THE SSX CONCEPT
Jerry E. Pournelle, Ph.D.
The intuitively obvious way to get to orbit is to build a rocket ship that will go there, fly around in space, and return to Earth for refueling and reuse. It's what Buck Rogers did. This is so obviously the 'right way' that people have to be taught why we don't do it.
Explaining why we have not yet built that kind of ship, and why we can and ought to build one now, is going to take a little work. Not a lot. Fortunately the math is simple: there's only one actual equation in this paper. It's presented in a couple of forms, but it's still only one equation, and it's a pretty simple one. It's called the classical rocket equation, and studying it can teach us a lot about what has happened to the space program. You don't have to be a rocket scientist to follow the argument.
The classical rocket equation:
(1) M0/M1 = e to the exponent (v/c)
Where M0 is the initial mass of the rocket; M1 is the mass of the rocket at velocity v; c is the exhaust velocity, which is to say, the speed with which the propellant is ejected out the back end of the rocket in a direction opposite the line of flight; and e is the constant 2.71828... usually rounded to 2.7183.
The quantity M0/M1 is called the mass ratio, and is critical to the concept of working rockets.
In the above, we have assumed that the rocket started with an initial velocity of 0, and that it traveled through empty space unaffected by gravity. That formula can also be used to find a change in velocity of a rocket in empty space: substitute for v the quantity delta-v (or change in v, often written as delta-vee), where M0 is the mass at the initial velocity and M1 the mass at the final velocity.
That equation is often written as
(2) delta-vee = (Ve)*ln(M0/M1)
where Ve is the exhaust velocity; why the conventional representation changes from 'c' to Ve is historical and utterly unimportant, and if you write that as
(3) v = c times ln(M0/M1)
you'll see that it's really formula (1) rewritten. It used to be that you had to look up the natural logarithm (the log to the base e) of the Mass Ratio in a book of tables, but nowadays a TI-30 calculator costing $9.95 at a drug store gives it to you at the punch of a key.
A rocket taking off from earth would experience both atmospheric resistance and gravitational attraction. That doesn't change the fundamental usefulness of the rocket equation, but it does complicate calculating the result. You may think of atmospheric resistance and gravity as 'consumers' of velocity: that is, the final velocity of a real rocket will be lower than that calculated by the rocket equation.
Gravity consumes velocity. How much is a function of flight time. To a first approximation this will be about 25,000 feet per second. In practice, you design for the delta-vee required to get into orbit (about 5 miles/second or 25,000 feet per second), add an allowance for loss of delta-vee due to gravity, and if you can achieve that despite the atmospheric resistance, you've got your ship.
Atmospheric resistance harms us in two ways: not only does it provide physical resistance to the rocket's movement through the air (drag), but it also hinders the escape of the propellant out the back end: which is to say it slows that down, or lowers the c in the equation.
The faster the rocket goes the worse the drag. On the other hand, if you're headed for orbit, the faster the rocket, the less time spent in the atmosphere. Drag is a complex function of the size, shape, presented surface area, atmospheric density and temperature, winds, and flight profile. Artillery officers are well acquainted with this, since the effects can be quite significant on the trajectory of a round traveling five miles or more.
Drag has the effect of consuming about 1500 to 2000 feet per second of the rocket's velocity.
The result of all this is that a first cut analysis shows that to achieve orbit, a rocket needs about 30,000 feet per second total velocity. It won't ever go that fast, because gravity and drag will have consumed some 5,000 feet per second, but your engines have to add that much energy to the vehicle.
APPROXIMATE REQUIRED VELOCITY TO ORBIT
Orbital Velocity 25,000 feet/second
Allowance for drag 2,000
Allowance for gravitational pull 2,500
Safety Margin 500
TOTAL 30,000 feet/second
You can find out a lot just by playing with the rocket equation, and many of us did: in my case, in high school, after discovering this and much more in Willy Ley's classic Rockets and Space Travel. (Willy's book kept appearing in revised editions over the years: I think the last one was Rockets, Missiles, Space Travel, and Man in Space. It's still one of the best introductory works ever done.)
For one thing, if you want a final velocity of twice your exhaust velocity, you need a mass ratio of e squared, and your calculator (or slide rule when I was in high school; I sure wish I'd had a TI-30) shows that's 7.4.
Well. The orbital velocity of the Earth is about 25,000 feet per second (fps) even without allowances for drag and gravitational pull while you're going up. The best exhaust velocity you'll get from ordinary chemical fuels is around 7,000 fps. Twice 7,000 fps can't even in theory get us to orbit. Neither can three times that 7,000 fps: and at 3 times exhaust velocity, we see from our equations that the mass ratio is e-cubed, or just over 20. Mass ratios of 20 and more weren't considered possible in those days. They're nothing like easy now with the best and most modern materials. Thus, even without allowance for gravity and drag losses, we can see that getting to orbit isn't easy.
