why it’s sometimes best not to compromise

February 27, 2013

shuttle booster concept art

Pretty much everyone with a TV on Earth knows how the space shuttle looks. Shaped like a fat, boxy plane strapped to a giant orange fuel tank with two rockets on the side, it’s been shown on countless news channels for decades. And it completely different from what how it should look, the original design meant for something more akin to an advanced version of the Virgin Galactic space planes. What you see above is the original North American Rockwell/General Dynamics concept art of the shuttle’s booster, a sleek, rocket-assisted plane that would haul the shuttle on its back high into the atmosphere and releasing it into orbit. Every part of the vehicle was meant to be reusable to lower launch costs, and it would’ve looked like something from the future even if we compared it to today’s clean, modern designs. So what went wrong? Why did we end up in orbit on a fat, boxy plane that didn’t really work as promised and fell short of its lofty goals?

In the same post on Beyond Apollo featuring the artwork which offered Rockwell’s and General Dynamics’ dreams of exploring the final frontier, there’s a very blunt answer to this question. It’s all about money in the short term. Because the shuttle was meant to be developed first, corners were cut and the plan changed to strap it to a big, orange foam tank just for a little while, until a more fitting booster vehicle could be built. But once the launch stack was operational and got to orbit, the need for reusable boosters now fell away and the agency was content spending what it did to keep the shuttle flying. A short term compromise to save money now quickly turned into a permanent solution to cash flow problems despite the fact that it scaled back a necessarily costly and ambitious plan. The engineers were tasked with building a space-faring Ferrari meant to be a stepping stone to building and supporting massive space stations that would launch missions to other worlds. Then they were then told to build a pickup truck instead.

There’s a saying in the technology world. You can get it made well, quickly, or cheaply; now pick which two. Sometimes quick and cheap will not work. Sometimes you have to bite the bullet, stick to the plan, and refuse to compromise to get what you wanted to build done the way it should’ve been built, not the way that the bean counters and politicians would like it wrapped up because the quants addicted to dashboards and keeping budgets as low as possible would rather you sit still and not build anything, especially something ambitious. I can certainly understand the need for fiscal responsibility and keeping an eye on how things are going, but I also realize that a big idea requires a big investment and keeping your eye on the prize over the long term rather than glued to every cent and productivity metric that day. And when I look at the photos of what might have been if the shuttle vision was followed through just as NASA and its contractors planned in the beginning stages of the project, all that comes to mind is "what a damn shame…"

Share
  • http://wading-in.net/walkabout Al Denelsbeck

    Another crucial factor in all of it is that new technology does not have a budget – there’s no reasonable way to predict what something new will actually end up costing, yet Congress (and every common pundit) wanted to whine and moan about NASA not working within their budget. So there’s another algorithm to be recognized: select either the goal, or how much you want to spend. You might end up with both only by luck.

    Add in all of the special interests that were accommodated, as well as the military’s requirements for payloads (which supposedly altered the initial dimensions,) and you have a lot of fingers in the pie. The committee’s camel we ended up with really wasn’t bad in consideration.

    I’m always a little wary of initial designs, since they rarely come to pass. A stacked pair of aircraft designed to both launch in the higher machs and land with conventional turbines is daunting criteria. The lifting body program came about because, very simply, “you don’t put wings on the nose of a rocket” – stability is next to impossible. A standard winged craft couldn’t really have a rocket booster behind it, so something that could fit within a nose cone was tried. Eventually, the design of the shuttle booster with wings at the rear of the assembly was selected. I’m trying to imagine all of the drag/stall/pressure issues with stacked or staggered wings on the booster design pictured in the article.

  • Paul451

    NASA’s problem wasn’t that they were told to build a pick-up truck, it was that when given a budget that would only afford a pickup truck, kept trying to build a Ferrari. A 100 ton dry-mass space plane? Having never built any type of space-plane before. Dumb.

    Many of the things that NASA folk still blame on others are their own fault. For example, they continued to believe that after Apollo they would be “rewarded” with an even greater budget. Ignoring that even Kennedy mocked the “Apollo Applications” proposals when meeting Jim Webb, saying that he wasn’t interested in space flight, he was only interested in showing up the Russians (by building a bigger rocket, demonstrated by doing something the Russians couldn’t.) Also ignoring that Congress had been chipping away at Apollo’s budget, cancelling hardware, even cancelling missions. Dumb.

