# (Semi-) Technical Aspects of SSTO

(Semi-) Technical Aspects of SSTO This paper will try to give some idea of why SSTO makes technical sense and is a reasonable idea. We'll concentrate on the overall issues, trying to give you the right general idea without getting bogged down in obscure detail. Be warned we will oversimplify a bit at times.

### Why Is Building An SSTO Challenging?

Getting a one-stage reusable rocket into orbit doesn't look impossible, but it does look challenging. Here's why.

The hard part of getting into orbit is not reaching orbital altitude, but reaching orbital velocity. Orbital velocity is about 18,000 mph. To this, you have to add something for reaching orbital altitude and for fighting air resistance along the way, but these complications don't actually add very much. The total fuel requirement is what would be needed to accelerate to 20 to 21,000 mph.

So how much is that? (If you don't want to know the math, skip to the next paragraph for the results.) The "rocket equation" is (desired velocity) = (exhaust velocity) x [ln (launch weight / dry weight)], where "ln" is the natural logarithm. The exhaust velocity is determined by choice of fuels and design of engines, but 7,000 mph is about right if you don't use liquid hydrogen, and 10,000 mph if you do.

The bottom line is that the launch weight has to be about 20 times the dry weight (the weight including everything except fuels) if you don't use liquid hydrogen, and about 8 times the dry weight if you do. This sounds like hydrogen would be the obvious choice of fuel, but in practice hydrogen has two serious problems. First, it is extremely bulky, meaning that hydrogen tanks have to be very big; the Space Shuttle External Tank is mostly hydrogen tank, with only the nose of the tank containing oxygen. Second, some of the same properties that make hydrogen do well on the weight ratio make it difficult to build hydrogen engines with high thrust, and a rocket does need enough thrust to lift off! Both of these problems tend to drive up the dry weight, by requiring bigger and heavier tanks and engines.

So how bad is this? Well, it's not good. Even with hydrogen, an SSTO launcher which weighs (say) 800,000 lbs at launch has to be 7/8ths fuel. We've got 100,000 lbs for tanks to hold 700,000 lbs of fuel, engines to lift an 800,000 lb vehicle, a heatshield to protect the whole thing on return, structure to hold it all together at high acceleration...and some payload to make it all worthwhile. Most of the dry weight has to go for the vehicle itself; only a small part of it can be payload. (That is, the "payload fraction" is quite small.) To get any payload at all, we need to work hard at making the vehicle ("dry") very lightweight.

The big problem here is: What happens if the vehicle isn't quite as light as the designer thought it would be? All rockets, and most aircraft for that matter, gain weight during development, as optimistic estimates are replaced by real numbers. An SSTO vehicle doesn't have much room for such weight growth, because every extra pound of vehicle means one less pound for that small payload fraction. Particularly if we're trying to build an SSTO vehicle for the first time, there's a high risk that the actual payload will be smaller than planned.

That is the ultimate reason why nobody has yet built an SSTO space launcher: Its performance is hard to predict. Megaprojects like the Shuttle can't afford unpredictability -- they are so expensive that they must succeed. SSTO is better suited to an experimental vehicle, like the historic "X-planes," to establish that the concept works and get a good look at how well it performs.

### Why Does SSTO Look Feasible Now?

The closest thing to SSTO so far has been the Atlas expendable launcher. The Atlas, without the Centaur upper stage that is now a standard part of it, has "1.5" stages: It drops two of its three engines (but nothing else) midway up. Without an upper stage, Atlas can put modest payloads into orbit: John Glenn rode to orbit on an Atlas. The first Atlas orbital mission was flown late in 1958. But the step from 1.5 stages to 1 stage has never really been attempted.

Actually, people have been proposing SSTO launchers for many years. The idea has always looked like it just might work. For example, the Shuttle program looked at SSTO designs briefly. Mostly, nobody has tried an SSTO launcher because everybody was waiting for somebody else to try it first.

There are a few things that are crucial to success of an SSTO launcher. It needs very lightweight structural materials. It needs very efficient engines. It needs a very light heatshield. And it needs a way of landing gently that doesn't add much weight.

Materials for structures and heatshields have been improving steadily over the years. The NASP program in particular has helped with this. It now looks fairly certain that an SSTO can be light enough. Existing engines do look efficient enough for SSTO, provided they can somehow adapt automatically to the outside air pressure. The nozzle of a rocket engine designed to be fired in sea-level air is subtly different from that of an engine designed for use in space, and an SSTO engine has to work well in both conditions. (The technical buzzword for what's wanted is an "altitude-compensating" nozzle.) Solutions to this problem actually are not lacking, but nobody has yet flown one of them. Probably the simplest one, which has been tentatively selected for the DC-Y, is just a nozzle which telescopes, so its length can be varied to match outside conditions. Making nozzles that telescope is not hard -- many existing rocket nozzles, like those of the Trident missile, telescope for compact storage -- but nobody has yet flown one that changes length while firing. However, it doesn't look difficult, and there are other approaches if this one turns out to have problems.

