February 07, 2003
Space Travel Will Be Enabled By Non-Space Technologies
Rand Simberg argues that we need low cost reusable space launch vehicles.
We need to recognize that we have a chicken and egg problem. We will only get low costs and reliability with high activity levels, and we will only get high activity levels with vehicles designed to sustain them, at low cost (and that means not throwing them away).
In the comments section of that post Michael Mealling argues that only a business approach to space will make space development happen.
IMHO, there are two methods: 1) we all build businesses unrelated to space and create enough wealth among us that we can pay to have that value network built for us (there is imperical evidence that this works) 2) we figure out disruptive technologies/products/business methods that change the underlying assumptions about space and its relationship to people on the planet. The first one is tractable and relatively easy. The second is much more fun and potentially paradigm changing but extremely hard.
Let me argue a different viewpoint: The vast bulk of the technologies that will eventually enable significant human movement into space will come from outside the aerospace industry and will not come from people whose motive it is to develop technologies that will enable the development of cheap safe spaceflight. The US Department of Defense will have an FY2004 budget of around $379 billion dollars. In spite of this the DOD increasingly looks for ways to more rapidly incorporate civilian technologies into military weapons systems. NASA, with a budget of only $15 billion dollars (little of which goes to the development of new space launch technology) is even more in the position of user of the best new private sector technologies (and then only when it gets around to designing something new).
NASA has been locked for years into supporting the continued use of old technologies to produce sentimentally appealing human space missions in the short term. Whether the fault for this lies in NASA or Congress or Presidents or the American people is really besides the point. Because of the continued inability of NASA to focus on long term technological development the technological advances that will some day enable the economic development and colonization of space will not come from NASA funding.
It makes sense for NASA to abandon the Space Shuttle and ISS in order to focus on new technology development. But my own prediction is that the only way that is going to happen in the short term is if the loss of the Columbia is found to have been due to a design flaw in the Space Shuttle that can't easily be fixed. NASA and Congress are too committed to the Space Shuttle and ISS. The film clips the Shuttle missions create are seen as glorious in the minds of too much of the public. Political leaders are not at all eager to educate the public to see the Shuttle and ISS as big mistakes (after all, who made those mistakes?). Nor are they going to tell the public that the deaths of the astronauts who die on Shuttle flights do not contribute to the advance of our ability to move out into space (even though that is obviously the case). In the face of the widespread belief in myths about what our current human space flight program accomplishes it seems unlikely that NASA will be ordered to abandon the Shuttle. It seems even more unlikely that NASA will instead be assigned as its the top priority the development of new space-enabling technologies.
Given that NASA is unlikely to become more effective and that other national space programs are less well-funded and even less ambitious where does that leave the future of manned space travel? We need to make very large technological strides in order to get out of our current rut of high costs and low safety and reliability for human space launch. But until future Shuttle losses eventually end the Space Shuttle program by attrition NASA is not going to put much effort into radical technological advances. Even when NASA gets around to developing a new type of shuttle it will do so in such a hurry to meet an immediate need (yet another Shuttle loss being the most likely proximate cause) that the new design will just incorporate the best technologies available at that point. Therefore NASA will not try out many experimental design concepts as prototypes and will instead opt to pursue a fairly conservative design utilising existing knowledge.
Luckily there is a silver lining in this pessimistic story. The overall rate of scientific and technological advance is accelerating. While Moore's Law may slow down the rate of increase in processor speeds the rate of advance in computer microprocessors (eventually using quantum computing or biomolecular computing) will still produce computers that are orders of magnitude faster in the next few decades. Also, fiber optics and mass storage will continue their own rapid rates of advance. All of these technologies along with advances in mathematical algorithms for simulating designs and physical phenomena will combine to provide better computer aided design and engineering tools. Therefore future spacecraft development efforts will be able to produce much more optimized designs.
General physics, chemistry, and biology continue to advance. Advances in materials science and nanotechnology will provide many new materials and fabrication techniques for use in space launcher design. New types of structural and sensor materials will enable the implementation of spacecraft whose performance greatly exceed the best spacecraft that could be built today. Computer advances combined with sensor advances to make new kinds of control systems will enable the creation of designs that would otherwise not be possible.
The development of a significant human presence in space could in theory be accelerated by a focused attempt to develop enabling technologies specific to spaceflight. Before the advent of computers with sufficient throughput to simulate the performance of advanced supersonic ramjet designs and other advanced design approaches it would be possible to develop many prototype concepts and to try many prototype materials in prototype experimental spacecraft. Such an effort, while risky, might produce a much better design. But the political environment argues strongly against that the pursuit of such a high-risk high-payoff approach. Instead, advances in space launch technology will have to await the creation of a large range of enabling technologies which will originally be developed for other purposes.
Space enthusiasts who do not like this prognosis do have one option: promote arguments to the general public and to opinion leaders about the benefits of pursuing a more radical path for the development of space technologies. A reasonable component of such an argument would be to advocate the split of NASA to put its scientific space studies work (i.e. studying planets, asteroids, stars and all other stuff up there) into an agency dedicated to that purpose. Then another agency should be dedicated to the development of science and prototype technologies focused on lower cost launchers and human space travel.
While I agree with the criticisms of the shuttle program, I must disagree with the assertion that future space missions will result from research and development not targeted at space.
One only has to look at the semiconductor industry to see how it requires a combination of advances in physics and engineering with continuous experience to make progress. The most closely held secrets in that industry aren't the advances, they are the mundane details of the process technology.
In the same way, one only gets into space by building stuff, seeing what's wrong with it, and building some more. There are too many variables for computer simulation to take on the entire chore of practice and experience.
