LM Atlas V Man Rating Paper
An anonymous commenter sent me a link to this paper that LM is presenting at the Space 2006 conference this weekend in San Jose. While the tone definitely has some marketing hyperbole to it (calling Atlas V a "high flight rate" system, etc), but it addresses some of the key issues brought up by ESAS. I don't have time to do a full analysis (I'm on my lunch break), but here's a summary:
I could go on. There was some really interesting meat that is well worth reading for any company that wants to build crewed space transports. If Lockheed could do this, I bet that Boeing could also figure out how to make their vehicle meet NASA's official human rating rules without too much expense either. The Continual Employment Vehicle would have to be seriously Jenny-Craiged in order to launch on either of these vehicles, but there's a point at which you have to ask yourself: if existing vehicles are less expensive, require far less of an upfront investment, and actually meet NASA's own safety standards without having to be grandfathered in...what legitimate justification is still left for the Stick?
Oh, one final caveat. In all of these discussions about EELVs, I need to make clear, that even at high flight rates, I think these vehicles are too expensive in the long run. Cheaper than the Shaft and the Longfellow, sure, but still far too expensive for a sustainable space transportation infrastructure. However, once NASA has gotten out of the building/operating their own Earth-to-Orbit transportation business (you know, one of the selling points of the original VSE), and has made an architecture that can be launched using smaller vehicles, it makes it far easier for lower-cost commercial vehicles to eventually take the lead. If NASA were to adopt a drylaunch approach, and go with commercially available human, propellant, and cargo launches, it would be in a great position to benefit from continuous improvements in low-cost space transportation as they become available. Right now, even with Griffin's "we'll buy propellants from a depot to refuel EDS's if they become available" approach, NASA is locking itself into two vehicles which will be obsolete by the time they even enter service.
- Factors of Safety: NASA Human Rating requirements (NASA Standards 8705.2A as well as NASA Standard 5001) call for a 1.4 Factor of Safety on most of the subsystems of the rocket. For an Atlas V 401, carrying a 20klb capsule, it turns out that Atlas V meets and exceeds this criteria, as most of the systems were designed for handling the limit loads for flights carrying SRBs, heavier payloads, etc. Now, I'd never design something with that thin of a margin, but they don't have to deal with high cycle fatigue issues (since the vehicle is expendable), and they also have flight data from over 100 flights to back up those design criteria and analyses.
- Abort Loads: Even with the single engine Centaur upper stage, they were able to come up with a trajectory that is sufficiently shaped to avoid exceeding the peak abort reentry G's laid out in 8705.2A, while only losing between 5-10% of their payload capacity. And nobody but NASA can figure out how to force the requirements of an ETO capsule bad enough to make it too heavy to loft on an Atlas V 401.
- Another interesting point is that Lockheed wants to implement a lot of the Emergency Detection System hardware to every flight, even unmanned flights, in order to build up more datapoints, and get more flight experience with it. I do believe there was a canard floating around that making all the changes in order to "manrate" Atlas V or Delta IV would make the normal versions "too expensive" for unmanned customers...Apparently Lockheed disagrees.
I could go on. There was some really interesting meat that is well worth reading for any company that wants to build crewed space transports. If Lockheed could do this, I bet that Boeing could also figure out how to make their vehicle meet NASA's official human rating rules without too much expense either. The Continual Employment Vehicle would have to be seriously Jenny-Craiged in order to launch on either of these vehicles, but there's a point at which you have to ask yourself: if existing vehicles are less expensive, require far less of an upfront investment, and actually meet NASA's own safety standards without having to be grandfathered in...what legitimate justification is still left for the Stick?
Oh, one final caveat. In all of these discussions about EELVs, I need to make clear, that even at high flight rates, I think these vehicles are too expensive in the long run. Cheaper than the Shaft and the Longfellow, sure, but still far too expensive for a sustainable space transportation infrastructure. However, once NASA has gotten out of the building/operating their own Earth-to-Orbit transportation business (you know, one of the selling points of the original VSE), and has made an architecture that can be launched using smaller vehicles, it makes it far easier for lower-cost commercial vehicles to eventually take the lead. If NASA were to adopt a drylaunch approach, and go with commercially available human, propellant, and cargo launches, it would be in a great position to benefit from continuous improvements in low-cost space transportation as they become available. Right now, even with Griffin's "we'll buy propellants from a depot to refuel EDS's if they become available" approach, NASA is locking itself into two vehicles which will be obsolete by the time they even enter service.
posted by Jon Goff at 1:25 PM
15 Comments:
.
