HOW TO GET TO SPACE

 

 

Jerry E. Pournelle, Ph.D.

 

 

Whose Job Is It?

 

We have institutions for nearly everything now, but we have lost the institutional means of long range planning. Who looks after our grandchildren? Surely not the politicians, who can hardly think past the next election, and are generally content to put problems off to be solved when the politician is safely out of office.

 

We no longer have great families planning for the future because we don’t have continuity of wealth: for good or ill, there is no entailment of property, and while there is talk of abolishing death taxes, they have pretty well done their egalitarian work; and perhaps this is a good thing. Good or ill, few think about the next generation and fewer try to do anything about it, because they have little confidence of continuity of their positions and fortunes.

 

The corporations are useless. The discounted value of a dollar in thirty years is effectively zero. A corporation that invested significantly in research with a thirty-year payoff would be involved in a hostile takeover not long afterwards.

 

That leaves government.

 

But Government Always Mucks Things Up

 

It isn’t strictly true that government always mucks things up, but it’s often enough so. A, if not the, major purpose of government is to extract money from non-government and use it to hire and pay government employees. This guarantees that government will always expand; and there inevitably comes a point at which the addition of people to a project has a negative impact. Almost all long-standing government agencies and projects have people who are worse than merely useless, they are in the way; and the more conscientious they are about earning their pay the more they tend to get in the gears and bring progress to a halt.

It has certainly been that way in space R&;D and space operations. I needn’t give the history here. It can be summed up as early enthusiasm and success, expansion of the team, and then makework to keep the standing army busy or appearing to be busy. [1] New centers are then started in hopes of having a new cycle of enthusiasm and success, and that works for a while, but eventually the system is clogged with memoranda, policy directives, reviews, reviews of reviews, and the rest. Everyone is familiar with the process. Nothing can be done because nothing can be approved, and nothing can be approved because the only way to justify one’s importance is to ponder approvals, rewrite requirements statements, and require rewrites. And always to add complexities and more mission requirements.

 

Arsenals and research labs have accomplished miracles, but generally only in a crisis, and generally not for long. Soon they become part of the problem set rather than the solution set. Then they block anything not invented there, only they never invent anything. [2]

 

If government is the only long range institution and it tends to do more harm than good, where are we?

 

 

X Programs: An Exception to the Rule

 

 

One of the most spectacularly successful government projects in technological history was the X-program. Beginning with the X-1 and going on through the X-15 and beyond, X programs were instrumental – indeed essential – in making the United States far and away the technological leader in aviation and space technology. One example should suffice: the X-3 Stiletto, the first airplane to take off from the ground, go supersonic, and land to be flown again. This airplane was useless for operations. It could hardly maneuver. However, the Stiletto led directly to the Lockheed F-104 Starfighter, and that airplane dominated military airspace for a very long time.

There were other such successes for the X programs; indeed, the program was cancelled not because of its failure, but due to its success. It generated so much new military technology that the arms control strategists were afraid of it. Whether or not arms control was a good strategy, one thing was clear, the X programs were not compatible with it; they had to go and they did. [3]

 

What Are X Programs?

 

The X programs were so successful that they have generated a kind of mystique, and recently there have been attempts to capture that residual good will. Unfortunately, the two most spectacular, X-33 and X-34, are not, never were, and never were intended to be actual X programs; they took the name but nothing else.

Real X programs have certain characteristics, and while they are hardly all identical, we can tease out some common factors.

First, they are relatively small. They don’t have huge budgets. They don’t attract big attention and they are not good bait for large bureaucratic sharks or big companies. They are not profitable in the usual sense of the word. The payoff to a big company from participation in an X project is nearly all intangible: prestige at least within the profession, and some early technological advantages from having employed people who built and understood the new technologies identified or acquired by the project. Actual monetary gains are small and often negative.

 

Indeed, the payoff to everyone: sponsoring institution, contractors, the United States, and sometimes all mankind, is technology and experience and not much else. X-programs do not result in operational vehicles, and X-craft do not fly missions. Most of them can’t. The X-airplanes had no payloads unless you call a stick of Beeman’s chewing gum a payload. [4]

 

Next, X-programs have limited goals, and are over and done with in a relatively short time. They are not intended for career building, and ideally there should not be any career opportunities in X-program management. One does the project and gets out. It’s over. There’s no empire to be built because the project doesn’t last long enough to allow that. This is key to X-success.

 

Finally, the best X-projects are based in out of the way places, generally unpleasant places. It is no coincidence that some of the best X projects were done at Edwards. No one wants to build empires in the Mojave Desert. Edwards is far enough from Washington, or even Los Angeles, that there is some autonomy. It is also dull enough to encourage work, if only in the hopes of getting the project over with so you can return to civilization. I note that making life in Palmdale somewhat more pleasant than the Spartan existence of Edwards in the days of the X programs is probably a step backward.