We've seen that the Mass Ratio is the mass at takeoff divided by the mass at velocity. The mass at velocity -- in our case what gets to orbit -- is conventionally divided into two components: structure and payload. Payload is the spacecraft you wanted to get into orbit. Structure is everything else: fuel and oxygen tanks, shrouds, casings, rocket engines, pipes, etc. Note that the more structure, the lower the mass ratio. Since a high mass ratio is Good, there came about a mind set among rocketeers: Structure is Bad.
Note also that to get high mass ratios you need BIG rocket ships. That follows because there are minimum weights to many rocket parts. This is known as 'minimum gauge', meaning that you can't make it any thinner or lighter. Combustion chambers have to be rugged. Pipes must be thick enough to hold pressures that don't get smaller just because you're trying to make a small sized rocket. Thus the easy way to build a single stage to orbit rocket ship would be to make it BIG. This is the approach taken by Captain Truax with his Sea Dragon. Unfortunately, size costs. To make a larger ship get off the ground you have to add more engines, or make the ones you have more powerful. Either way is costly. There's also ground handling: clearly it's easier to move, store, and service a smaller ship than a larger one. Truax's Sea Dragon would be built in a shipyard, and towed out to sea to be launched from water.
Big also means inflexible. All your payload will go to the same orbit. There's merit to that, but there's more merit to a capability for putting smaller payloads into a number of different orbits, at least until the market expands to demand the heavier lift.
Back in the days when people first thought a lot about rocket ships, fuels with real exhaust velocities of 7,000 weren't all that easy to come by, nor were rocket engines capable of handling the corresponding temperatures, so speculation about SSTO was pretty theoretical.
Let's look at some possibilities (again from Willy's book):
|Propellant||Combustion Temperature degreee F||Exhaust Velocity: Sea Level||Exhaust Velocity: Average|
|Peroxide/ methyl alcohol||4160||6300||7420|
Those were with practical engines running at 300 pounds per square inch, which was what they knew how to do then. The 'average' value is a bit lower than the actual vacuum exhaust velocity. These are the kinds of numbers people thought about in 1940 - 1950.
(I have inserted the last two from XCOR figures; Max Hunter became very interested in Methane and Propane as an alternative to hydrogen, and this is a convenient place to preserve these numbers. JEP 2003)
So. How were we to get to orbit?
Some simply assumed we'd never do it. Vannevar Bush, Chairman of NASA's predecessor NACA and dean of US Science during World War II and just after, told the Congress of the United States that there was simply no possibility of Inter-Continental Ballistic Missiles. "You can leave that out of your thinking. I wish the American people would leave that out of their thinking," he testified to the House Military Committee in 1946. You couldn't build ICBM's and you certainly couldn't build Moon rockets and launch satellites.
Dr. Bush had great prestige, and his convictions led to great difficulties for the rocket program; but even as he was saying this, others knew better.
Equation (1) condemns us to remain on the Earth's surface unless we can either get very high exhaust velocities, or absurd mass ratios.
The American Rocket Society chose the second way.
Rockets consist of structure, tankage, guidance mechanisms; all kinds of stuff besides the 'payload' you want to get into orbit. Suppose you could throw that junk away when it wasn't needed any more? If you could fling the tanks overboard when they were empty? Then you'd save big time, because you wouldn't have to push that dead weight along.
That, of course, is precisely what a multi-stage rocket does. Incidentally, early books on rocketry generally contain no index reference to 'stages' or 'staging.' They spoke of 'step rockets'.
A typical two step rocket, as described in Willy's book:
Payload of second step 20 lbs
Structure of second step 180
Fuel, second step 400
Total of second step 600
Payload of first step
(total second step) then is 600
Structure of first step 4400
Fuel for first step 10000
Takeoff weight 15000
The takeoff weight is today generally called Gross Liftoff Weight, or GLOW. What's given here is about the size of a Viking sounding rocket.
Now what's our mass ratio here? Well, note that the second step starts at 600 and ends at 200 pounds, for a mass ratio of 3. The first step starts at 15000 and ends with 5000, so it's also a mass ratio of 3. However, for purposes of calculating the Mass Ratio of the total system, we can MULTIPLY the two stage ratios, to get a final mass ratio of nine; meaning that this rocket should get a final velocity of something over 2 times the fuel exhaust velocity; so that if it burned, say, ethyl alcohol and liquid oxygen, which has a sea level exhaust velocity of about 1 1/4 miles per second, we ought to get about 2 1/2 miles per second or 12,000 feet per second. In the real world, the first stage is unlikely to reach 6000 fps because of atmospheric drag and gravity, but the second should get most of the theoretical delta vee.
