It’s kind of amusing to note (as several others on the blogosphere have now) that the supposedly “safe, simple, soon” Crew Launch Vehicle is turning out to be as difficult to pull off as many of us have been saying all along. The fact that modifying an SSME for airstart would be extremely difficulty if not impossible is hardly news. I don’t have an exact number on how many SSME’s were destroyed trying to get its startup sequence worked out, but it was a disturbingly high number. And that was with a slow, ground-launched startup sequence. As a friend of mine quipped, “every blip or jiggle in the SSME startup sequence has at least one trashed engine proving its neccesity.”

But don’t worry folks, so what if there really is no off-the-shelf hardware on this vehicle, NASA says that this will have about a 2000:1 odds of blowing up on any given launch. They’re so smart that they knew the reliability to four significant figures before they even figured out how many segments the SRB was really going to need, or which engine was going to be used on the upper stage. Impressive.

I have to say that what NASA is doing with its COTS program, with Centennial Challenges, and with the money it’s setting aside for microgravity research, is actually pretty darned good. It’s just sad that the percentage of NASA’s budget over the next five years that is going to stuff that has a potential of opening up the space frontier for the rest of us is such a small percentage of the money getting blown on the Space Shuttle, and on legalized graft like the Shaft.

Michael and several others like to point out that the Shaft and finishing up ISS with the Shuttle is more or less the price we have to pay for things like COTS, but it still doesn’t make it any less galling to see so much money being thrown down such an obviously useless hole like this.

I still think that if the Shaft is supposed to be such a great deal, such a superior space launch vehicle, that NASA should require ATK to come up with all but $1B of the funding, and use the rest of that money saved on actually promoting a thriving commercial space transportation infrastructure. Why should the nation have to foot the bill when ATK pulls the good ol’ bait-and-switch on us?

I feel Shafted.

The following two tabs change content below.
Jonathan Goff

Jonathan Goff

President/CEO at Altius Space Machines
Jonathan Goff is a space technologist, inventor, and serial space entrepreneur who created the Selenian Boondocks blog. Jon was a co-founder of Masten Space Systems, and is the founder and CEO of Altius Space Machines, a space robotics startup in Broomfield, CO. His family includes his wife, Tiffany, and five boys: Jarom (deceased), Jonathan, James, Peter, and Andrew. Jon has a BS in Manufacturing Engineering (1999) and an MS in Mechanical Engineering (2007) from Brigham Young University, and served an LDS proselytizing mission in Olongapo, Philippines from 2000-2002.
This entry was posted in Uncategorized. Bookmark the permalink.

19 Responses to Shafted

  1. Paul Dietz says:

    I was under the impression the 1 in 2000 was the chance of killing a given crew member, not the chance of the vehicle failing. Losing the crew requires both the launcher and the abort system fail, so a 1 in 2000 chance seems more reasonable.

  2. Jon Goff says:

    The problem as I see it is that they were specifying some reliability numbers for the stick to 4 significant figures (something like 2147:1 for example) when they still don’t know how many SRB segments they’ll have to a single significant figure.

    The other thing that bugged me was that they were claiming a crew safety number that was something like 3-4 times better than if they launched on Delta or Atlas. That implies that they think they’ll get a better than 99.5% reliability on a vehicle that essentially has no flight history, and has a stage that can’t be shut down once it has been lit. I say it’s bullox.


  3. Anonymous says:

    First, the Spaceref piece is pretty darned vague. “Sources” report “problems,” without any indication of who the sources are or what the specific problems are. So isn’t it a bit soon to be jumping on the bandwagon?

    Second, they are still in the very early stages of design. Can you name a single rocket in history that did not have “problems” at some point in its development?

    Finally, you say that the vehicle “has no flight history.” But that depends upon how one wants to count history. You could claim that a man-rated Atlas also has no flight history either, since we haven’t flown one. SRBs and SSMEs have proven their reliability over a lot of launches. The CLV will use modified versions of both.

    As with all Spaceref articles, it is best to wait for confirmation by real media outlets before making any conclusions.