    Similarly, many like Al blame the Shuttle’s failure on the USAF wanting to use the Shuttle for launching and retrieving spy-sats, requiring massive cross-range ability in order to launch and land in one orbit, and a large payload bay. Except that the proposal didn’t come from the Air Force, it came from NASA management trying to attach itself to the DoD’s fat teat, thinking that if the much larger Shuttle had a military customer, then Congress would give them a bigger budget again. Dumb.

  • gfish3000

    Actually, NASA’s hopes for the shuttle were not nearly as dumb as you seem to think. Kennedy was dead long before the shuttle hit the drawing board. Congress wanted to do new things rather than go to the Moon and NASA thought it would give it something new, and with all the talk about building powerful spy satellites and delivering them to orbit, as well as having a vehicle that could capture, service, and repair them, of course NASA said it could build a vehicle to do it.

    The problem is that none of the big, bold talk ever cam to fruition. Everyone was so focused on what’s going on this fiscal year or that election cycle that there was no plan beyond it and longer term projects were scrubbed in the penny-wise, pound-foolish way governments usually try to save money. NASA asked for a Ferrari budget and presented a Ferrari. They were told that the Ferrari would be awesome, but here’s a pickup truck budget and since they landed on the Moon, they should figure out how to make it work.

    And yes, why not build a new 100 ton space plane from scratch? The shuttle itself ended up weighing about 72 tons so it wasn’t all that far off the mark. Likewise, how else do you build something new than by actually trying to build it? SpaceX also never built a rocket before and were told that there’s no way they’d be able to build a new type of rocket from scratch. Now they’re supplying the ISS. They tried, solved problems, found new ways of doing things, and kept trying until they succeeded. But of course they had an advantage over NASA: an owner with deep pockets and a long term vision which doesn’t depend on flaky electoral politics.

  • Paul451

    NASA’s hopes for the shuttle were not nearly as dumb

    Hopes for the shuttle? No. Design and execution, yes. Horribly dumb. The general idea of a reusable crew vehicle still makes sense. The idea of a reusable launcher still makes sense. The Shuttle did not make sense. Especially not on the budget they had (and anyone in 1971 who thought they were getting a bigger Apollo-style budget was… dumb.)

    Kennedy was dead

    My point about Kennedy was that even the very guy who issued the Moon challenge, the very example that people wave like a banner when they want to talk about “Presidential leadership” on NASA, even that guy had no interest beyond a narrow goal of beating the Russians. With him gone, and the Russians gone quiet, what on Earth would make anyone think that NASA was going to suddenly start getting monster budgets again?

    The shuttle itself ended up weighing about 72 tons so it wasn’t all that far off the mark.

    I had 104 tons for the Columbia orbiter, about 130 tons wet. That was my “100 ton space plane” reference.

    Likewise, how else do you build something new than by actually trying to build it?

    By learning how to build it. Step-wise. Learning your craft. Exactly as NASA did from it’s formation up to Apollo. And before that, through various military development. And before that, Von Braun’s V2 program.

    But with the shuttle they had only a tiny amount of X-plane research. Small piloted rocket-gliders. To jump from that to a 100 ton space plane (or 72 tons if that’s the case) in one step and expect it not to be a disaster is just wishful thinking. Frankly it’s shocking that the thing flew at all. Those old Apollo guys were something else.

    Similarly…

    SpaceX also never built a rocket before […] Now they’re supplying the ISS.

    They started with a one-engine Falcon 1. They then developed the Falcon 5 design, but felt comfortable jumping straight to Falcon 9. (IMO, some of their problems with the F9 engines probably comes from skipping this step.) Likewise they built their cargo capsule as conservatively as possible, with parachutes and ocean splash-down and borrowing from an improved heat-shield previously developed by NASA.

    Now they are on the fourth version of their engine, the second version of the Falcon9 design, and are working on the next iteration of their launcher (Falcon Heavy). And are upgrading their cargo capsule to carry crew. And have a first-stage test-vehicle to explore reusability (Grasshopper).

    Imagine if they’d started out jumping straight to their proposed Falcon XX fully reusable with 9 Merlin 2-sized engines. Do you think they’d have the success they are having? Do you think they would have saved money by skipping all the intermediate steps?