The ultimate staged system was Saturn/Apollo, which at liftoff stood 363 feet tall, with GLOW of 6,423,000 pounds. The first two stages dropped away to put about 110,000 pounds into orbit preparatory to heading for the Moon -- corresponding to an effective mass ratio of well over 50 had it been a single stage system.
Staging, in other words, can give quite high mass ratios; and that, of course, is how we got to space, to orbit and to the Moon. We built disintegrating totem poles, throwing away most of the rocket in order get velocity. It wasn't an elegant solution to the problem but it did the job.
However, it got us used to thinking of rockets as ammunition, rather than as airplanes. A lot of people never got over that. It also got us to thinking that the ship itself was useless junk: the only important thing was the 'payload' and the sooner you could get rid of everything that wasn't payload, the better off you'd be. That attitude is also still with us.
The absurdity of the disintegrating totem pole was pretty apparent, though, so some continued to work on single stage rockets; and since mass ratios of 20 looked fairly silly (and required enormous ships), they had to look at higher fuel exhaust velocities.
In Willy's engine hydrogen and oxygen at sea level have an exhaust velocity of 10,150 feet per second: 25,000 fps divided by that is 2.46, and e to the 2.46 is 11.7, and while that's a pretty severe mass ratio, it's not utterly absurd. Moreover, as soon as you get above the atmosphere, hydrogen and oxygen give exhaust velocities in the order of 12000 fps, and that corresponds to mass ratios less than ten; and THAT is at the edge of what we knew how to do.
Single Stage to Orbit (SSTO)
SSTO Concepts Continue
Even in Willy Ley's day many rocketeers said "just wait until we have hydrogen engines." Willy was skeptical. Hydrogen was tough stuff to handle. It didn't burn properly in the engines they could build then, it takes insulation to keep it cold, and it's not very dense, so the tanks have to be big. That's all heavy and means a lot of 'structure' which by definition is dead weight. Many people including von Braun thought like that. Hydrogen engines would be great, but they would also have their problems.
The result was that most rocket engineers went down the staging path, which was, after all, the only path to go down if we wanted to accomplish anything in the 50's and 60's; but a few kept looking at Single Stage to Orbit or SSTO. Among them were Phil Bono, Robert Salkeld and Gerry Driggers. Driggers was a professional engineer, and at one time the President of the L5 Society. Salkeld was an associate of Dr. Muller. They, and others, were able to keep the notion of SSTO alive, but by 1980 it was just barely alive.
Shuttle, meanwhile, was originally conceived as a two step rocket with both steps recoverable. One Boeing design called for both steps to be liquid rockets, and both would be manned. The first step flew to the edges of the atmosphere -- say 80,000 feet -- then separated and returned to Earth. The second step flew on to orbit. In at least one design concept, only the first step had wings. The second was a lifting-body design.
There were other Shuttle concepts, but all involved recoverable systems. The notion was to fly to space, do something, return to Earth, refuel, and fly again. Some single-step Shuttle systems were examined, but the initial costs were thought to be very high for a winged vehicle -- radically new engines would be needed -- and most Shuttle concepts were for a two-step system.
Then something horrible happened. Precisely what isn't important; but Shuttle changed from a reasonably elegant reusable system to the Monster That Devoured The Budget.
Partly this was due to engineering competition: Shuttle engines were designed to high performance, to run at high temperatures at very high pressures: something more appropriate to the ammunition concept than reusability.
There was also a desire to put costs off, to stretch things out; to keep initial costs low even if that made final costs higher.
For whatever reason, far from a low cost per launch system, Shuttle real costs per launch grew to a billion dollars each. Even the official cost is now over $500 million per launch.
In 1979 it wasn't entirely clear that Shuttle would become a monster. Some of us had suspicions, but after all, Shuttle was the only game in town; so the space community was asked to swallow its doubts and support NASA to get Shuttle flying, warts and all. Have faith, we were told; and we had faith. Hundreds of Shuttle launches were planned. NASA sold people on the notion of "Getaway Specials", low cost experiments that anyone, even high schools, could do in space. Shuttle was going to make American a spacefaring nation.
That didn't happen. The number of Shuttle flights scheduled was scaled down, then down again. Precisely what happened to Shuttle isn't relevant to this paper, so I'll pass over it, except to note that Shuttle was designed in a way that requires a very large crew of experienced people to keep it flying. You cannot operate Shuttle without a big standing army of technically trained people. It also requires critical ground handling facilities: now that the Vandenberg effort has been abandoned there's only one place in the world that you can launch a Shuttle from.
For all that, we all cheered when Columbia flew, and we hoped for great things.
However, as early as 1980 some people said that Shuttle wasn't going to be what we wanted it to be. One group, led by Carl Sagan and Bruce Murray, said that Shuttle would eat the space sciences budget. Unfortunately, their valid critiques of shuttle were lost because of their vehement attacks on the whole concept of man in space. As Larry Niven once pointed out, every time Sagan convinced someone there was no need for man in space, he as like as not lost a supporter for space. Few Americans wanted to spend money to make the universe safe for robots. They wanted to see heroes go to space and return to ticker tape parades.
