The most obvious answer is, of course, the Space Shuttle.
That’s right, I said it. No other vehicle in the history of manned space travel has lost two full crews during flight. The whole design is an abomination of committee-designed inefficiencies right from the start. Even Wernher von Braun thought it was an absolutely awful design.
没错。没有其他的人造航天器在飞行时像航天飞机这样损失了两班宇航员。整个设计一开始就被委员会嫌弃，连 Wernher von Braun 也认为这个设计很不好。
Start with the original design which was a delta-winged “space plane” at the top of a Saturn V like rocket, comparable to the X-20 DynoSoar project. This design, known as “on-axis” means that the shuttle sits at the top, safe from exhaust blasts and debris from launch, and all of its surfaces except the un-exposed tail-end are free to be designed however is needed for space-flight and re-entry.
最早的设计是在土星5号上类似火箭的三角翼“太空飞机”，与 X-20 DynoSoar 相当。这种被称为“同轴”的设计让航天飞机位于顶部，可以安全地防止排气爆破和发射时产生的碎屑，并且除未暴露的尾端外，其所有表面都可以自由设计，并且在太空飞行和再入大气时确实需要重新设计。
【译者注：Boeing X-20 Dyna-Soar，是用来执行军方任务的“太空飞机”。项目从1957年持续到1963年，后因为航天飞船项目开始而被取消。】
The DynoSoar had its issues though, and it was never launched in an actual rocket configuration. Several of the test pilots quit the program after the Air Force squabbled about which booster to use, including one named Neil Armstrong.
But the design was basically a good one. Von Braun had proposed a shuttle, but his design was more like what we see with Virgin Galactic’s Space Ship Two design, with a carrier that takes the shuttle into the stratosphere, where the main shuttle engines then ignite to carry it to orbit. In von Braun’s plan, the entire system was reusable.
That, however, didn’t make Congress and the companies like Boeing and Lockheed very happy. Re-usable meant they couldn’t sell you a new $100,000,000 rocket engine on every flight. Re-usable meant that assembly jobs were a one-time thing, and then congressional districts lost jobs and money. On the other hand, expendable meant the gravy train kept running.
So, as Apollo drew to a close, they pointed at the DynoSoar program as the model, and said, “This is the shuttle we want.”
But, then the committees got a hold of it. Simply getting people to and from space wasn’t enough. NASA wanted a way to get astronauts into space, but NASA had a limited budget. The Air Force wanted to get satellites to space and they had a much larger budget, so the Air Force got brought into the discussion. The Air Force, however, wanted cargo, not people. Other groups like the NRO (National Reconnaissance Office) and the like said they needed to put up satellites the size of school buses for doing satellite reconnaisance, i.e. spy satellites.
但是，然后委员会掌握了它。仅仅让人们进出太空还远远不够。 NASA想要一种将宇航员送入太空的方法，但是NASA的预算有限。空军希望将卫星送上太空，而且预算要大得多，因此空军参与了讨论。但是，空军要的是货物，而不是人员。 NRO（国家侦察局）等其他组织则表示，他们需要安装像校车大小的卫星来进行卫星侦察，即间谍卫星。
That means that the small, manageable X-20 sized shuttle was tossed out in favor of the giant cargo space of what came to be the Space Shuttle. But, more cargo space means more “space-plane” to house it. That raises the mass of the shuttle. More mass means bigger wings to land it, and more powerful rockets to lift it.
The whole thing became an almost comical cycle. With the shuttle needing that much lift, the engines became so expensive and time consuming to build that even Lockheed said they could no longer be expendable, and had to be re-used to make the shuttle even vaguely affordable. That meant the engines had to be recovered, which meant they had to be part of the shuttle. With engines mounted in the tail of the shuttle, it couldn’t sit on top of the stack any more.
And that’s vitally important. In the event of an accident, a capsule mounted at the top of a rocket stack has a chance to escape. When you’re strapped side-by-side to a giant bomb, as Challenger showed, there’s nowhere to go when something goes wrong.
Worse, strapping to the side of the stack (off-axis for those who are picky about such terms) leads to all kinds of other compromises. The piping for fuel must now go directly through the heat shields that cover the bottom of the shuttle, creating headaches in just how to ensure that the covering is safe for a 3500 degree re-entry.