Furthermore, there are many issues related to space travel which just don't have much importance in other areas. For example, the materials engineering needed to make hypersonic craft is not needed for much else. Likewise the development of large rockets and the myriad of specialized systems therein.
If NASA doesn't do the job, and the market is not sufficient for private enterprise, only the DOD is likely to be working on this. Of course, it was DOD work that lead to the creation of the integrated circuit, and lots of other innovations.
The other place innovation may come from is China and perhaps Japan. China has an active space program and may, for nationalistic and military reasons make a lot of progress. However, it would be better if a free country was also making that progress and developing the technological infrastructure.
John, The materials engineering needed to make hypersonic ramjets will be easier to do once nanotech assemblers are perfected for other applications. While the exact material needed for some ramjet aircraft will be different the techniques for arranging atoms to construct it with a sufficient degree of control will be developed for commercial needs to make many different kinds of ground-based devices.
A broad increase in our ability to simulate designs, simulate physical processes, and to control the deposition of material at the atomic level are all being developed for other reasons but will eventually be used to make spacecraft that are much cheaper and more reliable.
Randall, I am not contending that continued non-space progress will aid in space work - and in some cases may provide break-through technologies. But these are just enabling technologies... turning them into the aircraft requires trying and doing.
I don't expect a broad increase in our ability to simulate, because we already have a fantastic ability to do so. Additional Moore's law increases in computer power are nice, but I don't think they will *qualitatively* effect these efforts. Furthermore, computer simulation always requires validation. All simulation that I am aware of uses "parameterization" to adjust for either limits in the simulation resolution or limits in the input knowledge. The former limits are of course reduced by improved computing power, but the latter are not.
For an example of what computer simulation yields when you cannot test the results, just look at the global climate models that predict many varieties of global warming. They can't be tested because the only test is to wait decades and see what happens.
So I argue that as a practical matter, you can only go so far with models. While *in theory* you can model almost anything you care about, in practice you need feedback from the real world.
John, There are lots of physical processes for which we lack the computational thru-put to simulate. Protein folding is an example of one. Protein folding simulations are being done now. But they take a long time. The codes for doing scramjet modelling is another. Again, you can find plenty of people writing models and testing the models. But the models are too simplified and still take too long to run to allow a real design effort that would try many different designs. I've read hardcopy articles on this in the past but can't find any good ones on the web at the moment on comparative size of various classical modelling problems. However, I'm going to keep looking and probably do a post on it within a week or two.
I definitely agree that most technological developements that space industrialization is going to take advantage of are going to be from outside the industry. The fact that this is so proves my point mentioned earlier in the article: the problems in the space industry are business related, not technological. The reason technology will come from outside the industry is that there are real business reasons that the new technology was needed in that other, external industry.
As I've mentioned on Rand Simberg's sight several times, the technologies needed to get into space cheaply were developed in the 60s. Sure, with our new experience with materials we can make them better still, but that isn't a requirement for cheap access to space. What's needed are compelling products and services that people are willing to spend money on. Solve that problem and the technology (if needed) will follow.
I would suggest that folks who are really interested in solving this problem start learning distruptive business practices instead of engineering. Sure, its a technology based industry, but its still all about business.
There are lots of space related products and services that lots of people are already willing to spend lots of money on. When it comes right down to it, WTBS, HBO and long distance telephony are space related services.
I am not exactly sure who pays for GPS, but the service is very cheap for many users.
The question remains: What is the business case for direct human occupation of space or other celestial bodies?
For the most part, I think either staying on Earth has to get a lot more expensive or going into space has to get a lot less expensive before we will have a justifiable business case for any human presence in space.
"For the most part, I think either staying on Earth has to get a lot more expensive or going into space has to get a lot less expensive before we will have a justifiable business case for any human presence in space."
Which misses the point that it's a chicken-egg problem. It will not become less expensive until we start doing a lot more of it.
A key point about chicken-egg problems is we do not know which came first. (Well, I have an answer for that, but it's not suitable for a family friendly website.)
In another thread, I used the analogy of Europe trying to colonize America by canoe. It wouldn't matter how many times Europeans set out in tiny groups by canoe to explore America, the cost of colonizing America by canoe would remain high. Safer vehicles with better propulsion and navigation were required to make the venture feasible. Colonization became unstoppable once a critical mass of enabling technologies were developed (ie. sextants, timepieces, sails, durable and stable hull designs, scurvy preventives etc.)
The specific technologies were not developed for transatlantic colonization; even though, they were required to make transatlantic colonization an economic reality. Instead, they were developed for more pedestrian uses--facilitating trade and naval expeditions along existing routes for instance.
There are lots of technologies we can develop to make delivery of existing space based products and services easier or cheaper. Scramjets are one such technology. I suspect the first supersonic combustion ramjets will be unmanned delivery vehicles (probably missiles actually). Perhaps, scramjets will be the critical piece that makes a permanent human presence outside our atmosphere unstoppable. Of course, that will probably depend on safety.
More likely, it will also require improved hull designs.
How reliable can we make reusable launch vehicles? How safe can we make them? Heck, someday we may abandon the whole reusable kick and decide it's easier to leave the metal up there to smelt into something different, but until then we need hulls we can safely return to ground.
Regardless, we need vehicles with better failure modes.
We also need faster turnaround times for vehicles. How long does the average shuttle spend on the ground between flights? I've seen airlines offload, groom, fill and pull back a jet in as little as 30 minutes. Even with another 30 minutes taxi-ing around the tarmac, that's only an hour between touchdown and takeoff.
Unless you want to build a launch vehicle capable of carrying thousands of people at a time, fast turnaround is necessary for colonization, which again means better reliability.