and there that their potential is almost painful to see. SpaceX and Bigelow are just intermediates between government/contractor space and what these garage firms represent, and that is achingly exciting. Developing mechanical hardware like software would revolutionize many other sectors of the economy, not just space, if it were to succeed. On the other hand, the concept of a "patch" might become a little more dramatic.

Posted by Brian Swiderski at April 24, 2007 02:04 AM

Orbit's just 10 years away! Great news! Hey, so's breakeven fusion last time I heard (pick any date last 40 years) In fact, if I remember right, Rotary Rocket was just 10 years away from orbit in 1997. Wow, that means orbit may happen this year!

Hate to be the pessimist, but by nature I'm a bit of a realist. Rich Pournelle is describing progress on the very bottom of the capability curve and prognosticating on the timeline of the very top. He doesn't have the data to back it up, and I'll place a 100 dollars on it today if there are any honest takers that XCOR will not be in orbit in 2017.

Posted by tom at April 24, 2007 06:02 AM

"I'll place a 100 dollars on it today if there are any honest takers that XCOR will not be in orbit in 2017."

That's what you call a win-win bet.

Posted by Brian Swiderski at April 24, 2007 06:42 AM

...swallowing a doofus factor of 100 or more has always been tough for me.

It was never a "doofus" factor. There are many factors, none of which have anything to do with the competence of NASA employees. And they do in fact mean orders of magnitude higher cost.

Posted by Rand Simberg at April 24, 2007 06:43 AM

Armadillo has gotten through countless iterations of totally different rocket hardware for $3M so far. they probably have most varied current experience ATM with different types of rockets, and AFAIK they dont do much simulation or modelling.

They haven't had to do anything complex enough (e.g., aerodynamics) to require modeling. As for DC-X, it was LOX/hydrogen. That alone is probably sufficient to account for a huge cost differential.

Posted by Rand Simberg at April 24, 2007 06:45 AM

Brian's quick summary of the nascent private space companies was right on the money. Did someone hijack his account?

Posted by philw at April 24, 2007 07:26 AM

philw:
Same author, same reality.

Posted by Brian Swiderski at April 24, 2007 07:29 AM

As for DC-X, it was LOX/hydrogen. That alone is probably sufficient to account for a huge cost differential.
I would submit that for the capability that it had, LOX/hydrogen was clearly overkill and thus a poor choice.
Of course, i know that it was intended to evolve far beyond where it ended up, but nevertheless in retrospect you can make assessment that they could easily have deferred going the complex ( H2 ) route until its performance was needed, and it may just have saved the project a bunch of money and they could have afforded a second test article.
Or maybe not.
I am not trying to diss DC-X, im just pointing out that for practically equal delivered capability, they spent an order of magnitude more. What was under the hood is somewhat beside the point.

Armadillo will very likely go much faster and higher than DC-X ever did, largely influenced by the fact that they did _not_ choose the expensive route from the get go.

Posted by kert at April 24, 2007 08:27 AM

He doesn't have the data to back it up, and I'll place a 100 dollars on it today if there are any honest takers that XCOR will not be in orbit in 2017.
Would you want to take the same bet, but include a few other names as a OR clause in it ?
I see difficult scaling problems in HTHL approach, and thats what they are banking on. I'd willing to put my money ( and i actually have, in a way ) on vertical takeoff approaches.
So if you are willing to expand to a few more current suborbital startups, i would be a honest taker of that bet.

Posted by kert at April 24, 2007 08:36 AM

I would submit that for the capability that it had, LOX/hydrogen was clearly overkill and thus a poor choice.

That's because you apparently don't understand what the program objectives were. It was a technology demonstrator, and it was supposed to demonstrate a fast turnaround of a LOX/hydrogen vehicle.

philw, historically, Brian seems to be sane when (and apparently only when) posting on space-related topics.

Posted by Rand Simberg at April 24, 2007 08:55 AM

it was supposed to demonstrate a fast turnaround of a LOX/hydrogen vehicle
I personally wouldnt know, but all reference materials i have consulted say that it was it was supposed to demonstrate a fast turnaround of a rocket vehicle, not specifically LOX/hydrogen. Actually, more texts say "fast turnaround spacecraft" which it never became, of course.
Some places like Mark Wade's Astronautix site say simply "demonstrate VTVL"

If you have more precise info, could you point to reference documents ? I would gladly help correct 0nline materials spreading misinformation.

