Bye Bye NASA

Miscellaneous and Personal Stuff I Must Share
41

That’s not what I said. Tuckerfan argued that NASA should just take Rutan’s word for it that he knows what he’s doing, rather than have him prove it by constantly providing reports and documentation. Nobody deserves this level of trust, especially someone who has never done it before.

Also, the discussion is whether NASA should award the CEV major contract to Rutan, and even if he can provide the documentation, I’d say no. A smaller proof-of-concept project, perhaps, but not an entire multi-billion dollar project which will become the core of NASA’s human spaceflight program for the next 20 years. It’s not a responsible use of taxpayers’ money.

The CEV project itself is the problem. They’ve already made the decision to rely on existing launch systems and build a manned capsule system with existing technologies. If that’s what they want to do, then the best people to do it is the people who are in posession of the existing technologies - Boeing, Northrop and L-M. But I’d much rather see a dozen $100-million projects investigating new launch systems, including reusable spacecraft - and Rutan would surely deserve one of those contracts.

I seem to recall you telling us about a Japanese auto parts maker whose transmission lasted longer than the identical design made in the US. The only difference was the difference in manufacturing tolerances. But they both worked properly right out of the box, didn’t they? How can you tell them apart initially, except by going over the records of how tolerances are measured in each plant, what the measured values are, etc.?

They did multiple tests, and some of them showed a flaw. They made a wrong decision on which test results to trust, and on whether to delay the launch with additional tests.

I’m not following you here. We do use as much off-the-shelf components and legacy components as possible. And we keep using them on many different missions because that saves us the cost of testing new components. That’s exactly why we often use older technologies where newer (untested) ones may perform better.

As for buying parts on eBay, I’m pretty sure that was for ground support computers. I myself have worked on satellite operations, and we used a temperamental 15-year old mainframe. I don’t see how you can avoid such a situation - the computer was new when the satellite was launched, and it was a widely available and supported off-the-shelf computer system. Even now it’s supported by the manufacturer, but they often have to hunt around for replacement parts. Porting the software to a newer computer would be highly dangerous and expensive - we would have to test it thoroughly, because a wrong sequence of commands can damage or kill the satellite. Whereas an occasional crash or downtime is not a big deal.

42

This issue isn’t so much his (or rather, Scaled Composites and their partners in the t/Space Consortium) lack of “experience with manned orbital flight” as it is their lack of experience in being the prime contractor on such a massive project. This is an enormous undertaking–on the same scale as the Apollo program or the Manhattan Project–and for better or worse, one major component of that is “systems engineering”; i.e. coordination, oversight, and the dreaded documentation. (And quite aside from the argument of whether the amount of paperwork required by NASA standards is necessary or not, it is the status quo with NASA and isn’t going to change…so any prime on the CEV and M2M has to be prepared to deal with it.)

If Burt Rutan wants to do things his own way, with minimal paperwork and documenation, I say good on ya, mate. But making a suborbital spaceplane, for which the prime fabrication and assembly is done by one company (the engines for SS1 and the control systems were subcontracted) is on an entirely different scale from developing a multistage, transorbital space-going vessel. It isn’t a linear increase in difficulty; every individual system multiplies the complexity, and those constraints and requirements have to be communicated and factored up and down the various chains. It isn’t as if each system is independent and can be plugged in seperately; every factor is a trade off between a number of issues, including weight, packaging, efficiency, reliability and redunancy, cost, et cetera.

I’m not saying that Rutan (or rather, his “team”) couldn’t build such a system, though I think that his fabled brilliance and self-promoted invunerability would suffer some hits as he goes through the same learning curve that NASA and its present contractors did in the 'Sixties, and I’m not going to argue that the mass of paperwork and documenation required by NASA is all necessary, but bear in mind that these documentation procedures weren’t (mostly) developed just by some desk-sitter with a hatred for leafy green things but rather in response to a need for traceability. It’s quite possible–converging on a locus of certainty–that these procedures could be significantly streamlined, but they can’t be dismissed. And whoever is going to manage this project is going to either have to cope with the current requirements of NASA or devote the effort to making a convincing argument for change (which, given NASA’s current conservativism, is a Sysiphean task if ever there was one).