  4. Jon Goff says:

    While it is true that a single data point isn’t much to go off of, this isn’t really just a single data point. Anyone who has any experience whatsoever with rocket development, particularly anyone who knows anything about the particular systems in question knows that ATK has been bending the truth a bit with its salesmanship of the CLV.

    Problems with making the SSME air startable are not surprising at all if you know the system, and that is exactly my point. The number of modifications required to make the SSME a good and reliable air-startable upper stage engine are so large that the engine you have at the end of your development is no long an SSME. It is a brand new engine with no flight history. The number of modifications that need to be performed to the SRB to use it on the CLV aren’t quite so severe. I mean, there will be a bit of commonality, and you can extrapolate a bit on its reliability, but it isn’t the same as what has been flown. The rest of the stage systems for the upper stage (tanks, plumbing, structures, etc) are also completely new and unproven. Anyone with even a decent amount of background knowledge about the systems in question knew that the claim that you could just use the systems off the shelf was total BS.

    I’m not saying that I expected them to do this without problems, just that when you already have vehicles that would work just fine as a launcher, without any modifications at all, the arguments that ATK made for the stick are bogus.


  5. kert says:

    coming back to data points .. youd actually have to have at least 2001 data points to claim 2000:1 chance of crew being killed.
    i dont think i will find anyone willing to make a bet that this thing will ever launch 2000 times ?

  6. Leland says:


    I think you are mirroring Jon’s arguments against ATK. ATK may not be vague about the reliability, and they may be overly forthcoming about “safe, simple, soon”. But they are very vague on how they make the leap from a vehicle never built to knowing it can be done, “safe, simple, soon”.

    Jon points out several other datapoints that counter ATK, primarily the launches of Delta and Atlas rockets. Early discussions considered man rating these vehicles for CEV, might they have been “safe, simple, soon” as well? Well, no, many believed they could be made safe (man-rated safe), but not simply or necessarily sooner than other vehicles. Nice, healthy, scepticism… we should have the same for the ATK proposal.

    The space program has a lengthy history in providing too many significant digits while discussing reliability data. My all time favorite was a MTBF for a ISS rotary bearing with 10 significant digits. The units were hours, but if converted to years, would still be an expected failure of a mechanical device every 100,000 years or so (well maybe not ‘or so’… again 10 significant digits). That’s pretty amazing analysis. One wonders how they calibrated their test.

  7. Iain McClatchie says:


    I don’t think the EELVs are an alternative to the Shaft.

    Delta would require the D4H configuration, laughably expensive. Of course, the $5B to develop the Shaft would buy 20 D4H launches, or maybe 30 with lower costs from higher volume. Hmm… not much of a laugh there.

    If I understand correctly, and please correct me if I’m wrong, the Atlas uses a Russian-derived main boost engine. I don’t know where the engine is manufactured, but if the Russians maintain any control over the engine technology (even patents or IP controls of some sort), they would end up controlling the whole booster and thus have a big lever on the U.S. manned space program.

    And so, if the EELVs are not an alternative, what is? To NASA, Falcon V is too speculative, and the rest of you are… beyond speculative.

    But all this rests on my supposition that the supply or use of Glushko RD-180s is controlled by Russia.

  8. Kelly Starks says:

    > they were claiming a crew safety number that was
    >something like 3-4 times better than if they launched
    >on Delta or Atlas.< If your referring to the 1 out of 2000 number, isn’t that about 40 times better? All current manned and unmanned launchers (man ratings a joke given the reliability numbers are the same) have about a 2% catastrophic failure rate, and escape systems often to usually kill or cripple the crew. So for CEV on “The Stick” (a extremely violent and unsafe launch config) to claim a projected 1 out of 2000 crew loss rate — Well, NASA lies a lot.
    Ambivalent said…

    >…the Atlas uses a Russian-derived main boost engine. I
    >don’t know where the engine is manufactured, but if
    > the Russians maintain any control over the engine technology
    >(even patents or IP controls of some sort), they would end up
    >controlling the whole booster and thus have a big lever on the U.S. manned space program.

    The RD-180s are made by Pratt & Whitney under license from Russia since it’s a cut down Russian RD-170 design. So theirs no real leverage on the Russian side since the licenses are signed off.