To many of the rocket scientists, though, those astronauts were merely 'structure', junk that worsened the mass ratio. We didn't need man in space. Skylab demonstrated that people could be useful after all, but Skylab was an odd mission, not part of the 'main stream' effort; as demonstrated when NASA took a fully operational SKYLAB and the Saturn rocket that could have launched it, and made them into museum displays.
Shuttle was to be the Space Transportation System, and all other systems for putting stuff into orbit were abandoned: indeed, as mentioned, the last fully working Saturn rockets, each capable of putting into orbit about five times a typical Shuttle payload, were destroyed.
Not everyone agreed that Shuttle was the optimum way to go. One lonely voice was Gary Hudson, who was not part of the 'men aren't needed in space' movement.
A Diversion for Definitions
A moment off here for reminding readers what terms mean. SSTO is "Single Stage to Orbit." LOX is Liquid Oxygen, which is a cryogenic fuel. "Cryogenic" means that it must be kept at very low temperatures. LOX is cold stuff, but it's a lot warmer than Liquid Hydrogen. The RL-10 is a Pratt and Whitney 'expander cycle' engine that burns LOX and liquid hydrogen. It was designed as an upper stage, and was employed in the Atlas/Centaur rockets which so greatly aided space science in the early days.
"Storable" fuels are those which can be kept in normal temperatures. Rockets carry both fuel and the stuff to burn the fuel in (generally something that supplies oxygen if not oxygen itself). It's a lot easier to have storable fuels than storable oxidants. Storable fuels can be relatively nice stuff like propane and kerosenes -- the various JP and RP fuels, such as RP-5 = Rocket Propellant 5 are kerosene like. Storable oxidants tend to be nastier stuff, like red fuming nitric acid. Of course hydrazine is a pretty nasty storable fuel; both those latter were used in early air defense missiles.
GLOW means "Gross Liftoff Weight", ie the weight of the fueled ship as it takes off. Dry weight means the empty weight of the ship before you fuel it. Payload can have several meanings, but usually refers to what you wanted to put into orbit other than the ship; for a reusable SSTO the traditional division of the ship into 'structure' and 'payload' may not make a lot of sense.
Council means the Citizens Advisory Counciil on National Space Policy, a group that meets periodically to give advice to national authority on space matters.
In 1980 Hudson presented to the first meeting of the Citizens Advisory Council on National Space Policy a briefing on the Phoenix concept. Phoenix was a fully reusable single stage to orbit LOX Hydrogen ship which he had brought to preliminary design description stage.
The Phoenix concept presented to the Council was a 450,000 pound GLOW reusable vehicle designed to orbit a 5-10 ton payload manned or unmanned. The powerplant would have been either an aerospike -- see below -- or multiple bell nozzle engines. It was to be fabricated from the technology of the day, principally aluminum and composites, and featured active water cooling to handle the heat generated during re-entry. This cooling system, which Boeing called a water wall, was studied in the early 1960's as part of the Boeing Dyna-Soar proposal.
Hudson's proposal was politely heard at the 1980 meeting, but the idea of SSTO was dismissed without much discussion. In retrospect that was an error, and as Chairman I take responsibility for it. I can only plead that Hudson's concept did not even get a second. No one at the meeting -- which included Salkeld, known to be enthusiastic for SSTO -- spoke up in favor of a full debate, and thus SSTO was quietly dropped in favor of space plans involving Shuttle missions. In those days we thought there would BE Shuttle missions. We knew those missions would not be cheap: but we never dreamed that they'd go to a billion a mission. And Hans Mark, incoming Deputy Director of NASA, had personally asked us for support for the Shuttle program.
We wanted to believe. We were still being asked to have faith, and many of us swallowed our doubts to present a common front in favor of "the space program"; and 'the space program' in those days meant Shuttle.
By 1986 it was clear that the US space program was in trouble. People spent their entire professional lives anticipating planetary and space science data from space missions that were long delayed or canceled outright. Even our weather satellite observations were in danger of severe degradation or being lost altogether. The only commercial exploitations of space were communications and some commercialization of ground observation data: and both of those programs were badly underfunded.
It was obvious that if space activity were to grow, it would need more commercial support: space should be a source of money, not a sink for government funding.
In 1986 the Council reports strongly recommended new launch systems. "If you want to be a spacefaring nation you must: Build more rocket ships. Fly more rocket ships," it concluded. We had not as yet fastened on any particular system; there was still hope that a variety of programs might be funded.
By 1987 it was clear that there would not be a variety of programs, nor was Shuttle going to take up the slack. By then the French getting far more commercial space business than we were. Even Red China got in the act. In a final irony, a Chinese refrigerator company was to launch the two commercial satellites the Shuttle recovered and brought down for repair.