This biggest problem, actually, with NASA isn’t the mass of paperwork, but rather their risk-adverse/risk-incognizant attitudes which have paradoxically both slowed program development AND lead to the Shuttle disasters. The cultural hazard-blindness that led to ignoring the (obvious and recognized) problems that led to the Challenger explosion are the same that resulting in Columbia’s catastrophic failure. Unfortunately, the root cause of part of this problem–the detrimental impact upon NASA as a whole and individual engineers and technicians who are even so bold as to acknowledge a problem–extends beyond NASA. They are held accountable for every failure or holdup since the days of William Proxmire, even if those delays are due to unforseen (and often unforseeable) technical hurdles. And so, it is better, career-wise and budget-wise, to remain head-in-sand. (Witness Linda Ham’s self-serving evasions as an example of this pervasive culture within NASA.)

A couple of other comments:

Tuckerfan, try not to be too jealous of scr4. The space program (and the aerospace industry in general) looks a lot sexier from outside than it does from within. Honestly (and this will no doubt come as a shock to you) the automotive industry is better coordinated, better run, and often more highly engineered and refined than the products that come out of the aerospace industry. You would be astonished at how ad hoc many of the engineering solutions in aerospace are. There’s some really advanced, cutting-edge materials science, controls theory, and computational fluids analysis done, for sure, but the vast bulk of the work is (as with any engineering projects) minutae and really boring details, up to and including arguments over nomenclature and what kind of formatting to use on the drawings. (I swear, if I get stuck in another argument over on-sheet vs. off-sheet bills, I’m going to draw blood.)

It also doesn’t bring in the babes. “Hey baby, come home with me. I’m a rocket scientist!” Nope, doesn’t work. :frowning:

Also, regarding the funding of CEV and M2M: it’s not underfunded in the way in which the Shuttle was underfunded–that is, trying to cram every possible use into one reusable launch vehicle–but rather, underfunded in the sense that there isn’t enough money to even get this thing off the ground. The budget I’ve seen for this thing over a decade is less, in adjusted dollars, than what NASA spend in one peak year during Apollo. Mind you, we will be able to make use of some legacy information, so it’s not quite starting at the same point, but nonetheless, there just isn’t the kind of funding committment to guarantee progress. And from what we’ve seen lately, it doesn’t appear that Congress will, barring some kind of shift in public opinion (such as a newly energized space race), be willing to authorize additional monies they way they did during the Johnson Administration. My prognosis is that this program will die a quiet, strangled death with the coming of the next Presidential Administration.

This doesn’t bar private industry from engaging in their own development, of course, and hats off to Rutan or anyone else who can put together a group of investors and engineering concerns to make it happen. But I’ve got a nasty feeling that we’re going to be giving over the dominance of space to somebody in Asia. The (monetary) payoff for such a venture is a quarter century (or more) out, and American businesses barely look beyond the next fiscal quarter.

But I’ve no crystal ball, and my pessimistic view may be entirely bent. I sure hope so.

Stranger

43

No funding? But what about all these committed goals by our fearless leader to put a base on the moon and a man on Mars?

That’s an attack on Bush, yes, but honestly, if we’re not going to give the proper funding to NASA and these projects, why do we set these sorts of goals? I’m not nearly as well informed as probably anybody in this thread, but from what I’m reading, it doesn’t sound like the CEV or M2M will come about in the near future, mostly because there isn’t the money. What is the point, then?

44

I never said that Rutan should be given a free pass as far as all the paperwork goes, but if Rutan (and many others both inside and outside NASA) say there’s too much paperwork, then it seems to me a priority at NASA should be ditching that paperwork. After all, it’s most likely not being read by some high graduate, but by some degreed professional who’s getting a fat salary. A salary which could be better spent on research.

And it is a responsible use of taxpayers money to hand it over to folks with a long history of “creative accounting?”

Agreed.

