Why SpaceShipOne Never Did, Never Will, And None Of Its Direct Descendants Ever Will, Orbit The Earth
by Karen Pease

In the immediate aftermath of SpaceShipOne's historic flight, many were rightfully elated at the example of what a small group of people can accomplish if they set their minds (and pocketbooks!) to a task. A new wave of space enthusiasm poured over the geek community.

Unfortunately, misconceptions have abounded about what this flight really was, and the nature of space travel itself. After having to constantly correct the same misconceptions over and over, I figured it was time to write a FAQ. This document will be presented in the form of a typical argument over the significance of SS1.


* NASA is toast! SpaceShipOne accomplished on 25 million what NASA spends billions on!

No, it didn't. Apart from the fact that the cost was "at least 25 million" (they refuse to disclose the actual expenses), the most minimal non-experimental launches that NASA does are to LEO (Low Earth Orbit). SS1 didn't even come remotely close to LEO. Anything less than LEO is just a joyride. What Rutan did was far more comparable to building an airplane than a space ship.

* Nuh uh! It went into space!

It went into "space", to very loosely define the term, for roughly 3 minutes of almost-complete weightlessness (shorter than some people can even hold their breath). A sense of "weight" occurs when the forces acting uniformly on your body meet a positional resistance. For example, on the ground, the force applied from the ground to your feet which counters the uniform force of gravity on your body (and the corresponding force that your feet apply to your legs, then your spine, etc) give you a sense of weight. Likewise, on an airplane, the lift from the wings, applied to the plane and then to your seat, serves the same purpose. In an airless freefall, you get no resistive force, so gravity acts uniformly on your body and you feel weightless. Consequently, the 3 minutes corresponds to the period in which atmospheric resistance was insignificant, and thus in "space".

However, being in orbit is quite different. To maintain a stable LEO orbit, you need a velocity of over 7,800 meters per second. This speed is so high that it is usually several times higher than the very exhaust that left the rocket to get it up to this speed! Since air resistance of macroscopic objects is generally proportional to the velocity squared, what was sparse enough to be "weightless" to the SpaceShipOne pilot would be significant resistance for any orbital craft. Consequently, LEO craft typically orbit at least 300 km in altitude.

* So what? It was close.

No. It wasn't. The additional altitude needed for useful space flight is the easy part. It's the orbital velocity that is the hard part. In addition to the needed 7,800+ delta-V for reaching orbital velocity, there's also generally around 2,000 delta V for drag and gravity losses. SpaceShipOne's delta-V is only 1,700 m/s.

* Well, that's easy - they'll just scale it up to be 6 times bigger!

Unfortunately, things don't work that way. Because not only do you have to carry up the mass of your craft, but you have to also carry up the mass of your fuel. And when you add the fuel, suddenly your craft weighs more, so you need more fuel. Etc. Consequently, craft scale up in geometric progression. The exponent is roughly determined by the craft's ISP, while the scaling factor is determined by the mass of the craft (which is largely due to the tank mass). Tank mass is itself largely depdentant on one factor: the pressure. Use of self-pressurized tanks with no turbopump, consequently, means a heavy spacecraft (*).

Unfortunately, SpaceShipOne comes in on the bad side of all of these. It has a low ISP (around 250) and a higher tank mass than would be necessary if it used a turbopump. The net result is that if you wanted to scale it up to be able to reach orbit, you'd have to have a White Knight that makes the AN-255 Cossack (the largest airplane ever built - designed as a carrier for the Buran shuttle) look small by comparison. It's not a realistic proposition.

This is best represented in the mass fraction of the rocket. A good LOX/LH engine will give you a 7:1 mass fraction - that is, 7 times more propellant than the combination of structural mass and payload. A good LOX/Kerosene rocket will give you 20:1. SpaceShipOne will be several hundred to one. Then, when you factor in the issue of how heavy its self-pressurized tanks are, you realize that you'll never be able to scale it up.

