Rebel Engineers Talk To NASA
Hi. I just saw this...Betruger wrote:I'm not arguing this. I'm curious to see how the arguments from 93143 and someone else on another forum play out. Unfortunately the other debater kind of refuses to play nice because he's jaded with debate.. From debating global warming with deniers. So if I want to see any action (actual brass tacks debate), I have to kick start things myself by either getting 93143 to show his math on the other forum or ask him nicely to tell me the basis he's arguing on, in brief, so that I can at least see how things stand.
It's the most satisfaction I'm going to get out of this whole new space policy thing. It's too time consuming to wade thru all the emotional arguments from everyone involved that otherwise would present polite, informed debate.
Here's the external thread.. Debater #2 is BadAndy.
I hope trying to spark this debate wasn't done too impolitely. In my experience there's more to learn from watching two sides of a particular subject of contention debate things, than just reading either one under vacuum.
I'll try to get a well-supported response together. Right now I can't do it off the top of my head...
-
kunkmiester
- Posts: 892
- Joined: Thu Mar 12, 2009 3:51 pm
- Contact:
From badAndy:
I wonder if anyone has really done the numbers for this. I had an idea that's a bit more suited to permanent assembly, but I'm wondering if there's any real study.Exactly. And let me remind everybody that the simplest "space assembly" is automated docking of self-capturing assemblies, and goddam, we've known how to do that for years. There is a mass penalty for this, it isn't that large. Only in the situation where the structural/mass penalties of that are unacceptable should one go to more complex assembly-in-space, which can also be robotic or teleoperated.
Evil is evil, no matter how small
-
Josh Cryer
- Posts: 526
- Joined: Sun Aug 31, 2008 7:19 am
Here's a more recent writeup about Ares I / Ares V: http://www.popularmechanics.com/science ... 45250.html
Science is what we have learned about how not to fool ourselves about the way the world is.
-
Heath_h49008
- Posts: 71
- Joined: Wed Jan 27, 2010 9:12 pm
- Location: Michigan
It would seem like something that should have been well fleshed out when we redesigned the ISS in 1993(?)kunkmiester wrote:From badAndy:I wonder if anyone has really done the numbers for this. I had an idea that's a bit more suited to permanent assembly, but I'm wondering if there's any real study.Exactly. And let me remind everybody that the simplest "space assembly" is automated docking of self-capturing assemblies, and goddam, we've known how to do that for years. There is a mass penalty for this, it isn't that large. Only in the situation where the structural/mass penalties of that are unacceptable should one go to more complex assembly-in-space, which can also be robotic or teleoperated.
If this craft exists. And can actually be turned around in 24-48 hours. It won't matter if it can carry 40,20 or 1 ton into orbit... It will own the market in any case. Pure H2 and O2 are not that expensive... What are your actual expenses? Tank maintenance? Swap fresh engines onto the airframe? Ablative heat shielding?
There isn't a scenario where this bird could cost anywhere near what the "disposable" rockets do.
-
kunkmiester
- Posts: 892
- Joined: Thu Mar 12, 2009 3:51 pm
- Contact:
I hope you weren't expecting a detailed breakdown... I'm just a Ph.D. student doing CFD of rocket engines; I don't have the background or data access to comprehensively prove my point. (Really, I was just running my mouth off based on the impression I've gotten from a couple of years of following this stuff on the Internet.) But I've taken a stab at it.
It's a bit of a broadside right now. Please tell me if it seems to need any changes/additions/deletions to constitute a decent argument.
Skipjack? Stay out of it unless you can be constructive, please...
_____________________________________________________________
Okay, first off I'm not a fan of Ares V. It was too big and too expensive - in fact it was so big that its payload volume capacity suffered (it would have had to fit through the VAB doors with the PLF on top)... not to mention that it was programmatically tied to Ares I...