The shuttle is now in the path of any debris breaking away from the fuel tank. (This was opposite what you saw on the Saturn V launches, where great sheets of ice would peel away from the cryogenic tanks on the rocket, falling harmlessly along the surface of the rocket.) This meant the fuel tank had to have extreme insulation to prevent ice from forming, because the ice would now crash into the shuttle, cutting through its fragile tiles that protected it from reentry. So, thick foam insulation was placed over the entire tank, adding more weight, and more headaches, because the slightest flaw in how the foam bonded to the surface meant chunks of foam peeling off the tank and striking the shuttle. This was what would lead to the destruction of Columbia.
The main weight at launch is in the expendable fuel tank, carrying nearly 1.6 million pounds of hydrogen and oxygen, 20 times the mass of the shuttle itself. If you think of the engines on the shuttle as lifting that weight, then imagine that the center of gravity is nearly 20 feet off-line from the lift of the engines that are lifting it. That meant the shuttle, and the fuel tank, had to have reinforced structures to support all that off-axis thrust. And that meant even more weight to be lifted to orbit in the shuttle, and nearly to orbit in the fuel tank.
The numbers got so bad, the only way to solve the problem was to strap on giant, solid-fuel rocket boosters to get this travesty off the pad. But solid fuel has a problem. It has exactly two settings, “Not Lit” and “Full Throttle”. Now, to be technical, the SRBs had fuel mixes pre-loaded that created a thrust profile to match the flight events like Max-Q, and other adjustments, however, this was not an exact science and usually there were 2–3 second windows where the thrust profile would change, and the shuttle main engines would have to compensate.
Additionally, the SRBs were prone to “thrust oscillation” (think bouncing up and down like a pogo stick) events, vibration along the direction of travel, which tended to shake everything violently. Again, these were also there to lift the 1.6M pounds (800 tons) of fuel in the external tank. These boosters were also mounted off-axis, meaning they were also lifting on a lever arm 20 feet off-axis from the center of mass.
This meant more reinforcement and more weight. In the end, the STS (Space Transit System) was burning about 3.8 million pounds of fuel (1.6M in the external tank, plus 1.1 million in each booster) to lift the 200,000 pound shuttle into orbit. That’s a 5% fuel to weight ratio, which is actually rather good, but remember, the shuttle wasn’t actually the cargo.
The cargo was whatever satellite or bit of the space station they were trying to launch into orbit.
The maximum weight that a shuttle could take to LEO (low-earth orbit) was 65,000 pounds. So, the shuttle burned 3,800,000 pounds of fuel to get a maximum of 65,000 pounds to orbit. The average cost per flight on a shuttle was about $450,000,000 per mission. That’s a cost per pound of $6923 and change to low earth orbit, and the shuttle almost never flew at maximum cargo weight.
To put it in perspective, while the Saturn V was more expensive (about $1.16B per flight in modern dollars) it could carry 261,000 pounds of cargo to orbit, a price per pound of about $4445. And they were throwing the entire rocket away on each flight.
从另一个角度来看，虽然土星V较贵（按现代美元计算，每次飞行约$ 1.16B），但它可以运载261,000磅的货物进入轨道，每磅的价格约$ 4445。而且他们在每次飞行中都将整个火箭扔掉了。
To put that into perspective, the entire ISS weighs about 440 tons right now. The Saturn V could have lofted that in four flights, and almost could do it in three. (130.5 tons per flight.) The Space Shuttle had 36 designated ISS missions to loft parts of the station, and several parts were added by Russia, Japan, and even the ESA without the help of NASA. In fact, the current mass includes several portions brought up by SpaceX and Orbital as well.
All this doesn’t even mention the fragility of the tile system on the shuttle. Every mission required the inspection of every tile, and the replacement of hundreds if not thousands that were damaged on each flight.
The Main Engines had to be totally refurbished after every flight, the entire turbopump assembly removed, and rebuilt. The idea that was sold….
…was nothing at all like what was actually delivered.
When added to the Shuttle’s dismal safety record, the broken promises of re-usability — Congress was promised it would fly, “once a week with only minor refurbishment between flights” — there’s no choice about the worst rocket design ever put into use, and that’s the Space Shuttle.
(Edited to correct the term “Pogo” for SRBs — which is reserved for on-axis oscillations in liquid fueled rockets — to the correct term, which is “Thrust Oscillation”.)