* Ok, so Rutan can't just scale it up - but he's going to use all of the experience he got building it to get better ISP and tank mass next time.

First off, doing it the way he did (i.e., specifically designed to reach a prize) means that 90% of the development has gone in a direction that can't advance up to orbit, and would need to be started from scratch. But there's a much more fundamental problem at hand. First, we look at a diagram of how SS1's engine works:

Nice and simple, isn't it? It doesn't just look simple because it is a schematic representation - it actually is that simple. That's exactly why it was built this way. They were given a relatively easy challenge, and so chose a relatively simple engine. By choosing such a simple engine, they were able to have it built by a tiny outfit that has the capability to make such engines quite cheaply; no complicated manufacturing, assembly, or most importantly, maintinance is needed. But what if you want to get real (orbital)-level performance out of an engine? The required complexity increases very quickly. Here's the SSME (Space Shuttle Main Engine):

Complicated, isn't it? It has several thousand nontrivial parts; here's a page on how it works; it'd take me too long to describe here. It operates at temperatures ranging from just a little above absolute zero to above the boiling point of iron, and still manages to be reusable. It's probably one of the most impressive machines ever built in the sort of conditions it can tolerate. It makes million dollar jet engines look like child's toys by comparison (and actually has a lot of similarities to jet engines, except that its components are subject to much harsher conditions and need much more extreme performance, in addition to not benefitting from mass production as jet engines do).

Now, it doesn't take an SSME to get to orbit (although it helps!). But it still takes engines far more complicated and difficult to build than the ones Rutan got from SpaceDev(**). At the very least, you're going to need single-stage turbopumps, much better nozzle alloys, and a more elaborate cooling system. These things make *real* rockets become complex rather quickly.

* Ok, ok, lets say he spends a couple tens of millions buying engines from Lockheed or some other company. And lets say he completely redesigns the rest of the ship. He'll still be a better pick than NASA.

Apart from the fact that this is essentially starting over, no, he won't. Rutan builds aircraft. SpaceShipOne's flight envelope, too, was like an airplane, albeit a fast one. But Rutan doesn't have the engineering experience working on spacecraft. The biggest challenges in reusable spacecraft have to deal with producing lightweight tanks, high ISP/high thrust engines, and dealing with the heat of reentry from orbit. Rutan has some experience in the first, but little in the others.

Rutan works in a (relatively) small shop with relatively few employees. NASA employs many thousands - and with reason, because, as we've already discussed, reaching orbit is a huge challenge. He's not as used to dealing with congressional mandates or the different branches of the government (although he has a little military contracting experience). Etc.

* He doesn't need experience! He's Rutan!

Idol worship is great, but in the real world, experience is extremely important.

* Well, NASA has a bad track record! Compare his track record to Rutan's!

NASA's track record is actually quite impressive. Despite what you hear from the media or certain geek-oriented websites, the Shuttle has the best safety record in the world for any manned launch system with a fair number of launches under its belt, at around 98% (if you doubt it, find a counterexample. I'll get you started: go to www.astronautix.com, and start browsing (***)). And this was with the fact that the Shuttle was only given about half of the budget it needed during development, and had to make huge design compromises. Rutan, on the other hand, nearly killed his test pilot by launching in high wind shear conditions, and launching before resolving the cause of wild rolls at rocket ignition. With just a small handful of flights. On a task that is incredibly easy compared to reaching orbit. Some view the rocketplane tourism industry as a disaster waiting to happen.

Well, what about budget? The Space Shuttle is, of course, everyone's favorite budgetary punching bag. Yet, despite the fact that its costs are higher due to cuts during the design and manufacture of the shuttle, it isn't that unreasonable. Launching cargo on the shuttle will cost you around 13,000$/kg, compared to around 10,000$ per kilogram for the ESA's Ariane V, and 7,000$/kg for Russian Proton rockets and Chinese Long March rockets (the latter two benefitting from cheap labour). So, while it's high, it's not unreasonable - and its safety record easily beats that of the other three). Rutan is looking to charge 200,000$ per person for a ride on a craft similar to SS1; assuming 100kg, that's 2,000$ per kilogram to for 1/6th the delta-V of orbit (and, as we discussed, there's geometric scaling up to orbital velocity, not linear). The "out of the ballpark figure" is SpaceShipOne.