Jupiter, on the other hand, is relatively cheap (DIRECT say about $12B, with margin, for the entire development program including upper stage; The Aerospace Corporation says their numbers are good) and dead easy - MSFC claims they could fly a J-1x0 prototype in 2012, using mostly existing parts. And there's apparently a group of corporations that's been in talks to build a Jupiter-like inline SDLV commercially, which would probably make it even cheaper. The incremental cost of a J-1x0 launch looks to be less than $200M if the tanks, engines, and SRBs are bought in reasonable quantities.
Now realize that the IMLEO for this scheme using EELVs is going to go up substantially, partly because lots of small structures are not as volumetrically efficient as a few large ones, and partly because all the individual pieces - including the propellant tankers - need to be (or be carried by) spacecraft in their own right, with maneuvering engines and propellant of their own, RCS, power, avionics... not to mention that spacecraft tend to cost significantly more than their launchers; there's at least another year of your HSF budget gone, and we haven't even touched orbital ops yet, never mind mission support...
You could sort of get around this with an orbital tug, but we don't have one, it wouldn't be cheap, and it would need fueling (you can't do LEO assembly on solar ion). Also, I doubt a single tug would work for an entire mission assembly sequence without repair/refurbishment, so it isn't a given that it would help at all...
Going with an inflatable transit habitat won't help - the DRM already has one. That's where the name TransHab comes from...
Then you have the extra cost of development and testing to get the automated assembly to work right (this sort of thing can get very expensive), the chance that a docking operation could fail and endanger the entire stack (hey, it's happened to the Russians before, and they weren't doing cryogenic propellant transfer)... the extra cost and risk involved in a segmented heat shield/entry aeroshell, because EELVs can't put up a big enough monolithic one (PLF size matters here, of course, but surprisingly mass does too; the aeroshells for the DRM are 40 tons each)...
Falcon 9 Heavy could probably do the launching more cheaply, but then so could Jupiter (they're in roughly the same $/kg class, especially if Jupiter is operated commercially - and if SpaceX can hit their cost targets, which isn't a given). And with Jupiter you don't have the time-consuming and expensive orbital integration problem.
Something like Zubrin's Mars Direct plan would be easier to fit on EELV-class lift - in fact, it's been done - but the guy who did it has admitted that (a) his plan isn't actually "for less" unless a dramatic improvement in launch cost is realized; the idea is to provide an "anchor tenant" for large volumes of small launches and break the current market deadlock, leading to cheaper RLVs, which as others have pointed out could just as easily be done with propellant depots serving spacecraft launched on HLVs, and (b) his plan doesn't work as presented due to aerocapture issues (ie: heat shield size).
There's probably a reason the Augustine commission ended up insisting that heavy lift was necessary for a robust exploration program, even while favouring the idea of commercially-supplied propellant depots (and even while ignoring the J-1x0, which is what makes DIRECT work, and adding a huge margin onto a cost structure that already had significant margins in it, while trying to cram too much other stuff into the budget, so the SD-HLV development ended up stretched way out and dominated by fixed costs. Hey, they were in a hurry...). A number of the Commission members went in thinking EELV-class was enough, but they changed their minds. Medium lift plus prop transfer/depots can give us the moon (suboptimally), but beyond that the high per-launch cost and the headaches of on-orbit integration make it far cheaper and easier to use a heavy lifter, even if you take the HLV's development costs into account.
Now, if you want large multi-MWe nuclear reactors in space, current EELVs will not be enough. Ad Astra's smallest Mars mission requires a 48 mT nuclear power system, and their largest (the famed 39-day transit) needs an 800 mT system... no, you cannot assemble an 800-ton nuclear power plant out of 20-ton pieces in space...
ULA's lunar proposal, using the ACES concept, assumed a $10,000/kg launch cost. So that's probably in the ballpark for the upgraded (40 mT) EELVs at the required launch rate for a couple of lunar missions per year (an ESAS-class lunar mission requires roughly 200 mT IMLEO).