Posted by kert at April 24, 2007 10:16 AM

Someone pointed out that there has been over a 20 year history of failed start-ups in commercializing cheap orbital access. This is certainly true. However, there are significanly more space start-ups now than there has ever been in the past. If there are 10-12 of these companies, it is likely that 1 or 2 of them will make it.

There have also been significant changes in the underlying conditions necessary for the design and manufacture of space vehicles.

The reduction of costs associated with computer aided engineering is quite profound. Much of the costs associated with the development of new aircraft (and space vehicles) is in the engineering and design. Computers now make this cheap, cheap, cheap.

When I was in college (mid 80's), Boeing had the only Cray supercomputer in the pacific northwest. I have no idea how present day laptops and PCs compare to the mid 80's Cray, but the difference cannot be that significant.

The other part with manufacturing space vehicles is the metal-bending. Yes, this is fairly technical with precision machined components, sensors, and controls. However, I doubt this would be that much more difficult or complex than manufacturing high-vacuum process equipment (CVD, PVD, and the like) and there are lots of small 10-20 person companies that make this stuff (I used to work for one in Japan).

Also, the machine tool industry is one of the first industries to get going in a developing country (Panang, in Malaysia, has over 20 machine shops that make tools of reasonable quality). This is an industry that the East Asian economies are exceling in. There are countless manufacturers of machine tools just like there are countless manufacturers of speciality chemicals and about any other man-made product in China. Check out Alibaba.com or globalsources.com and you get the idea.

Eclipse Aviation is an example of a start-up that is utilizing these developments of cost reduction to revolutionize aircraft manufacturing. Companies like XCOR and the like will do the same for space.

I am guardedly optimistic about space stuff for the first time in my life (since I joined L-5 as a kid).

Posted by Kurt9 at April 24, 2007 10:18 AM

Straight from the horses:
http://www.hq.nasa.gov/office/pao/History/x-33/dcx-facts.htm

The Ballistic Missile Defense Organization's Single Stage Rocket Technology (SSRT) program is chartered to demonstrate the practicality, reliability, operability and cost efficiency of a fully reusable, rapid turnaround single stage rocket, with the ultimate goal of aircraft-like operations of reusable launch vehicles (RLVs).

The program is focused on using existing technologies and systems to demonstrate the feasibility of building both suborbital and orbital RLVs which are able to fly into space, return to the launch site, and be serviced and ready for the next mission within three days. Such a suborbital RLV could potentially support many of BMDO's planned suborbital system tests and experiments.

Not a peep about hydrogen. Where did i go wrong ?

Posted by kert at April 24, 2007 10:21 AM

In fact, im reading that switching the fuel combination from H2/LOX to something that would require less expensive ground infrastructure support was considered under the very DC-X program to improve on the real targetted benchmarks ( low cost and fast turnaround )
So Rand would you care to set the story straight ?

Posted by kert at April 24, 2007 10:27 AM

Rand: "Brian seems to be sane when (and apparently only when) posting on space-related topics."

Only your emotions, not my reasoning ability, are what change across subjects. I treat all of them as problem-solving challenges, enjoy considering exotic ideas, and am disgusted when people turn off their minds and resort to ideological self-gratification.

Why do you support space, Rand? I do because I love humanity, and want it to thrive and evolve without limit; because I love life, and want it to thrive and evolve without limit; and because I want to experience, and know that others will experience, the endless beauty of this universe. It is not political or emotional escapism for me, nor a masochistic desire to seek hardship, nor a blank canvas on which to draw utopias, but simply the pursuit of something truly amazing--to participate in the birth of the new; to crest a hill within your tiny valley and know vistas never imagined.

Yet even staring into the infinite, at the promise of eons of continuously branching humanity and history, I don't lose sight of what comprises it: The lives of individual people, living day to day in communities of some kind, and facing pressures and problems reflecting or expounded from the basic nature of our species and past events. I love individual human beings and cultures just as much as humankind as a whole, and would see them flourish if I can.