    Because that’s what NASA did. Skipping over all the steps between X-Planes and Apollo capsules, and building a giant space plane. And it’s what they are doing with SLS. And it’s what they are doing with Orion. And it’s what they did with JWST. And it seems to be what they’ve done with virtually every major program since the end of Apollo.

    They say the definition of madness is doing the same thing over and over and expecting a different result. By that definition, NASA isn’t dumb, it’s just nuts.

  • gfish3000

    I doubt that NASA was really planning the original shuttle design all in one go. The problem is that to save money they decided to go fast and cheap and again, I believe you’re applying 20/20 hindsight to a much more complex situation than you present. You said that a reusable, human-rated vehicle makes sense but fault NASA for thinking that it could cash on the goodwill its Moon missions generated when the military was talking about space planes and aerospace companies in Europe were trying to design them. It’s Monday morning quarterbacking.

  • Paul451

    “It’s Monday morning quarterbacking.”

    Okay, fine. But you know what they did on Tuesday? The tried to follow up the Shuttle with a giant SSTO space plane, built in one version, having never first shown that SSTO’s were even technologically possible, let alone giant manned reusable SSTO space planes.

    Then guess what they did on Wednesday? Tried to build a giant 12 man space station. In one step.

    And guess what they did on Thursday…

    I wouldn’t mind if they did this once and learned their lesson, but they do it in nearly every single major program. And they’re doing it now with SLS/Orion and with JWST. And when you call them on it, they blame funding, or lack of leadership, or other agencies, or Congress, or anyone but themselves and their pathological unwillingness to learn from their mistakes.

  • gfish3000

    Again, you’re making the matters too black and white. NASA is treated in very paradoxical ways for Congress. First and foremost, they’re an easy source of highly paid government STEM jobs which is why Texas, Florida, Alabama, and California are all about keeping NASA programs going and why senators from these states fight tooth and nail to keep money flowing and projects built. What matters is that the agency has work to do.

    However, when it comes to actual work, they don’t want to spend the money to build it right while expecting that NASA somehow manages to figure things out quickly, easily, and cheaply because hey, these guys went to the Moon, the can figure out [insert project name and roadblocks here]. NASA is a political entity which means that what you’re seeing as a refusal to learn from its own mistakes, is usually pressure from Congress to keep cranking out program after program and keep cash and jobs flowing to their districts.

    Were NASA a company or even a GSE, it would behave far differently than it does today because it could find better places to built and test spacecraft, and come up with phased concepts that get it where it needs to go like SpaceX and Bigelow. If Elon Musk had Congress breathing down his neck about getting that Falcon 9 into orbit ASAP and build his facility where a couple of Congress members can later brag about bringing jobs from Washington, I guarantee you that things would have went a lot differently for him.

  • http://wading-in.net/walkabout Al Denelsbeck

    The X-plane projects weren’t a handful of piloted rocket gliders, they were part of a multi-decade program exploring the dynamics of drag, pressure, high-altitude engine operation, and on and on, incorporating stuff learned from the X-15, the YF-12/SR-71, the XB-70 Valkyrie, the entire lifting body program, scramjets… those who have actually looked at the program find it rather hard to characterize in such dismissive terms, nor to understand how you possibly end up with the “huge leap to a giant space plane.” And since the shuttle worked as intended for three decades, it’s difficult to fathom your point.

    As for the comment about building a 12-man space station in “one step,” you could probably stand to Google “Skylab,” and the extended shuttle programs with the various science labs onboard. Then there’s this completely unwarranted idea that they incorporate information gained through the Russian stations and build on that. Crazy, I know – they should have been reinventing the wheel. And again, since it’s still operating as intended, your point remains indistinct.

    Let me forestall further accusations of sides, or of sucking up to NASA, since I don’t think in terms of allegiances, to either organizations or ideas. I simply find it more useful to view as much info as possible as objectively as I can, without trying to support preconceived notions.

  • Paul451

    nor to understand how you possibly end up with the “huge leap to a giant space plane.”

    Fine, then what were the space-planes flown into orbit between the end of Apollo and the first Shuttle flight in 1981? I must have missed them.