We'll talk about landing methods in more detail later, but this is one issue that will be resolved, and soon. One of the primary goals of the DC-X1 experimental craft is to fly SSTO landing maneuvers and prove that they will work.

So...with materials under control, engines looking feasible, and landing having been test-flown, we should be able to build an SSTO X-vehicle demonstrator. The X-vehicle's performance may not quite match predictions, but if it works at all, it will make all other launchers obsolete.

### Why A Rocket?

As witness the NASP (X-30) program, air-breathing engines do look like an attractive alternative to rockets.Much of the weight of fuel in a rocket is oxygen, and an air-breathing engine gets its oxygen from the air rather than having to carry it along. However, on a closer look, the choice is not so clear-cut. The biggest problem of using air-breathing engines for spaceflight is that they simply don't work very well at really high speeds. An air-breathing engine tries to accelerate air by heating it. This works well at low speed. Unfortunately, accelerating air that is already moving at hypersonic speed is difficult, all the more so when it has to be done by heating air that is already extremely hot. The problem only gets worse if the engine has to work over an enormous range of speeds: NASP's scramjet engines would start to function at perhaps Mach 4, but orbital speeds are roughly Mach 25. Nobody has ever built an air-breathing engine that can do this...but rockets do it every week.

Air-breathing engines have other problems too. For one thing, to use them, one obviously has to fly within the atmosphere. At truly high speeds, this means major heating problems due to air friction. It also means a lot of drag due to air resistance, adding to the burden that an air-breathing engine has to overcome. Rocket-based launchers, including SSTO, do most of their accelerating in vacuum, away from these problems. Perhaps the biggest problem of air-breathing engines for spaceflight is that they are heavy. The best military jet engines have thrust:weight ratios of about 8:1. (This is at low speed; hypersonic scramjets are not nearly that good.) The Space Shuttle Main Engine's thrust:weight ratio, by comparison, is 70:1 (at any speed). The oxygen in a rocket's tanks is burned off on the way to orbit, but the engines have to be carried all the way, and air-breathing engines weigh a lot more.

And what's the payoff? The X-30, if it is built, and if it works perfectly, will just be able to get into orbit with a small payload. This is about the same as SSTO, at ten times the cost. Where is the gain from air-breathing engines?

The fact is, rockets are perfectly good engines for a space launcher. Rockets are light, powerful, well understood, and work fine at any speed without needing air. Oxygen may be heavy, but it is cheap (about five cents a pound) and compact. Finally, rocket engines are available off the shelf, while hypersonic air-breathing engines are still research projects. Practical space launchers should use rockets -- and, so, SSTO does.

### Why No Wings?

With light, powerful engines like rockets, there is no need to land or take off horizontally on a runway, and no particular reason to do so. Runway takeoffs and landing are touchy procedures with little room for error, which is why a student pilot spends much of his time on them. Given adequate power, vertical takeoffs and landings are easier. In particular, a vertical landing is much more tolerant of error than a horizontal one, because the pilot can always stop, straighten out a mistake, and then continue. Harrier pilots confirm this: Their comment is "it's easier to stop and then land, than to land and then try to stop." What if you don't have adequate power? Then you are in deep trouble even if your craft takes off and lands horizontally. As witness the El Al crash in Amsterdam, even airliners often don't survive major loss of power at low altitude. To make a safe horizontal landing, especially in less-than-ideal weather conditions, you must have enough power to abandon a bad landing approach and try again. Shuttle-style gliding landings are dangerous, and airline crews go to great lengths to avoid them; the Shuttle program, with the nation's best test pilots doing the flying and no effort spared to help them, has already had one near-crash in its first fifty flights. Routine access to space requires powered landings.

If we are going to rely on powered landings, we must make sure that power will be available. Airliners do this by having more than one engine, and being able to fly with one engine out. SSTO is designed to survive a single engine failure at the moment of liftoff, and a second failure later. Since (at least) 7/8ths of the takeoff weight of SSTO is fuel, it will be much lighter at landing than at takeoff. Given good design, it will have enough power for landing even if several engines fail. If SSTO has an engine failure soon after liftoff, it will follow much the same procedure as an airliner: It will hover to burn off most of its fuel (this is about as quick as an airliner's fuel dumping), and then land, with tanks nearly empty to minimize weight and fire hazard.

Note that in an emergency, vertical landing has one major advantage over horizontal landing: Horizontal landing requires a runway, preferably a long one with a favorable wind, while a vertical landing just requires a small flat spot with no combustible materials nearby. A few years ago, a Royal Navy Harrier pilot had a major electronics failure and was unable to return to his carrier. He made an emergency landing on the deck of a Spanish container ship. The Harrier suffered minor damage; any other aircraft would have been lost, and the pilot would have had to risk ejection and recovery from the sea.

Given vertical landing and takeoff, is there any other use for wings? One: Crossrange capability, the ability to steer to one side during reentry, so as to land at a point that is not below the orbit track. The Shuttle has quite a large crossrange capability, 1500 miles. However, if we examine the history of the Shuttle, we find that this was a requirement imposed by the military, to make the Shuttle capable of flying some demanding USAF missions. A civilian space launcher needs a crossrange capability of, at most, a few hundred miles, to let it make precision landings at convenient times. This is easily achieved with a wingless craft; the Apollo spacecraft could do it.