What about development cost? The complexity of building something like SpaceShipOne is equivalent to perhaps 4 or 5 of the "common" research projects that NASA does on developing new materials, novel parts, etc. Yet, if you search NASA's website, you'll find almost 10,000 papers that contain the words "novel" alone. NASA funds a *LOT* of research with its budget. Its costs are actually quite reasonable.

* Well, I don't believe you, but I'm not going to present counterevidence!

Suit yourself.

* I think you have ulterior motives! You work for NASA, don't you?!?

No. I already discussed my motives at the top of the page, and am rather dismayed at people whose arguments have been simply to attack me. My interest in rocketry is purely just that - interest. Much of my knowlege of rocketry mostly comes from research during the development of two rocket simulators - a completed simple python simulator covering basic orbital mechanics in a single plane, and a complex full-system partially completed C++ system in which I left off somewhere between completing simulation of Kirchoff's Laws for the electrical system and calculating volume, mass, pressure, and intertial tensors for tanks.

* Only private rocketry can save the industry!

While I don't see it as being such a black-and-white distinction, I am hopeful for private rocketry bringing down launch prices. However, a few points on this issue:

A) Private industry already really runs the rocketry market. Most of NASA's craft, for example, are built and at least partially operated by large private players. So, if the intent is to criticize "government" involvement, the criticism needs to be on the contracting process.

B) If the intent is to promote completely privately owned and operated rocketry services with custom-developed rockets, those already exist, too - for example, Orbital's Pegasus and Sea Launch's Zenit. Both companies launch modified existing rockets, but there is nothing wrong with standing on the shoulders of giants. Note that the launch services aren't hugely profitable, nor are their costs dramatically lower than the other players for similar sized payloads.

C) If the intent is to promote "relatively" small companies with largely independently developed rockets, those exist, too - for example, there is good reason to be hopeful about SpaceX's new Falcon rocket series.

* Well, since you're such a pessimist, what good do you think came of all this? Do you think the X-prize was worthless?

No. In fact, I think it was wonderful. Now a whole new generation of geeks have caught space fever - and that can only be a good thing. Plus, the more excursions in aircraft-style vehicles (like SS1) to the edge of the atmosphere - and eventually perhaps even real orbital spacecraft - the more demand there is for superalloys, and the more people's interest in space get perked. We don't dedicate enough resources toward space, in my opinion, and so I'm very thankful for the publicity that the Ansari X-prize brought about.

* - There are benefits to pressure-stabilization of tanks (as you switch stresses from shear to tensile), but there is a limit to this effectiveness. In almost all cases, if you're using the pressure from your tank without turbopumps to power your engine, you're far beyond any benefits you could receive.

** - The other option is massively staged craft like Otrag; however, then you have a huge number of stage-separation risks and a vehicle of monstrous scale for a small payload (which raises labor cost concerns).

*** - NASA's best competitor on the safety record, for any craft with a large number of launches under it's belt, is the Soyuz. At a first glance, Soyuz may look safer: It has less fatalities. However, it also carries a lot less people - the shuttle averages 6 to 7, while Soyuz carries 2-3. Furthermore, NASA has relaunched the same astronauts more often than the Soviets/Russians relaunched Cosmonauts, increasing the odds of an individual person's death. When you look at craft failure rates, the Shuttle is better. When you factor in Soyuz ground-crew deaths, it's a no-contest. And yes, while the Soyuz hasn't had a manned death in decades, the craft continues to fail disturbingly often on unmanned launches; it seems they've just been rather lucky on which craft failed. Their near misses - such as a craft breaking through a frozen lake and sinking, or nearly rolling off a cliff - are also quite disturbing.