    There was early military development through the ’50s following on from the Nazi work, then pre-Mercury animal flights, then Mercury, then Gemini, then Apollo. Then a decade gap… Then a 100 ton space plane. You don’t see what’s wrong with this picture?

    And did they learn their lesson? Did they adopt any of the many proposals they had to incrementally upgrade the next version of the Shuttle, done while the researchers and engineers were fresh from the first one? Nope. They tried to develop a giant SSTO space-plane (in one step, having never shown that even a small expendable SSTO rocket was technologically possible).

    The only reason the shuttle flew at all, IMO, was because those Apollo guys were the culmination of decades of incremental development. They could do rocket design in their sleep. In the same way that the SpaceX engineers have developed their knowledge incrementally; they learned their craft. And that’s what NASA stopped doing in their major programs since the end of Apollo. And they make that same mistake over and over, and wonder why it keeps going wrong.

    And since the shuttle worked as intended for three decades, it’s difficult to fathom your point.

    A reliable low-cost fully-reusable space truck? The shuttle flew. It most certainly didn’t “work as intended”. It consumed around $3b/yr for up to 6 flights per year (in its peak year, 1985, I think).

    Then there’s this completely unwarranted idea that they incorporate information gained through the Russian stations and build on that.

    That was in the late ’90s, when NASA was struggling to build even the cut-down four-man Alpha proposal without help. (In 1996, Clinton ordered a major “What can you actually build, not what you wish you could build” review.) By then, they had spent a decade burning through the entire original Freedom Space Station budget without a single piece of flyable hardware, had cut back the design several times, and finally had to ask the Russians how to do it.

    And yes, Russia’s 30 years of incremental space-station experience helped. Imagine that. What possible lesson could we learn from that…hmmm?

  • Paul451

    The reason NASA centres are spread out across the US is because that’s how Jim Webb created NASA in order to secure Congressional support for Apollo during the ’60s. Even then, they lost funding for entire missions, for major pieces of hardware.

    NASA’s problems with Congress are not new. Their problems with funding are not new. So why do they now repeat the same mistakes over and over and act surprised each time. That’s like blaming gravity for falling off a cliff. After the tenth time, I think I can say “maybe you’re just clumsy?”

    NASA’s culture was considered a major factor in the Feynman review following the Challenger accident. It was considered a major factor in the Columbia accident. It was clearly a factor in the failed Constellation project, in the $9b JWST project, and now with SLS.

  • http://wading-in.net/walkabout Al Denelsbeck

    Fine, then what were the space-planes flown into orbit between the end of Apollo and the first Shuttle flight in 1981? I must have missed them.

    Your entire argument is that they did not fly a space plane before they flew a space plane, is that correct?

    Now, please don’t make me provide links to the X-plane programs you’ve been frothing about and the numerous tests of airfoil performance in “edge-of-space” conditions, which is precisely what you need to know for the crucial difference between the moon programs and a reusable entry vehicle.

    They tried to develop a giant SSTO space-plane (in one step, having never shown that even a small expendable SSTO rocket was technologically possible).

    I could have sworn you mentioned the Mercury and Gemini programs, which were all SSTO. I’m guessing you also aren’t aware that the SRBs used came from suborbital, but extra-atmospheric, tried-and-true ICBM platforms?

    The only reason the shuttle flew at all, IMO, was because those Apollo guys were the culmination of decades of incremental development. They could do rocket design in their sleep.

    I can only guess that you think the NASA of the shuttle design is a different organization than the NASA of the Apollo design. Otherwise your point makes no sense.

    A reliable low-cost fully-reusable space truck? The shuttle flew. It most certainly didn’t “work as intended”. It consumed around $3b/yr for up to 6 flights per year (in its peak year, 1985 I think, typically it
    was more like 4/yr).

    Yes, I addressed that point in my original comment. The design proposal was not what resulted, both in terms of budget and artists’ concepts. Welcome to reality. You are invited to point to a program that fell exactly as proposed. Until then you’re only repeating the pundits’ vapid excoriations.

    By then, they had spent a decade burning through the entire original Freedom Space Station budget without a single piece of flyable hardware, had cut back the design several times, and finally had to ask the Russians how to do it.

    Considering that the budget was for design, not construction, this is exactly what you should have expected – provided, of course, that you weren’t simply trying to feed your pet rant. As for cutting back the design, how many times do you have to hear comments about Congressional funding before you disintegrate from the effort of ignoring them?