(not related with Atlas man-rating)
I think that a (small SM, TEI-only) Orion is like a (single purpose) "CorkScrew", while, a (big SM, multi purpose) Orion may be the "SwissKnife" of space exploration since it can perform autonomous (manned and unmanned) missions also without the LSAM.
I explain my opinion in details (with a curious image and a list of TEN advantages of the bigSM Orion) in my latest article [ "CorkScrew Orion or SwissKnife Orion?" ] here: http://www.gaetanomarano.it/articles/014swissCEV.html
.
In all of these discussions about EELVs, I need to make clear, that even at high flight rates, I think these vehicles are too expensive in the long run. Cheaper than the Shaft and the Longfellow, sure, but still far too expensive for a sustainable space transportation infrastructure.
Too expensive compared to reusable craft? So far we've only got one data point on those, and they're not exactly cheap - and it is precisely the high cycle fatigue issues that you mention that are a big part of the problem.
I like the idea of a partly-reusable craft. The only part of a manned spacecraft that has to return to the earth is the part that carries the men. So, if one was to launch a variant of the HL20 on top of an expendable booster, then the reusable part is very small, and getting it ready for the next flight requires far less maintenance than a fully-reusable SSTO or TSTO craft would.
Ed,
Too expensive compared to reusable craft? So far we've only got one data point on those, and they're not exactly cheap - and it is precisely the high cycle fatigue issues that you mention that are a big part of the problem.
Not so much as you would think. Beware of trying to plot arbitrary functions through a single data point. Sure, you can get any curve you want to fit, but you have absolutely no evidence that your curve represents reality.
Airplanes fly for years, without high cycle fatigue becoming that much of an issue. It just means that you have to make the parts a little bit thicker. If you design to a more liberal 1.7-2.0 factor of safety, fatigue issues pretty much go away.
I like the idea of a partly-reusable craft. The only part of a manned spacecraft that has to return to the earth is the part that carries the men. So, if one was to launch a variant of the HL20 on top of an expendable booster, then the reusable part is very small, and getting it ready for the next flight requires far less maintenance than a fully-reusable SSTO or TSTO craft would.
Sure, but you also would save a whole lot less by that approach. Honestly, I think we have far too few of data points to really conclude that RLVs are inherently expensive. Most of the problems with Shuttle came from the fact that the design (like that of the Shaft and Longfellow) are mostly politically driven. They get extra points for employing more people in more places. The incentives are inherently wrong.
A TSTO RLV built with a more reasonable target payload (say in the 1-10klb range), with a small crew, and not setup to be a lab or a space motel, and not designed to do a return to launch site after a single orbit, will end up being a lot easier of an engineering challenge.
~Jon
Well, Jon, since we only have the one data point on reusable craft, I guess that saying that expendables are too expensive in the long run can't really be supported either: too expensive compared to what? Certainly not compared to our existing data point. I guess this must remain an open question until there are more data points for reusable spacecraft, or until another paradigm like the space elevator becomes a reality, so that we have something else to which we can compare the expendables.
Ed,
[EELVs are] too expensive compared to what?
Compared to where the price needs to get to to really open up new markets, and to allow for us to become a truly spacefaring society? Even if they got their prices back down to half of what they were originally priced at (say $40M for the vehicle, and $10M for the reusable capsule), that would only barely start getting to the part of the demand curve where the number of tickets per year gets interesting.
It's a step in the right direction, but not a sufficient step. I think that RLVs can do substantially better, and that's not just based on faith. We have partial datapoints from several other RLV projects such as X-15, DC-X, and SS1, and all of them point to the same conclusion--that the Space Shuttle is a lousy data point that can be improved on substantially.
~Jon
Jon, I think part of the issue with aircraft is that the industry has developed the different 'checks' (including the one that financiers fear, the 'D' Check. Very costly, that one) to address the wearoutability of the different parts. Some parts need to be checked a lot. Some don't, but do need to be checked regularly. Some we don't think need to be checked often actually do but we don't usually figure that out until someone dies.
Now, if I were running the shuttle program (which I'm not nor would ever want to), I'd be expecting that the forensic analysis would be compiled, analyzed and published, leading over time to a greater sophistication in the operation of the vehicle. But of course it is a developmental vehicle, so its lessons will never be applied in operations (presumably).
I actually applaud LM's indication that they want to put more sensors in all vehicles to gather exactly this kind of data. By riding higher up the learning curve they will hopefully learn how to simplify the overall sensor requirements, leading to cheaper requirements farther out. That's good strategic thinking.