Finally, wings are a liability in several important ways. They are heavy. They are difficult to protect against reentry heat. And they make the vehicle much more susceptible to wind gusts during landing and takeoff (this is a significant limitation on Shuttle launches).

SSTO does not need wings, would suffer by carrying them, and hence does not have them.

### Why Will It Be Cheap And Reliable?

This is a good question. The Shuttle was supposed to be cheap and reliable, and is neither. However, there is reason for hope for SSTO.

The Shuttle's costs come mainly from the tremendous army of people needed to inspect and refurbish it after each flight. SSTO should get by with many fewer.

The basic SSTO concept opens major possibilities for simple, quick refurbishment. With no discarded parts, nothing needs to be replaced. With no separating parts, there is no need to re-assemble anything. In principle, an SSTO vehicle should be able to "turn around" like an airliner, with little more than refuelling. Of course, this is easier said than done. But there is no real reason why SSTO should need much more. Its electronics experiences stresses not much worse than those of an airliner -- certainly no worse than those of a jet fighter. Its structure and heatshield, designed to fly many times, will have sufficient margins that they will not need inspection and repair after every flight. Most space-vehicle components don't inherently need any more attention than airliner components. The one obvious exception is the engines, which do indeed run at much higher power levels than airliner engines. But even here, airliner principles can be applied: The way to make engines last a long time is to run them at less than 100% power. SSTO engines have it easy in one respect: They only have to run for about ten minutes at the start of the flight and two or three minutes at the end.

Still, the Shuttle engines certainly are not a shining example of low maintenance and durability. However, it's important to realize that the Shuttle engines are not the only reusable rocket engines. Most liquid-fuel engines could be re-used, were it not that the launchers carrying them are thrown away after every flight. And the durability record of these other engines -- although limited to test stands -- is much better. The RL-10 engine, used in the DC-X1, is rated to fire for over an hour, in one continuous burn or with up to ten restarts, with no maintenance. Several other engines have comparable records. Conservatively-designed engines are nowhere near as flakey and troublesome as the Shuttle engine. Here again, the DC-X1 has supplied solid evidence. Although its engines and other systems are not the same ones that an SSTO X-vehicle demonstrator would use, they should be representative enough to demonstrate rapid, low-effort refurbishment, and the DC-X1 program has done so. Airliners typically operate at about three times fuel costs. The fuel cost for an SSTO vehicle would be a few dollars per pound of payload. It may be a bit optimistic to try to apply airline experience to the first version of a radically new vehicle. However, even advanced aircraft typically cost no more than ten times fuel cost. Even if SSTO comes nowhere near these predictions, it should still have no trouble beating existing launchers, which cost five thousand dollars per pound of payload -- on up. We can look at this another way: Head counts. Airlines typically have about 150 people per aircraft, and most of those sell tickets or look after passengers' needs. Perhaps a better example is the SR-71, which is like SSTO in that it was an advanced craft, pushing the frontiers of technology, operated in quite small numbers. Although it is hard to get exact numbers because of secrecy, it appears that USAF SR-71 operations averaged perhaps one flight per day, using perhaps eight flight-ready aircraft, with a total staff of about 400 people. That's 50 per aircraft. If SSTO can operate at such levels -- and there is every reason to think it can -- it should have no trouble beating existing launchers, which typically have several thousand people involved in preparations for each and every launch. (NASA's Shuttle ground crew is 18,000 for a fleet of four orbiters flying about eight flights a year.)

As for reliability, the crucial reason for thinking that SSTO will do a lot better than existing launchers is simple: Testing. It should be feasible and affordable to test an SSTO launcher as thoroughly as an aircraft. This is vastly more thorough than for any launcher. The F-15 fighter flew over 1,500 test flights before it was released for military service. No space launcher on Earth has flown that many times, and the only one that even comes close is an old Soviet design. It is no wonder that the Shuttle is somewhat unreliable, when it was declared "operational" after a grand total of four test flights. By aircraft standards, the Shuttle is still in early testing. Some expendable launchers have been declared operational after two tests. Each and every SSTO vehicle can be tested many times before it carries real payloads. Moreover, since SSTO can survive most single failures, it can be tested under extremes of flight conditions, like an aircraft. For example, unlike Challenger, an SSTO vehicle would launch with passengers and cargo in freezing temperatures only after multiple test flights in such conditions. There will always be surprises when a new craft is flown in new conditions, but SSTO should encounter -- and survive -- most of them in test flights.

### The Bottom Line

Although there is reason for some uncertainty about the exact performance of the first SSTO spacecraft, the basic approach being taken is sensible and reasonable. The test flights of the DC-X1 test craft have resolved many technical concerns. Nobody can be sure about costs and reliability until an SSTO X-vehicle demonstrator is flying, but there is reason to believe that SSTO should be much better than current launchers. If the program is carried through to a flying prototype, it really could revolutionize spaceflight.