The Council met to consider particular systems. We were prepared for extensive debates. During the years 1980-1989 there had been considerable interest in the National Aerospace Plane, NASP, sometimes known as the "Orient Express". This was a winged hypersonic ship that employed air scoops: oxygen would be drawn from the atmosphere, saving the weight of tankage, and of course lowering the liftoff weight and thus presumably improving the mass ratio.
When Hudson again brought up the Phoenix concept in the Fall 1983 Council Meeting, the chief opposition was from advocates of winged vehicles. Winged vehicles seemed very logical to many of us. After all, if you're going to operate like an airline, it's reasonable that the ships look and act like airplanes. The result was that the Council endorsed NASP research, and said nothing about a wingless SSTO.
By 1988 most of the Council members had lost their enthusiasm for winged vehicles. In particular, Max Hunter had become a convert to wingless multi-engine SSTO ships, and had developed a number of SSTO concepts while at Lockheed.
Astonishingly, in the 1988 meeting held to choose a specific line of space ship development, there were almost no debates. Just about everyone present recommended a Single Stage to Orbit Vertical Takeoff/Vertical Landing fully reusable multi-engine system. From having no adherents other than Gary Hudson in 1980, SSTO had become the unanimous choice of a distinguished group of rocket scientists and engineers.
The ship recommended was called SSX, Space Ship Experimental. It was explicitly intended as part of a revival of the highly successful X-airplane program.
SSX was recommended by two different analyses. The first was an engineering analysis that indicated that it could be done, probably with existing engines. There was no agreement on how much payload SSX might carry. Gordon Woodcock was concerned about payload: the ship might not have any at all. His final calculations showed a probable payload of 5,000 pounds using existing engines; but while that sounds like a comfortable number, when we are dealing with mass fractions around 10 -- which we must in a single stage design -- then it doesn't take much change in engine performance or tankage weight to wipe out all the payload.
SSX was to be a "hydrogen rocket" as Willy Ley would have called it. In 1947 the hydrogen engine was an exotic concept. By 1989 the Pratt and Whitney RL-10 hydrogen engine had been tested thousands of times without failures. There had been other changes.
The RL-10 had been flown often enough to give hard performance data:
Performance of the RL-10
(Isp, or 'specific impulse,' is now usually used instead of exhaust velocity. It is nothing more than the exhaust velocity divided by gravity, or 32.17 feet per second per second. The main reason Isp is generally used now is that it's a smaller number, but it is sometimes described as an index of efficiency: it is the pounds of thrust per pound of fuel per second of acceleration.)
Assuming we need 30,000 feet per second, at an exhaust velocity of 10615 (Isp of 330) we need a mass ratio of 16.88, which is beyond achievement; but at 14478 (Isp = 450) it's only 7.6. Since the rocket spends more time in vacuum, the average exhaust velocity during the flight is going to be higher than 10615; precisely what it will be can be calculated, but the calculations are sensitive to the flight path, and that's sensitive to the structural weight; meaning there's enough room for experts to disagree. However, the numbers are intriguing, and recall, these are figures for well tested, reliable, existing engines.
After three days of intensive meetings, we concluded that the SSX could be built, for under a billion dollars, and in under four years time. This was an engineering, not a political, analysis: one rule of the Council is that we will not paralyze ourselves with political pessimism. If a concept is technically feasible, we will recommend it and hope that it will find a political -- or commercial -- champion. We saw no technical reasons why SSX could not be constructed using existing technologies and generally employing modifications of existing equipment.
A second meeting, held in Fall 1988, confirmed this conclusion.
* * *
The other line of argument for SSX was more theoretical.
Start with this. Most people in the space community still think of ships as ammunition rather than aircraft. What happens if you think of them as aircraft?
Airlines typically operate at a small multiple of fuel costs. It takes roughly the same fuel to fly a pound to Australia as it does to put that pound in orbit. Granted that liquid hydrogen will be rarer and more expensive than jet fuel, the actual costs of space flight with Shuttle are not several times fuel costs but several hundred times those costs. Rockets are not less efficient than jet engines. Why, then should space flight cost so much?
The answer wasn't hard to find. First, the cost of a ticket to Sydney would be a very great deal more than a few times fuel costs if they had to throw away the airplane after you took a trip in it. Clearly expendables -- ammunition -- could never be as cheap as airline style operations. You have to reuse the ship.
That line of thinking had led to Shuttle, and it certainly wasn't cheap. What went wrong there?
That answer wasn't hard to find either. Typical airlines have 120 employees per airplane, and most of those sell tickets. The highly technical SR-71 program had about 48 employees per airplane, many of them data analysts. By contrast, Shuttle has 20,000 people to operate 4 vehicles. The typical annual cost of a technical employee today (including benefits, retirement, office facilities, etc.) is over $125,000 a year, meaning that a standing army of 20,000 costs at least $2.5 billion a year before they launch anything at all.