    On the Plus side the RD-170s were excellent engines, designed for reuse on their shuttle, and tested out as capable of 20 flights. So the RD-180s should be a strong engine to unless Pratt really screwed up. And at $10 million a copy they are ridiculously cheep by US standards.

  9. Tom Cuddihy says:

    “they were claiming a crew safety number that was something like 3-4 times better than if they launched on Delta or Atlas. That implies that they think they’ll get a better than 99.5% reliability “

    No, it doesn’t. It implies that crew safety (including combined booster reliability and launch abort surviveability) is highly favorable on a Stick as opposed to an EELV. According to Scott Horowitz, the main problem with survivability on top of an EELV is apparently the thrust profile and especially loft. If you have an abort late in the launch (as in mid way into the second stage) of an EELV, according to Horowitz, you have a huge chance of exceeding surviveability limits on ballistic return of the ejected capsule due to g-loading. I don’t know if that’s true, but that’s what Horowitz claimed in person last June.

    Let me quote from ATK’s SAIC study,
    “The simplest designs of the EELVs, which offer the greatest potential for inherent
    reliability, are the single core variants. These single core EELVs with an effective
    crew escape system should provide the greatest crew safety. Unfortunately, the
    single core EELVs are unable to meet the performance needs for the CEV
    mission, so the higher performance, more complex, less reliable multi-core
    “heavy” variants are required making it difficult to achieve the ascent risk goal
    proposed by the astronaut office for PLOC to be better than 1 in 1,000. This
    dilemma motivated the search for a launch vehicle that could preserve the
    simplicity of a single core propulsion system that utilizes highly reliable humanrated
    heritage components with sufficient performance to meet the CEV mission
    needs. The result of this effort is a 2-stage launch vehicle utilizing a single
    Space Shuttle Solid Rocket Booster (SRB) for the first stage, and a single J-2S
    engine for the second stage.”

  10. Tom Cuddihy says:

    btw, I know the quote doesn’t support what Horowitz told us in person–so I have some reservations about how accurate that launch abort surviveability issue was. It may be that the astronaut office made some very untenable assumptions that led to that conclusion.

    But the quote at least makes their thinking clear.

  11. Anonymous says:

    See Apparently NASA is going to use the J-2S instead of the SSME.

  12. Tom Cuddihy says:

    It is notable that the SSME option was not part of the original ‘stick’ proposal by the astronaut office: it had always assumed use of the J2S for the reasons that are now abundantly clear. SSME was ‘inserted’ in a bid to placate members of Congress about maintaining shuttle infrastructure. With Tom Delay out of the way, they probably have a lot freer hand now. (also, it’s much easier to ‘peel off’ the SSME requirement than if they had never included it in the proposal.)

  13. Anonymous says:

    The talk about reliability by NASA these days is totally incoherent.

    If you want to read real reliability numbers dig into the NTRS server and download the Saturn 1B manual (that the astronauts studied). The reliablity of the Saturn 1B first and second stages was stated as 0.84. Now with the crew ejection system that went up to 0.95.

    This was after more than 20 successful launches were under their belt as this was the Apollo/Soyuz mission version of the doc.

    Too many smart people at NASA with no real experience in launch vehicles (including Mike Griffin).

  14. Mark says:

    “SSME was ‘inserted’ in a bid to placate members of Congress about maintaining shuttle infrastructure. With Tom Delay out of the way, they probably have a lot freer hand now.”

    I’m pretty sure that there are no rocket fabricators in Tom Delay’s district. I suspect that the first statement has no evidence to support it either. Most of the “shuttle infrastructure” involves turning around the orbiter fleet which is going away regardless of what gets picked for VSE. Indeed, a great many shuttle workers are close to retirement and will be gone in any case. A whole new generation is going to have to be hired to run VSE, which in my opinion is all for the good.

  15. Tom Cuddihy says:

    Mark, you don’t know what you’re talking about, although I admit to mixing my some of own speculation with fact.

    I wasn’t trying to imply that Delay was worried about losing jobs at Marshall connected to the SSME. (Delay’s district is in Texas, and has JSC). But he was the majority leader and has never been supportive of reducing or radically changing the NASA beaurocracy. It’s true that I don’t KNOW that SSME was inserted for the purpose of dragging along reluctant GOP members who like the status quo.