DIRECT's cost and schedule estimates have been looked at by The Aerospace Corporation, and have been confirmed to be reasonable. This is not really surprising, since Jupiter is essentially a slightly modified STS stack, meaning the costs are not at all difficult to quantify.
Funny how all the naysayers simply confine themselves to saying the cost estimates are lowballed, without saying why they believe this or providing any alternative estimates...
Yes, as many have pointed out, the cost and schedule estimators for DIRECT are the same as the ones for Ares. That is, they're the same ones that put Ares I IOC in 2015 at 0% confidence, long before the Augustine Commission shone a spotlight on the problem. It was management that was painting the rosy picture. If the guys doing the estimates for DIRECT thought Ares was going to work, why were they working on DIRECT?
Besides, as mentioned above, Aerospace has already confirmed their numbers.
The Shuttle stack is actually pretty efficient when you remove the orbiter. Just because it's bigger doesn't mean it's vastly more expensive.
(Please note that some of the cost estimates on the DIRECT website haven't been updated since DIRECT v2.0, and thus include the J-2X development and RS-68 man-rating and infrastructure changeover costs. DIRECT v3.0 uses all existing engines - RL-10 should be easy to man-rate; they need it for the lander anyway, and since it's for the upper stage it doesn't need to be ready right away - and is thus cheaper and faster to develop.)
Operate the Jupiter commercially and it could start looking even better...
Back to Mars: Basically if you want something sustainable beyond flags and footprints, you need heavy lift, because otherwise it's too expensive and time-consuming to get a mission off.
If the heavy lift is cheap and sustainable enough (Jupiter, especially commercialized, should be; Ares V wasn't), you can then do a bunch of stuff with it - not just Mars. Large telescopes, Europa subsurface probes, manned NEO missions, manned lunar missions (of course), large one-shot space stations...
...
And what's this about teleoperation of robots from Earth? Beyond the moon? You're nuts. It makes me all shivery and antsy just thinking of trying to teleoperate a robot with a 40-minute signal round trip (Mars at its furthest point)... Spirit and Opportunity were probably as teleoperated as is feasible on Mars, from Earth. Even the moon is iffy for real-time work because of the nearly 3-second lag; to really do decent teleoperations work you'd need to be at L1 or L2...
_____________________________________________________________
It's a bit of a broadside right now. Please tell me if it seems to need any changes/additions/deletions to constitute a decent argument.
Skipjack? Stay out of it unless you can be constructive, please...
_____________________________________________________________
Okay, first off I'm not a fan of Ares V. It was too big and too expensive - in fact it was so big that its payload volume capacity suffered (it would have had to fit through the VAB doors with the PLF on top)... not to mention that it was programmatically tied to Ares I...
Jupiter, on the other hand, is relatively cheap (DIRECT say about $12B, with margin, for the entire development program including upper stage; The Aerospace Corporation says their numbers are good) and dead easy - MSFC claims they could fly a J-1x0 prototype in 2012, using mostly existing parts. And there's apparently a group of corporations that's been in talks to build a Jupiter-like inline SDLV commercially, which would probably make it even cheaper. The incremental cost of a J-1x0 launch looks to be less than $200M if the tanks, engines, and SRBs are bought in reasonable quantities.
How about a thousand-ton Mars mission? The IMLEO requirement for the current DRM (which uses nuclear propulsion, BTW) would require about 40 EELV Heavy launches, which would still be close to $200M per flight even with economies of scale factored in. That's $7B right there - most of NASA's HSF budget for an entire year blown on launches (assuming you could even launch that many EELVs in a year, which I seriously doubt). Jupiter, even operated by NASA, could do it for a little over half that, with a much more comfortable launch schedule.That statement is asserted with no proof. I don't accept it at all, unless what he means by "current technology" is NASA's current technology IN PLACE, AS IT IS. Then duh, but that's circular reasoning.You cannot build a plausible Mars mission out of 20-ton chunks, not with current technology and launch vehicles.
Tell me something which cannot be built reasonably efficiently from 20 ton pieces.