 

The typical X project focused on a needed technology. Although the technique is applicable to many areas of technology, I’ll stay with aerospace for the moment. One designs the best ship possible given existing technology. There are to be few to no stretches or reaches: we are not looking for new technology, we are looking to see what the best we have can do – and thereby identify what’s needed next.

 

The ship is built. Typically there will be three vehicles (tail numbers in the jargon). The first is flown to find out its capabilities. Then those limits are tested, and tested again. Frequently tail number One is destroyed in the test process, although that’s not inevitable. Using what was learned from One, Two is modified and flown to its limits, and kept flying until there is no more to learn. Number Three makes a few token flights and goes to the Smithsonian.

 

From that process we learn what we can do, and what more we need to know.

 

Can’t Computers Do That?

 

 The usual objection to X projects is that they are costly. “It’s silly in this day and age to design ships and fly them until the wings come off. Do it right in the first place. Do your tests with computer simulations.”

The simplest answer to that is the X-33 fiasco. This was supposed to give us a great leap forward to reusable spaceships. There would be a series of expensive ground tests to provide input for computer simulations, simulated tests of designs, and the smooth design and flight of a new ship.

It hasn’t worked out that way, and it won’t, and when it is done we will not have learned much. [5]

 The truth is that computers are great for some functions, but not very good for testing experimental design concepts. How could they be? Extrapolation is never as accurate as interpolation, and we all know it.

Let me give a simple example. A rocket must be about 90% fuel or it won’t get to orbit. This is obvious from the rocket equation. [6] Multi-stage rockets have even more extreme mass ratios.

Of the 10% that’s left, some will be structure – tankage and aeroshell, control surfaces, heat shielding, etc. – and necessary avionics and power. Indeed, this will be about 90% of what’s left: meaning that the payload, if any, will be no more than 1% of the Gross LiftOff Weight (GLOW). For a 600,000 pound ship this translates as 6,000 or fewer pounds of payload. How much? It makes a difference whether 6 pounds or 6,000 pounds gets to orbit. But of course we don’t know, and our computers can’t tell us. The true payload of that ship is lost in the noise around the third decimal place, and no computer model I know of will find it.

We don’t have inputs accurate to the second decimal point; how can we have any confidence in computer generated three decimal answers? And clearly we cannot.

 

There comes a time when you must fly something if you hope to learn more. We need data for those simulations. We are at that stage now in a critical area of space technology.

 

What Do We Need?

 

I’ve done this in reverse order. Having told you how to get there, I’m going to talk about where we need to go.

 

First and most obvious, the real cost of space is access. We over-build satellites because they have to last a long time, and they have to last a long time because it costs too much to replace them. Technology outstrips us every time: by the time a satellite or other spacecraft is launched it is generally obsolete, sometimes hopelessly so, as Moore’s Law takes another run along the exponential curve.

Ideally we would build satellites to last for a few years, then replace them with smaller, more powerful, and cheaper spacecraft; but we don’t do that. Instead we lock in electronics that are outdated before the design is completed, and design those systems to last a long time. This only makes sense if the launch costs are enormous. But of course they are.

 

The real key to space exploration is developing cheap and reliable ways of getting there. It’s pointless to speculate on whether we need humans in space when the costs are so high. My own view is that if access to the Moon cost anything like what access to the South Pole costs, we’d be there like a shot: we certainly have not been content to explore Antarctica with robots. But until the costs come a long way down, we have to be content with robots, and getting them to orbit and beyond is all but prohibitively expensive.

 

Sir Arthur C. Clarke said that if the human race is to survive, then for all but a very brief period of our history the word ship will mean space ship. This generation is the first to have the capability of developing true space ships. This is the time to do it, and we know how.

 

If we want to learn more about using space and the resources beyond our planet, we must build more space ships and fly more space ships.

 

What’s a Space Ship?

 

A ship isn’t a missile. It’s not ammunition. A ship is a vehicle. It goes places, gets there, comes back, is refueled, and goes again. We don’t throw the ship away when it has reached its destination.

 

The obvious way to go to space is the way Buck Rogers and Flash Gordon did it: fly to space, bop around for a while, then land and refuel. Everyone knows this, the trick is how to do it, but let’s look at why.

 

Airline operations cost a small multiple (typically 3 to 5) of fuel costs. It costs about the same in fuel costs to fly a pound to Sydney, Australia from the United States as it does to put that pound in orbit. It shouldn’t cost more in dollars, either. Of course the airline doesn’t roll the airplane into the sea after it gets to Sydney.