I have to agree on your timeline. Expendables are not a long range solution (except at the upper end of mass to orbit), TSTOs will probably work out next, followed (but not replaced) by SSTOs.
Part of it is going to be learning what needs to be looked at/replaced often, and what can go 10/20/40 flights. That's what happens when you climb the learning curve. Makes me wonder where the Shuttle's learning curve has been going...
Most of the problems with Shuttle came from the fact that the design (like that of the Shaft and Longfellow) are mostly politically driven.
If all you mean by "politically driven" is "constrained by what Congress would appropriate" ... well, every engineering design ever has been cost-constrained, wherever the money comes from.
If you mean STS designers tried to please several user "constituencies"... again, that happens in commercial markets as well as in government programs. Or does MSS have no interest in finding the widest possible range of customers?
I wouldn't deny that once STS was built, those involved in its operations became a standing interest group... or that maintaining rice bowls has played a part in ESAS. But it's glib to say STS design was "mostly politically driven"... and glibly dismissive of some engineers I think you'd have respected if you'd watched them at work on it.
If i read the LMA paper correctly, they are quoting about 20klbs to LEO.
This falls way short of way the ORION CEV weighs, or even apollo for that matter. While the CM fits into this weight, adding in the service module and the launch abort system, will easily push you pass this 20k number even for apollo. With the CEV (ORION), carrying 6 to ISS and so being larger - this is not an option. Now, you must recall that ORION is designed for the lunar mission, and ISS is a fallout mission. It's over designed for ISS. But then so is shuttle as far as crew is concerned. If you seriously look at the requirements for lunar missions, you will not be on a 401, you'll need 3 common core boosters to even get close.
If i read the LMA paper correctly, they are quoting about 20klbs to LEO. This falls way short of way the ORION CEV weighs, or even apollo for that matter. While the CM fits into this weight, adding in the service module and the launch abort system, will easily push you pass this 20k number even for apollo.
Well, how much does 'GE Apollo' on steroids weigh?
Anonymous,
I accidentally blew off my first reply, so this one will be a bit shorter.
If i read the LMA paper correctly, they are quoting about 20klbs to LEO.
Yup.
This falls way short of way the ORION CEV weighs, or even apollo for that matter. While the CM fits into this weight, adding in the service module and the launch abort system, will easily push you pass this 20k number even for apollo.
Actually, the 20klb is more than enough for a large cabin (I think they were thinking 5-8 people, though they weren't explicit in this exact paper), as well as a small service module/launch escape system. The thing people need to realize is that it's been 40 years. The CSM was designed using slide rules and draftboards. Electronics are lighter, composites are available now, manufacturing processes have greatly improved, solar panels are now available instead of just using fuel cells, etc, etc. Basically, just because Apollo did something one way doesn't mean it's the best possible way to do it today.
Now, you must recall that ORION is designed for the lunar mission, and ISS is a fallout mission. It's over designed for ISS. But then so is shuttle as far as crew is concerned.
I think that you sum up the problem with the CEV and the Shuttle very well. The problem is that both of them are trying to be way too much in one vehicle. Not only are they a crew transport, but they're also a motel, and the CEV has this big huge rocket on it because Ares V is too pansy to actually put the whole stack into Lunar Orbit and bring the CEV back (because that would require more on-orbit assembly or *gasp* figuring out how to do on-orbit propellant transfer).
This isn't the best way of doing things. NASA would be far better served if they just developed a "mission module" that included the long-term crew accomodations with a docking port that was compatible with all of the major planned commercial crew transportation vehilces (like what SpaceX, Kistler, LM, Boeing, and others are working on). For propulsion actually make a real lunar transfer vehicle that has the capability of putting the whole stack all the way into lunar orbit (none of this "we'll send you into a lunar transfer orbit, and then let you do the rest with the LSAM" cop-out garbage). Sure, that would require learning how to transfer propellants on orbit, and sure, that would make NASA act like a customer, instead of some centrally planned and operated RoketBureau.
If you seriously look at the requirements for lunar missions, you will not be on a 401, you'll need 3 common core boosters to even get close.
If you seriously look at the requirements, and then look at what is the most cost effective way of meeting them, you wouldn't try to put all of that into a single launch. By forcing things to be launched on a vehicle that is too big to be commercially viable, you greatly decrease the safety and robustness of your system while simultaneously driving up costs. You can't divorce engineering from economics without getting really stupid results.