Now realize that the IMLEO for this scheme using EELVs is going to go up substantially, partly because lots of small structures are not as volumetrically efficient as a few large ones, and partly because all the individual pieces - including the propellant tankers - need to be (or be carried by) spacecraft in their own right, with maneuvering engines and propellant of their own, RCS, power, avionics... not to mention that spacecraft tend to cost significantly more than their launchers; there's at least another year of your HSF budget gone, and we haven't even touched orbital ops yet, never mind mission support...
You could sort of get around this with an orbital tug, but we don't have one, it wouldn't be cheap, and it would need fueling (you can't do LEO assembly on solar ion). Also, I doubt a single tug would work for an entire mission assembly sequence without repair/refurbishment, so it isn't a given that it would help at all...
Going with an inflatable transit habitat won't help - the DRM already has one. That's where the name TransHab comes from...
Then you have the extra cost of development and testing to get the automated assembly to work right (this sort of thing can get very expensive), the chance that a docking operation could fail and endanger the entire stack (hey, it's happened to the Russians before, and they weren't doing cryogenic propellant transfer)... the extra cost and risk involved in a segmented heat shield/entry aeroshell, because EELVs can't put up a big enough monolithic one (PLF size matters here, of course, but surprisingly mass does too; the aeroshells for the DRM are 40 tons each)...
Falcon 9 Heavy could probably do the launching more cheaply, but then so could Jupiter (they're in roughly the same $/kg class, especially if Jupiter is operated commercially - and if SpaceX can hit their cost targets, which isn't a given). And with Jupiter you don't have the time-consuming and expensive orbital integration problem.
Something like Zubrin's Mars Direct plan would be easier to fit on EELV-class lift - in fact, it's been done - but the guy who did it has admitted that (a) his plan isn't actually "for less" unless a dramatic improvement in launch cost is realized; the idea is to provide an "anchor tenant" for large volumes of small launches and break the current market deadlock, leading to cheaper RLVs, which as others have pointed out could just as easily be done with propellant depots serving spacecraft launched on HLVs, and (b) his plan doesn't work as presented due to aerocapture issues (ie: heat shield size).
There's probably a reason the Augustine commission ended up insisting that heavy lift was necessary for a robust exploration program, even while favouring the idea of commercially-supplied propellant depots (and even while ignoring the J-1x0, which is what makes DIRECT work, and adding a huge margin onto a cost structure that already had significant margins in it, while trying to cram too much other stuff into the budget, so the SD-HLV development ended up stretched way out and dominated by fixed costs. Hey, they were in a hurry...). A number of the Commission members went in thinking EELV-class was enough, but they changed their minds. Medium lift plus prop transfer/depots can give us the moon (suboptimally), but beyond that the high per-launch cost and the headaches of on-orbit integration make it far cheaper and easier to use a heavy lifter, even if you take the HLV's development costs into account.
Now, if you want large multi-MWe nuclear reactors in space, current EELVs will not be enough. Ad Astra's smallest Mars mission requires a 48 mT nuclear power system, and their largest (the famed 39-day transit) needs an 800 mT system... no, you cannot assemble an 800-ton nuclear power plant out of 20-ton pieces in space...
Actually, those numbers are for the Atlas V.Also if you look at the claimThe "EELV" he's talking about is Delta, and he's assuming that the very optimistic claims for Jupiter costs (which of course are entirely speculative ... perhaps I should just be honest and saw "low balled" ... are real)Even with moon missions, if you want more than one per year, Jupiter is cheaper than EELV, and the difference only grows with mission rate.
I think the claim that Jupiter will beat any routinely operating COMMERCIAL launcher on kg/$ to LEO makes no sense. Of course Delta is not really a routinely operating commercial launcher.
ULA's lunar proposal, using the ACES concept, assumed a $10,000/kg launch cost. So that's probably in the ballpark for the upgraded (40 mT) EELVs at the required launch rate for a couple of lunar missions per year (an ESAS-class lunar mission requires roughly 200 mT IMLEO).