You don’t need all that paperwork to find that out, however. All you need is a copy of the original blueprints, suitable tools to measure the parts (calipers, micrometers, etc.) and someone who knows how to use them. Crikey, we don’t need the documentation given to the Pharoh’s architect to know how the pyramids were built.

More proof that when it get’s right down to it that the paperwork doesn’t matter, it’s what’s between the ears of the guy calling the shots that matters.

IIRC, some of it did actually fly on one or more of the shuttles.

Stranger, large scale systems integration isn’t exactly the bugaboo that it once was, after all we do a variation of it every time we connect to the net. In the case of the Apollo program, it was something that had never been tried before on such a scale. Figuring out how to do it the first time is the hard part. Oh, and most of my best work is done ad hoc (tain’t nuttin’ like whipping a rabbit out of your ass and having everyone call you “Scotty” :D), and I babes or not, I bet it pays better.

As for the future, well, Bigelow has pledged $500 million towards it, Paul Allen rummaged around in his sofa cushions and pulled out $20 million for it, and Branson’s wagering an unspecified amount towards it, and that’s just some of the players involved. It’s just a matter of regulatory action, since the laws haven’t yet caught up with the technology, Congress being tied up with more pressing matters like keeping feeding tubes in a corpse. :rolleyes: And because of that, it’s rumored that Branson’s going to have to with someone other than Rutan for VirginGalactic, since there’s apparently some legal issues with Branson owning the gear and him being a Brit. (Yeah, like Tony Blair’s sitting there in Number 10 Downing Street, rubbing his mitts together going, “Once I get my hands on Rutan’s suborbital technology, I shall rule the world!!! Bwahahahaha!”)

45

I didn’t catch this on preview, but since I deal directly with this issue I’d like to make special note of it. In recent years we’ve been under pressure to use COTS (commercial off the shelf) hardware, both in new development programs (like GMD and KEI) and on refurbishments of legacy equipment (MM, PK), the idea being that we’ll both save cost over mil-spec (effectively, custom requirements) components and have a wide background of empirical data. When you say mil-spec people think of $500 toilet seats and $600 hammers, and while part of that was sheer grift by contractors, some was also due to the estenstive testing and documentation requirements for a mil-spec hammer versus one you buy at the Ace store. They might be the exact same hammer, but the mil-spec one has been hardness tested, or has its handle selected from a hand-inspected lot, and so forth.

This is fine when it comes to common hardware like threaded fasterers and rivets (which are as safety critical in many commercial applications as they are in military ones); however, when it comes to more complex or less commonly used parts (actuators, controllers, sensors, et cetera) COTS is often either not sufficiently tested, lacks a true database of reliability (“Ever had one of these break?” I ask a vendor; “Nobody’s ever complained,” he replies…:rolleyes: ), or just plain doesn’t fit and requires retrofit and compromise of other components and structures in order to make it work. Mil-spec doesn’t guarantee quality, of course; oversights and deceptions happen. But at least you know what you’re specifying, and some purchasing agent can’t just go off and get some Chinese-made POS fastener that is “equivilent” (according to their say-so).

OTOH, you’ll notice that on the Shuttle and in the background they have commercial laptop computers velcro’d in all over the place–a situation where COTS is actually superior to the spec’d hardware though this is due more to the rate of technological advance. That highlights a valid issue, though; we shouldn’t be looking for one tool to do all jobs for the next 20-30 years. We should, as scr4 suggests, be spending moderate amounts of money on various concepts to develop different tools for different purposes (leapfrogging on the legacy of the previous), each being in planned use for no more than a decade (or whatever period of obsolescence seems appropriate), so that we are not locked into using a compromised or outdated design indefinitely.

NASA has the SwissChamp approach; one big tool for all applications. If you’ve ever used a SwissChamp knife, you know how awkward and unwieldy it is for any task, and far to large and heavy to fit in a pocket. (Admittedly, this was pushed upon them by Congressional requirements for the Shuttle.) Apollo Applications (after which the CEV program should be modelled) was instead intent of using economies of scale by basing their platform off of different configurations of the Saturn V rocket and command hardware dedicated to a specific mission range.