Posted by Brian Swiderski at April 24, 2007 10:29 AM

DC-X was premised on the original SSX (Space Ship Experimental) concept put forth by Max Hunter and Jerry Pournelle to Dan Quayle. The idea was to demonstrate technologies for SSTO, and it was assumed that such a vehicle would be LOX/hydrogen.

I don't necessarily agree that this is a good assumption, but it was the assumption. Had they used any other kind of propellants, critics would have complained that they didn't prove out the necessary technologies. Which, of course, they did anyway.

Posted by Rand Simberg at April 24, 2007 10:41 AM

Beautifully put Brian. I agree.

Posted by Lee Valentine at April 24, 2007 11:49 AM

and it was assumed that such a vehicle would be LOX/hydrogen
Aha, now we are getting somewhere. Hidden assumptions, not clearly stated as goals of the program. Note where our discussion wandered off: why did a government equivalent program deliver equal capability to a bunch of voluteer garage builders for at least tenfold increase in costs ?

Hidden assumptions.
Like, if ESAS architecture is judged purely on technical and cost merits, it does not stand a second of critique. Now if you take the hidden assumptions into account ( the need to keep the standing armies into account ), it sorta starts to make sense ...

Posted by kert at April 24, 2007 12:04 PM

I once asked the DC-X Deputy PM why they opted for
LOX-LH2, and the answer was:

"Once we had selected for VTVL, that pushed to deep throttle
on the primary propulsion system. Given the tight schedule,
an existing engine with minimum mods was needed, the
only engine in the thrust range still in production was
the Rl-10. P&W was willing to mod the nozzle for
sea level operations, and that was how we got to LH2.
While it was good we had atracability to LH for orbit,
all of that was second order behind a need for an existing
throttled engine".

Posted by anonymous at April 24, 2007 12:41 PM

so, if it would have been a op funded by private money, the long-term goals for low operations cost would have steered them away from LH to something less arcane, like it does with current suborbital builders ( Masten, Armadillo, TGV, Blue Origin, there are others )
Being a government program, near-term schedule pressure, i.e. vision being limited only to program milestones, cost be damned, dictated poor propulsion choice.

Does that sum the actual reasoning behind it up ?

Posted by kert at April 24, 2007 01:06 PM

Hidden assumptions, not clearly stated as goals of the program.

I don't know whether they were "hidden" or not. Go read the original RFP. Until you do so, stop complaining about it.

Posted by Rand Simberg at April 24, 2007 01:17 PM

kert

well in the case of the DC-X the Hydrogen decision was
essentially a compromise between cost, schedule and operations.

The DC-X was not meant for a long operational life, merely,
40-50 test flights and then if there was some luck 5-6
technology flights per year after that.

They had a pretty tough design requirement for deep throttle
and none of the storables or non-toxic engines had that
kind of throttle range.

in terms of deep throttle, you have the XLR-99
engine out of the X-15 which throttled to 40% kind of.

YOu had the RL-10 which throttles to 25% and you had
the LEM Descent engine which throttles to 10%.

now the LEMD was an ablator engine with really nasty fuels

the XLR-99 uses ammonia and has some real issues at low power.

The RL-10 was in production and did a decent throttle job.

There were two choices: Design new engines, which would
have been some serious coin and time or reconfigure
either the XLR-99 or RL-10 to new fuels.

Neither choice was palatable.

what do you suggest, and what do you think it would cost?

Posted by at April 25, 2007 04:03 PM

The fact that Armadillo only spent $3M is less interesting that the fact Armadillo jumped around so much.

Big organizations -- government and private -- have huge inertia. Getting them moving is difficult; changing their direction is equally difficult. When mistakes are made and committed to, reversing course can be nearly impossible. Decisions have to be coordinated with ever-expanding sets of stakeholders. Since rework is so difficult, there's a misguided attempt to get everything worked out first, and analysis paralysis replaces a 'just do it' philosophy.

Posted by Paul Dietz at April 26, 2007 09:31 AM

The manufacturing guys over at Evolving Excellence are always talking about lean in small companies. Amazing how even a t-shirt manufacturer in LA paying above minimum wage can outcompete an Asian sweatshop when applying lean.

http://www.evolvingexcellence.com

Posted by Ken Tolbert at May 20, 2007 12:22 AM

Post a comment