~Jon
Not too sure what 'too big to be commercially viable is'. Certainly boeing and lockheed will both happily sell you a 3 stick configuration.
For just getting folks (and i don't think it would be more than 4) to ISS you could probably do it on single stick.
If you are going to the moon. It will take a 3 stick + the CaLV, or the CLV and CaLV. However you do it, you have to get the same mass up.
And while you could certainly make this a nicer system with an on orbit fuel depot. The simple fact is we don't know how to do that yet. NASA tends to do the simple, if not expensive, solution to things. A fuel depot, in my opinion, is not simpler than just building 2 really big rockets.
That said. Space X and RPK Kistler have got there chance. I would think that upon sucessfully building and demonstrating a ISS supply (manned or not) capability, that a obvious next business opportunity would be to start figuring out how to store fuel on orbit. I personally thing that a lox methane fuel depot + an LTV designed to use it ... all for rent would be a great idea. Take cargo to the ISS, rent LTV's to NASA to get crew/cargo to/from the moon...
Anonymous,
Not too sure what 'too big to be commercially viable is'. Certainly boeing and lockheed will both happily sell you a 3 stick configuration.
Actually, the evidence points in a different direction than you think. LM was supposed to deliver an Atlas V Heavy variant, but never actually finished it, and doesn't want to finish it unless they get more government funding. That alone should tell you that at least they don't think there's enough commercial demand to make it worth their while to bring the Atlas V Heavy to market. SpaceX is in the same boat. They've talked about their Falcon IX Heavy vehicles, but their position has always been--we'll actually finish development on them if and only if we get enough customers who want it to justify the costs. So far they haven't.
Even Boeing has only flown, what 2 Delta IV Heavy launches? The fact that LM is perfectly willing to let Boeing have that market unchallenged should tell you something about what people in the industry really think about the prospects for that market. The only potential commercial customer that I've seen who could use a launcher that big is Bigelow, but who knows how many launches per year that would be. The fact that nobody new is jumping out of their seat trying to fund on their own dime the development of such heavy lift, shows pretty conclusively to me that nobody thinks the market is even close to there for such big vehicles outside of government use.
For just getting folks (and i don't think it would be more than 4) to ISS you could probably do it on single stick.
For 20klb, they should be able to fit 8 people easily. T/Space was talking 4 people on a 9klb to orbit booster, SpaceX is IIRC 8 people on a vehicle with almost identical LEO capacity as Atlas V. Almost everyone I've seen outside of NASA thinks that 20klb is plenty of mass for carrying 5-8 people if done correctly using modern materials and technologies.
As for the rest of your comments, I agree more with the last paragraph. The problem with the current plan is that it is really unsustainable, expensive, and a really crappy payback to cost ratio. Storing and handling cryogens on orbit isn't trivial, but it's not ridiculously tough either. If they wisely spent even a fraction of the CLV cost on funding multiple methods (say one business as usual contract, one or two COTS style fixed, pay-on-milestone projects, and some prizes for anyone else), they could likely have a workable solution before the decade is out.
Even if you don't postulate on-orbit propellant tranfer, you still don't need to launch all the parts for the CEV on a single launch. Breaking it up into a capsule, a mission module, and a propulsion module, you could at least get rid of the CLV, which would free up 5 billion to speed up other parts of the architecture. Or you could go to a 2.5 launch architecture, where you launch the CEV mission module and propulsion module, along with say the LSAM ascent module on one Ares V (minus the SRBs), and then launch the LSAM descent module and EDS on a second stripped down Ares V, and lastly send the crew up on whatever commercial vehicle (Atlas V, Falcon IX, or Kistler K-1) can deliver them. That at least would allow you to close down the solids line, only require one new booster, and get a higher flight rate out of that booster. It would also free up most of the CLV money to expedite the completion of Ares V and the other stuff.
Eh...but that would make too much sense, huh?
~Jon
And while you could certainly make this a nicer system with an on orbit fuel depot. The simple fact is we don't know how to do that yet.
Er.. i was under the impression that Progress M1 regularly refuels ISS.
Why cant an M1 or a similar system refuel lunar stacks ? Licence the relevant pieces from russians if you cant figure it out on your own.
ISS uses biprops. The discussion here is dealing with cryogenic fuels. Big difference.
Actually there's a Boeing powerpoint that tells about lessons learned from the shuttle program, and it says, among many other things, roughly that the design was too ambitious and unrealistic and the margins not big enough.
Post a Comment
<< Home