    But it is a fact that last May (12 May 2005 to be exact), when Scott Horowitz was heading up the ‘safe,simple,soon’ team at ATK, he came to talk to a large group of space systems students at the Naval Postgraduate School, followed up by a smaller session with students in my propulsion class (set up by his fellow former astronaut Dan Bursch). He talked about the details of the stick plan, and answered a lot of questions we threw at him.

    He showed us an excel spreadsheet “paper napkin analysis” that summarized the results of several different ‘stick’ derivatives, including 5 segment versions, multiple RL-10 upper stage versions, and even a ‘super cheap, but small payload kerosense upper stage.’ His team had selected the H2/LOX J2s as the best trade between payload and cost. There was no ‘SSME’ upper stage version.

    Somebody (not me) asked him why not SSME, and his answer at the time to us was that the J2S had been designed from the start to be an upper stage engine and would be cheaper to throw away then an SSME. I’m working from memory here of events from several months ago, but that’s what happened to my memory, if you have any doubts I can personally email you the ppts and Excel documents he gave us.

  16. Tom Cuddihy says:

    I should mention that as far as I remember, Horowitz never mentioned any concerns about airstarting an SSME–he was talking about cost. But his proposal was exclusively J2S, even though several months later NASA started leaking documents showing a ‘J2S/SSME’ stick. That’s what makes me think it was added in as a political ploy.

  17. Tom Cuddihy says:

    okay I’ve been checking the documents he gave us (not that I doubt my memory…but nobody’s perfect) and, sure enough, although the design features J2S and all the reliability / safety documents talk J2S, one of the ATK ppts shows an ‘evolution’ path that shows a ‘J2S/SSME’ upperstage on the 5-segment SRB version… hmm.

  18. Mark says:

    Tom – So I don’t know what I’m talking about, but you seem to be agreeing with my points. I’m not sure what you mean about Delay being “unwilling to change NASA bureacracy.” He’s an enthusiastic proponent of VSE which must, in any form, do that, even though the shuttle has been a cash cow for the Houston area for years. I doubt that he knows or cares about the virtues of one rocket system over another.

  19. Anonymous says:

    “With Tom Delay out of the way, they probably have a lot freer hand now. (also, it’s much easier to ‘peel off’ the SSME requirement than if they had never included it in the proposal.)”

    You’ve already said that you don’t have any evidence to support this conclusion, so why not assume a neutral stance until you have more evidence? I suggest reading the new issue of Space News, which has a longer article on this subject based on an interview with Horowitz. It clarifies the issue a little bit, but not completely.

    Essentially, he says that they were trying to reduce the amount of development that they would ultimately have to do, and shifting some of the development to the nearer term instead of the longer term in the hope of saving money over the long-term.

    The first version of the ESAS would have required doing separate development:

    4-segment SRB
    air-start SSME

    5-segment SRB
    ground-start SSME
    air-start J-2S

    Now what they have done is reduce that approach so that they are ONLY going to do the 5-segment SRB and NOT do an air-start SSME.

    The effect of this is to reduce some development risk (air-start SSME eliminated completely) and shift some of the development risk (5-segment SRB, J-2S) from say 2012 to about 2007 or 2008. So it moves some things up earlier, but overall reduces the things that they have to work on. Air-start SSME would have been a significantly different engine, whereas now all they are going to do is keep the existing SSME from shuttle. I suspect that this also reduces operating costs a bit. Building a lot of SSMEs that you throw away is probably expensive and so they want to throw away fewer of those.

    Now as to why the SSME was included in the ESAS version 1.1? It may not have been politics at all. They may have simply decided that it would take less development time to build an air-startable SSME than to restart the J-2S production line. Perhaps upon further reflection they decided that air-start SSME would require more development time (and money) than they thought, and so the benefits evaporated and they decided to switch to the J-2S anyway (after all, they planned on using it for the HLLV).

    Unfortunately, the Space News article does not settle this.

Leave a Reply

Your email address will not be published. Required fields are marked *