 

Although fuel cost drives airline operations costs, there’s another driver for space ops.  A typical airline will have about 110 employees per airplane, but half of those sell tickets. Sixty in operations and maintenance is more likely.  Now divide the number of NASA civil servants and contractor employees working on Shuttle by the number of Shuttles, and you will get some idea of why Shuttle flights are about a billion dollars a flight instead of the hundred million or so an airliner would charge. [7] Understand I am not accusing NASA of featherbedding Shuttle: all those people are needed, but that’s just the problem. Shuttle was in some respects designed to need them. When your goal is to employ a standing army, you will reach that goal one way or another.

 

 

Shuttle is rebuildable, not reusable, and that’s not what we need. A properly designed space ship will seek to minimize operations costs, not maximize performance. We have yet to build rockets whose design was driven by operations, and it’s time to do it.

 

Operations driven designs are the key to our future in space.

 

A Development Plan

 

There are probably a dozen paths to success, but let me lay out one of them.

Begin with a concept: we want a simple, reusable, spaceship design. We’re far more concerned with keeping total mission costs down than with optimizing performance. We’re looking to build a ship, not ammunition, and we want to keep things as simple as possible.

 

We also want a program to develop that ship that ensures we learn something at each step of the way: we don’t want an all-or-none concept. We want incremental testing, so that we can feel our way to the edge of the envelope. We also want multiple tail numbers so that if we get past the edge of the envelope we have a ship ready to take advantage of what we learned when we pranged the first one.

 

SSX

 

That brings us to what I call the SSX. We sold this concept to the National Space Council in 1989, but alas it was only partially implemented before X-33 came to dominate space development.

 

SSX is a Vertical Takeoff Vertical Landing Single Stage to Orbit rocket with multiple rocket engines. I used to think the right way to go was hydrogen-LOX and expander engines, but I’m now inclined to agree with Max Hunter that methane or propane with LOX presents far fewer operational problems at affordable decrements in performance.

 

The important thing about SSX is that it can be flown incrementally: don’t put much fuel in it and fly it a little way, then land it, see what happened, and do it again. This feature alone should dictate SSTO and VTVL for the first experimental reusable spaceship. Those may not be the configurations of operational spaceships, but they are the proper ones for today’s X projects. Get SSX accomplished and we’ll know a very great deal about reusable spaceship operations.

 

We started the SSX program with DC-X which flew at White Sands so long as the Air Force was in charge, but oddly enough, this potential rival to Shuttle toppled and burned the first time NASA was in charge. Pure coincidence, of course. But note that even the last flight was successful, and the only reason the ship burned was that a human mechanic hadn’t connected up a hydraulic line, so the landing gear didn’t deploy properly.

 

DC-X was a small scale model of the SSX we had proposed to the National Space Council in 1989.  DC-X demonstrated the feasibility of vertical landings. It showed that contrary to computer simulations, there was sufficient control authority at low speeds. It also showed that you can trick an F-15 flight control program into believing it is in charge of the weirdest airplane it ever conceived of…

 

The next step is a full SSX-1: around 600,000 pounds GLOW, meaning an empty weight of 60,000 pounds, well within the capabilities of off the shelf ground handling equipment.

 

Payload

 

SSX-1 will probably have a negative payload to orbit: that is, it won’t get there. We will and should overbuild it. Flight safety and recoverability are more important than shaving the mass fraction.

 

Remember.  X programs have no payloads, nor do they have mission requirements. The purpose of an X program is to learn how to build ships that do have payloads and can fulfill mission requirements.

 

Payoff

 

There will be many payoffs from an SSX-1. First, when it’s done we’ll have numbers to input: we’ll have a much better idea of what the mass fraction, and thus the payload, of an SSTO VTOL ship will be. We will know what size the ship must be to make orbit. Perhaps we need much better engines, or to go to some kind of Two Stage to Orbit (with recoverable first or zero stage). But we will know these things, not merely guess at them.

 

More likely we will find the stress points, bore holes in the structure to remove needless strength and weight, and come close to orbit with this ship. My guess is that we may never make orbit but we can scare it to death.

 

We’ll learn a lot about routine spacecraft operations, and that’s important: space ships are not wet navy ships, nor are they airplanes. They’re more like airplanes than ammunition, though, and we need ground crews used to that concept. We’ll learn a lot about flying to different azimuths. We’ll learn a lot about engine performance and altitude compensation and the real truth about engine layout. Because we don’t have to make orbit or come close to fly the ship and bring it home we can experiment with different arrangements and tricks: extensible bells, dynamic aerospike geometries, and so forth.

 

And finally, we’ll relearn what X projects are all about, and that will be the best payoff of all.