Clearly Shuttle was not designed for operational simplicity. That provided the primary design directive for SSX: its design would be driven by operations considerations, not by performance. Performance has to be "Good Enough'. It doesn't have to be a lot better. Better is the enemy of good enough. The important thing for SSX is simplicity of operations; low costs and low personnel requirements.
Another SSX design criterion was SAVABILITY. It ought to be able to endure an engine out on takeoff: to hover, burn off fuel, and land. Clearly a desirable feature, and one that has a vast influence on total space operations costs.
There are other benefits, but surely it is established that if you can build this kind of rocket ship, you would want to have one.
Gary Hudson's original Phoenix was a bit smaller than the SSX we envision.
We thought in terms of a ship with about 45,000 pounds of structure, and 5,000 pounds of payload. With a Mass Ratio of 10 this produces a GLOW of 500,000, and requires an average exhaust velocity of 13029 feet per second. This is just within the capability of existing engines, and the structure weights are thought to be achievable.
Clearly none of these round numbers is sacred. The truth is, until we begin to build that ship, we won't know precisely what those numbers are: and that's the rub, because if we size the ship small to keep the total costs down, we run the risk of not being able to make orbit, or of making orbit with no payload.
Thus some proposals for DC/Y, a possible implementation of the SSX concept, call for a GLOW of 1.4 million pounds. This makes for a large ship, complicating ground handling, and requiring considerable thrust which in turn requires engine developments. That large a ship will have increased costs and complexity of operation, so a smaller ship is preferred. However, the big ship is in a sense a 'conservative' design, in that it contains considerable reserves of payload mass that can be sacrificed to compensate for unforeseen problems.
The Council has no unanimous recommendations on sizing; there are members willing to defend a wide range of values. There is near unanimous agreement that we have no more to get out of continued analysis of existing data. It is time to FLY something; without new flight test data, it is impossible to choose the optimum design size.
We recommended that we build the SSX and get on with the program; at worst we'd learn what we need to know to build the proper size bird.
Design for Test Flights
At the Council meeting that recommended SSX it was unanimously agreed that a ship using RL-10 engines, modified for altitude compensation, would have a marginal orbital capability: it might not be able to make orbit, but it would be close to that capability. That ship could be flown incrementally. That is, it could take off not fully fueled and fly for short distances and times, thus allowing an exact determination of its capabilities. It could then be modified: add thrust, or reduce weight, to 'nickel and dime' its way to an orbital capability.
This incremental test philosophy seems to have been misunderstood -- in some cases rather deliberately misunderstood.
"Almost make orbit" is of course a pretty dangerous thing to do -- if you try to make orbit and fail. However, a ship that can "almost make orbit" can be used for a LOT of testing in test missions that don't at all try for orbit. At each test stage you gain a new appreciation of the ship's capabilities, and the next test is designed accordingly.
Given the experience of flying a ship that can 'almost make orbit' we would fully understand the requirements of a ship that DOES make orbit. This incremental approach to space ship design is similar to the X airplane programs that produced such dramatic results in US aviation.
The original SSX concept envisioned building a couple of ships that would "almost make orbit": with luck and skill, one of them might actually be sent on an orbital mission. It would probably have no payload other than itself, but so what? The payload of the X-1 was Major Yeager. The mission wasn't to deliver payload, but to fly faster than sound. Similarly, the mission of SSX-1 is to develop the means for airline-like space operations, not to deliver payload into space.
SSX-2 would take what was learned from SSX-1 to build a ship that would reliably go to orbit and return. Once again, the amount of payload wasn't critical; but everyone was certain that with engines developed from what was learned by FLYING SSX-1 we would have a payload of at least 9,000 pounds, and with data developed by FLYING that ship, we could improve SSX-2 payload considerably: some expected as much as 19,000 pounds.
Ammunition or Airplane?
Many rocket systems were developed for use as ICBM's. A missile needs performance and one-time reliability. There's no point in building the engine to last. It won't be used but once (and with luck never at all).
The result was that many rocket engineers fell into the habit of thinking of rocket motors as part of an artillery system; they were ammunition, and it was pointless to build them for anything else. Even when the system was to launch a satellite, or a manned spacecraft, it wasn't likely to be used again.
Shuttle was of course different, but except for Shuttle there wasn't much effort put into reusable, restartable, rocket engines.
Even so, there was an exception. The Pratt and Whitney RL-10 is ultra-reliable and test engines have been restarted and refired many times. The P&;W RL-10 was developed by engineers exclusively familiar with aircraft engines, not missiles, and many think that was a major influence on the engine design. In any event the RL-10 demonstrates that you can build reliable, reusable rocket engines if that's what you set out to do.