CEV, as currently conceived and funded, is a failure in the making. But nobody at NASA is willing to argue the point; after all, they want to keep their jobs. I know the feeling. :frowning:

Stranger

46

OK, maybe I should look at those claims from those inside NASA - do you have cites?

It’s better than giving money to someone who’s saying the equivalent of “I don’t need a babysitter, I’ll behave.”

And records showing what measurements were done on each item every step of the way, what off-the-shelf parts were used, who made them… It’s all necessary.

Here’s one example. If I order a computer module and requested a listing of every little part in the module, including detailed listing of all materials used, and detailed information on how the board was manufactured and cleaned, would that seem like excessive paperwork to you? But to me those are absolutely crucial, because I’m putting it inside an ultraviolet telescope. Certain materials can outgas and contaminate the UV optics, possibly to the point of making them completely opaque. So someone who can’t provide me with all that documentaiton ain’t getting my money - it’s not enough for them to tell me “oh yeah, it’ll work in space just fine.”

And what if I order two, one for flight and another for backup, and one capacitor on the flight system fails during thermal-vac test? I need to know who made the capacitor, what its specs and tolerances are supposed to be, whether the backup has a capacitor from the same batch, etc.

What’s the net got to do with it? :confused:

47

I’ll dig around and see if I can’t find you some this weekend. IIRC, there was some mention of it in the CAIB report.

And you’re putting words in Rutan’s mouth. He’s not wanting to ditch all the paperwork.

Que? The first thing you’re going to do is compare two trannys that failed, one American built, one Japanese built. Presumably you were smart enough to do destructive testing on your own, and aren’t relying on whatever broken units you have laying around in the shop. You’ll dissassemble them and look for any obvious differences, did they change the mounting of the torque converter or whatever? Next you look at the wear on the components, run some tests (takes maybe an hour or so) to see if there’s any differences in the materials used (and this can some times be determined by simple visual or tactile inspection), then you start with the less obvious stuff, like tolerances. If that doesn’t turn up anything, you can start digging through the paperwork, or, even better, talking to the people who built the thing.

This might not work in your situation, but I’d personally try to figure out how to slap the thing together so that it didn’t matter if the thing gave off hydroflouric acid it wouldn’t screw up my work. I realize that this isn’t always possible, but having stuck my head in enough machines over the years, I’ve found that most engineers don’t think too much about the total configuration of things, so while their solution does work, if they’d spent their time poking about in the finished product, they’d learn that a slight shift in one or two of the components would make the lives easier of everyone who had to use their product.

And if the only company willing to supply you that information has a hit or miss track record? Do you keep using them or do you call up one of the other companies who says they can’t provide you with all the paperwork, and try to see if there’s not some way you can reach some sort of arrangement whereby you get all the information you’re likely to need, without placing too much of a burden on them, especially since everyone else who’s used them hasn’t had any problems with their gear?

A.) It’s a large system which we intergrate with seamlessly for the most part. B.) It took a lot of work to get it work the first time, but thanks to some general guidelines that have been set out, anyone can connect to it easily, they don’t need to be a software engineer to do so. C.) One of the issues involved is that every computerized component on the CEV (of which there’s going to be zillions) is going to have to talk to a host of other components, and it seems to me that the fastest/cheapest way to make this happen in most cases would simply be to adapt the existing protocols used for LANs, internet, PC peripherals, and the like. to these components where ever possible, rather than trying to develop a new system from scratch. IIRC, the computers in SS1 ran on Linux and not their own unique OS. (And Rutan seems to like to design everything on Macs, which is kind of funny when you think about the fact that Paul Allen of M$ is paying for everything.)

48

Oh, scr4, now that I think about it, you and I were both probably talking out of our asses when it came to the kind of paperwork Rutan was having issues with. The kind of paperwork you’re talking about is probably required for FAA certs on Rutan’s planes, NASA no doubt requires a little bit more in those areas, since there’s probably little off-the-shelf hardware for things like avionics. this article has a pretty good breakdown on the paperwork and areas which are most likely the ones which put Rutan off his feed.