    And yes, Russia’s 30 years of incremental space-station experience helped. Imagine that. What possible lesson could we learn from that…hmmm?

    I know of several lessons that I take from it. I am also quite sure you’re not going to mention any of them, however, since not one of them translates to “NASA stupid bad durrrr…”

    And I hope that clearly expresses that, unless you’re actually capable of making sense, we’re done here. You’ve had your chance.

  • Paul451

    Your entire argument is that they did not fly a space plane before they flew a space plane, is that correct?

    [Sigh] You can’t be that thick. You can’t have really read my comments – talking about the importance of incremental development, using examples of pre-Apollo NASA, the more recently SpaceX development, and even the Russian space-station program – and think that I’m really saying something so stupid.

    I have no doubt that had NASA developed space-planes over several stages during the ’70s, culminating with a full Shuttle-scale space-plane – designed and built by people with not only Mercury/Gemini/Apollo under their belt but also two or three prior space-planes – they would have produced something vastly superior, safer and lower cost than the actual Shuttle. And had they continued that development (with funding freed up by not having to spend over half a billion dollars to launch each shuttle) through the ’80s and ’90s, the vehicle they’d be flying today would be worthy of a thirty year program.

    But it was madness to spend a decade trying to go straight to the end design, without learning about this entirely new style of vehicle (side-mounted orbiter, mixed solid and liquid rockets, reusable parts, massive space-plane and its reentry and servicing requirements) in smaller steps first. It neither saved time nor saved money.

    Imagine if that was how they had done Apollo. After Kennedy’s challenge, they scrapped the existing Mercury program, skipped Gemini entirely, and instead spend 8 years developing just the Saturn V/Apollo stack in one go. If the first flight after 1961 was the manned Apollo 8 mission. Because that is essentially what the first Shuttle launch in 1981 was. Do you think Apollo would have made the deadline? Do you think they would have saved money or time to skip Mercury, Gemini, the Titan II, Saturn 1 & 1b launches, etc etc?

    This is what they did instead: http://upload.wikimedia.org/wikipedia/commons/6/67/USAF_ICBM_and_NASA_Launch_Vehicle_Flight_Test_Successes_and_Failures_%28highlighted%29.png

    Put it another way, the idea of an unmanned version of the Shuttle-stack (Shuttle-C) was only explored 3 years after they launched the manned version. Likewise, the small HL-20 space-plane started development in 1989. It, unlike the shuttle, was a direct descendent of the X-plane lifting body work in the ’60s. It would have made an excellent first or second step towards the Shuttle during the early ’70s. Can’t you see how back-to-front that all was?

    [There have always been people who believed as I do. The Delta Clipper team is an example, deliberately choosing the simplest first step with DC-X. What Jerry Pournelle called “fly a little, break a little”. At the time, critics within NASA, even within McDonnell Douglas itself, said that nothing could be learned from a small demonstrator and that they were just wasting time/money not developing the full-scale version, as they did in the X-30 program.

    According to insiders, the problem is that NASA developed a post-Apollo culture where “paper studies” were considered superior to actual flight-tests. This was seen in the use of a computer model to satisfy themselves that Columbia wasn’t in danger after the foam-strike, rather than fire a chunk of foam at a piece of RCC-tile as they did only after the loss of crew. And before that, the Challenger loss, trusting theoretical-specifications over just freezing a piece of O-ring material, as Feynman famously did during his post-accident review.]

    I could have sworn you mentioned the Mercury and Gemini programs, which were all SSTO.

    Sorry, I shouldn’t have used aerospace terminology. SSTO stands for “Single-stage-to-orbit”. http://en.wikipedia.org/wiki/Single-stage-to-orbit
    It means a launcher where every part of the vehicle travels from Earth all the way into orbit.

    By contrast, Mercury used the two stage Atlas-D launcher to put the Mercury capsule and service module into an orbit insertion trajectory, with the service module rocket doing the final orbital burn and de-orbit, before it too was jettisoned. So three separate expendable stages, plus the capsule, making it a four-part vehicle. (Even the sub-orbital Mercury flights used a main-booster (Redstone) plus the service module plus the capsule.)

    Likewise Gemini, with the two-stage Titan II and Gemini’s own service module rocket.