It was unanimously agreed that the technical risk of the SSX program, through SSX-2 with a payload of at least 10,000 pounds, was a lot lower than the technical risk of Apollo: that is, there are far fewer technical unknowns in the SSX program than there were in going to the Moon when Kennedy announced that mission.
Exactly how far single stage to orbit -- SSTO -- technology could be taken wasn't predictable until we had built and FLOWN some of these machines. However, it was clear that any advances in material science -- such as were coming out of the NASP program -- would instantly benefit the SSX. In general, anything that makes NASP more feasible helps SSX at least as much.
There were some technical unknowns. One was the engine concept known as "aerospike". This is a method of mounting a number of engines in a pattern that allows their exhaust plumes to work in a way that does automatic altitude compensation.
Explaining aerospike and altitude compensation requires more technical detail than I want to give in this paper. The important thing to remember is that rockets are more efficient in a vacuum than in an atmosphere (which is why you want to get out of the atmosphere as quickly as possible when you fly rockets). Whatever is done to make the rocket work better in an atmosphere will detract from its performance in vacuum. It can also add weight. The aerospike concept tries to get around some of those limits.
Aerospike engines were built and static tested decades ago. The results are encouraging, but hardly decisive. You can find people who are convinced they can build a working aerospike design with existing engines, and others who are convinced that if you do build it you won't get enough altitude compensation to make it worth the effort. The consensus seems to be that aerospike is the proper way to do altitude compensation, but there are some doubts about the effects of dynamic pressures at certain critical altitudes between 25,000 and 75,000 feet. In that flight regime the effect of dynamic or slipstream pressures could be to lower Isp by as much as 30 seconds, corresponding to a lowering of exhaust velocity by 965 feet per second. However, this applies only in the regime around 50,000 feet, and the rocket doesn't stay there very long; so the total effect on average exhaust velocity (and thus mass ratio, and thus payload) for the flight isn't accurately known. The question won't be resolved without flight testing.
Since aerospike engines have great benefits, it seems reasonable to have a program to develop both aerospike and the proper engines for use with them. It wouldn't cost more than fifty million dollars a year, which is pretty trivial given the potential gain. In particular, SSX / SSTO benefits from the gains our analysis says you'll get with a working aerospike. However, SSTO will work without aerospike, and while the original SSX concept employed aerospike in the proposed design, it was not considered essential to the SSX concept. You have to do altitude compensation, but that can be done with variable geometry rocket bells.
Another technical unknown is re-entry geometry. If there's anything we understand about re-entry it's blunt cones, so that a nose first re-entry vehicle is easy to conceive. Moreover, nose first gives considerable increases in cross range maneuver capability for landing. However, since we want to land SSX tail down, and getting a blunt cone from nose first to tail first isn't easy, we either have to learn how to make that rotation or we have to let SSX re-enter tail first.
Tail first entry has many advantages. Among them is the possibility of running the engines at idle, letting the exhaust plume serve as the heat protection mechanism for the ship. Once again, this is not a concept critical to the success of SSX, but it does or could save structural weight. It is also another unknown: we can simulate all we like, but we won't know how well that works until we try it, and right now we aren't trying much of anything.
One bugaboo that isn't as big a problem as it looks is the tail first landing. This looks frightening, since if the fire goes out the ship falls: it can't possibly glide. Here is this enormous thing falling. . . However, once the ship has re-entered it is nearly empty. The terminal velocity due to atmosphere is something under 100 miles per hour. The empty weight of the ship is less than that of an airplane. No one wants to crash a space ship, but the consequences of crashing an SSX are a lot lower than those of crashing a shuttle or a 747.
The real debate over single stage to orbit involves materials and structures. Clearly we know how to build the ship. The question is, will the structure -- complete with frame, tanks, motors, controls, re-entry shield, landing gear, crew and cargo compartments -- be light enough to allow the ship to have a reasonable payload?
Some questions on structure weight were resolved in a series of classified experiments code named HAVE REGION. Recently declassified, these were conducted by the Strategic Defence Initiative Office. In HAVE REGION several contractors built scaled versions of sections of a single stage to orbit ship, using rather conservative mixtures of aluminum and composites. The results are not self interpreting, but structural experts have concluded that HAVE REGION demonstrated that vehicles having mass ratios and strengths better than those required for single stage to orbit can be constructed. The questions remaining have to do with costs.
Most experts who have studied single stage to orbit technology are agreed that an SSTO ship can be built, and that it will have some payload; and while there is debate over how much payload -- estimates range from 5,000 to 19,000 pounds to Low Earth Orbit (LEO) -- nearly everyone is agreed that a practical, savable, and resuable ship with about 9,000 pounds payload can be constructed. The question is, can a ship that lightly built be used enough times to justify its construction?
Note that this is question is as much economic as technical.
There are other technical risk factors in SSX design, and they all work the same way: if things turn out well that increases performance, and turning out the other way doesn't. If every one of those factors goes the wrong way, SSX won't have much payload. If every one of them goes REALLY the wrong way, it won't have any payload at all. However, there is no reason to suppose that they'll all go wrong. Things don't usually work that way. In fact, the way to bet any one of them is that it will go right: Clarke's Law states that if a venerable scientist tells you that something is possible, he's right, and if he says it's impossible he's most likely wrong; and experience has proven him correct. (Clarke, of course, remembers Vannevar Bush very well.)
The Real Risk
The real risk to SSX was always known to be organizational.
The first risk is capture by a group that wants to study it to death.
There are two major engineering design philosophies. One says, "Do things right. Don't waste money putzing around, really understand what you're doing, figure out in advance all that can go wrong, and then build it right in the first place."
The other says "Do the best you can, learn from that, and then go on from there. You'll learn more from one good flight than from a thousand computer simulations."
Both concepts have merit. The first can be thought of as the 'prototype' approach. You don't build anything unless you'd be willing to build a lot of them. The other is the "X vehicle" approach: X ships aren't prototypes. Build two, fly one until you prang it, use what you learned to modify the other one, and fly that until you augur it in or there's nothing more to learn.
We are convinced that SSTO requires the X concept. Alas, there's nothing so magical about SSTO that it is immune from the "let's do it right" approach. Capture by a 'do it right' group would certainly delay SSTO well into the next century.
Another danger is the 'requirements' group. X concepts are to develop TECHNOLOGIES, not vehicles; if you start hanging specific requirements on an experimental vehicle, you will end up trying to 'do it right'; which is generally fatal to an X program. Single Stage to Orbit ships will have many customers; but the potential users of SSTO should not dictate the development program.
The conclusion is obvious. We believe that SSX is possible; that an SSTO ship could be built and flown in fewer than four years and for less than a billion dollars. We believe that doing that will show how to design and build SSTO vehicles with payloads of about 15,000 pounds, operate at costs of 4 or 5 times fuel cost, and require about 50 technical persons per vehicle.
Such a ship would be 'savable': it could experience an engine out without disaster.
Savability plus low cost to orbit will have an enormous effect on payload design, as well as on commercial uses of space. Payloads need not be overdesigned, since the cost to orbit will not be enormous. Obsolete satellites can be replaced.
Airline style operations to space will lead to space commercial activities on a wide scale; look at what happened to air travel when reliable systems with low operating costs became available.
SSX, we believe, could lead to the ships that will make America a Spacefaring nation as promised by John Kennedy all those years ago.
ADDENDUM: The Lockheed disaster.
After DC/X flew Lockheed got NASA funds to do an program called X-33. It wasn't an X program. Nothing ever flew, and we learned little to nothing. It absorbed all the money that might have gone into a flyable SSX.
If we learned anything from that it is this: build things to fly with X project money. Build ships. Fly ships. Don't try to do it all in one great big leap with one ship: build X ships, and fly X ships.
- 30 -
APPENDIX: Some Data
The following data on LOX/LH2 engines is taken from Rocketdyne publications.
J-2 230,000 lb Thrust
427 sec Isp (Vac) = 13,715 fps e.v.
Area Ratio 27.5:1
O/F Mixture Ratio 5.5:1
Chamber Pressure 763 psia
Weight 3480 lb
J-2S 265,000 lb Thrust
435 sec Isp (Vac) = 13,970 fps e.v
Area Ratio 40:1
O/F Mixture Ratio 5.5:1
Chamber Pressure 1246 psia
Weight 3800 lb
SSME 513,000 lb Thrust
455 sec Isp (Vac) = 14614 fps e.v
Area Ratio 77.5:1
O/F Mixture Ratio 6.1:1
Chamber Pressure 3240 psia
Weight 6990 lb
Linear 200,000 lb Thrust
455 sec Isp (Vac) = 14614 fps e.v.
Area Ratio 115:1
O/F Mixture Ratio 5.5:1
Chamber Pressure 1224 psia
(The Linear aerospike was not build as a flight weight engine, so the engine weight is uncertain.)
RL10A-3-3A 16,500 lb Thrust
444 sec Isp (Vac)
Area Ratio -----
O/F Mixture Ratio 5.0:1
Chamber Pressure ----
Weight 310 lb
RL10A-4 20,800 lb Thrust
449 sec Isp (Vac)
Area Ratio -----
O/F Mixture Ratio 5.5:1
Chamber Pressure ----
Weight 370 lb
(The RL10 data is from a P&;W handout which did not give the area ratio and chamber pressure numbers. The new RL10A-4 uses an extendable nozzle to increase the available area ratio.)
All the above thrust numbers are vacuum thrust.