DIRECT's cost and schedule estimates have been looked at by The Aerospace Corporation, and have been confirmed to be reasonable. This is not really surprising, since Jupiter is essentially a slightly modified STS stack, meaning the costs are not at all difficult to quantify.
Funny how all the naysayers simply confine themselves to saying the cost estimates are lowballed, without saying why they believe this or providing any alternative estimates...
Yes, as many have pointed out, the cost and schedule estimators for DIRECT are the same as the ones for Ares. That is, they're the same ones that put Ares I IOC in 2015 at 0% confidence, long before the Augustine Commission shone a spotlight on the problem. It was management that was painting the rosy picture. If the guys doing the estimates for DIRECT thought Ares was going to work, why were they working on DIRECT?
Besides, as mentioned above, Aerospace has already confirmed their numbers.
The Shuttle stack is actually pretty efficient when you remove the orbiter. Just because it's bigger doesn't mean it's vastly more expensive.
(Please note that some of the cost estimates on the DIRECT website haven't been updated since DIRECT v2.0, and thus include the J-2X development and RS-68 man-rating and infrastructure changeover costs. DIRECT v3.0 uses all existing engines - RL-10 should be easy to man-rate; they need it for the lander anyway, and since it's for the upper stage it doesn't need to be ready right away - and is thus cheaper and faster to develop.)
One lunar mission per year is one J-2xx and one J-1x0 per year. That's not very many; current estimates are that such a launcher program would cost a bit over $2B/year, without the spacecraft factored in. That's a little more than $10,000/kg, or about how much the EELV people figured on being able to deliver in the ACES lunar architecture. This is probably why the costs for one or two lunar missions per year are roughly similar between the two concepts. After that, the Jupiter starts to get cheaper because the majority of the cost is fixed cost; the incremental cost of a launch is actually quite small relative to its payload.And one per year ... you so sure there would be one per year? One of my biggest objections to the big launchers is that they can't be even remotely economic unless you commit to a rather high launch rate.
Operate the Jupiter commercially and it could start looking even better...
Back to Mars: Basically if you want something sustainable beyond flags and footprints, you need heavy lift, because otherwise it's too expensive and time-consuming to get a mission off.
If the heavy lift is cheap and sustainable enough (Jupiter, especially commercialized, should be; Ares V wasn't), you can then do a bunch of stuff with it - not just Mars. Large telescopes, Europa subsurface probes, manned NEO missions, manned lunar missions (of course), large one-shot space stations...
...
And what's this about teleoperation of robots from Earth? Beyond the moon? You're nuts. It makes me all shivery and antsy just thinking of trying to teleoperate a robot with a 40-minute signal round trip (Mars at its furthest point)... Spirit and Opportunity were probably as teleoperated as is feasible on Mars, from Earth. Even the moon is iffy for real-time work because of the nearly 3-second lag; to really do decent teleoperations work you'd need to be at L1 or L2...
_____________________________________________________________
Last edited by 93143 on Mon Mar 01, 2010 5:24 pm, edited 1 time in total.
-
Heath_h49008
- Posts: 71
- Joined: Wed Jan 27, 2010 9:12 pm
- Location: Michigan
Sort of both. My first statement referred to the automated/self docking/self assembling structures. I thought all that was discussed back in the 1980s when the ISS was supposed to be larger and made of a bunch of Trans-hab modules. (at least in one of the designs)kunkmiester wrote:Huh? You've got me confused, Heath. I'm talking about something that puts itself together, and it sounds like you're talking about a reusable SSTO with significant payload.
The second paragraph was my opinion, (without enough numbers to back up) that if the Skylon TAV/SSTO was a reality, no disposable rocket could ever touch it on a kg/$ basis.
I dont know what exactly you are referring to. I am not the only one with this opinion, though. So shutting my off like that is not fair. If you dont want a discussion with people on the other side, then a forum like this might be wrong place for you.Skipjack? Stay out of it unless you can be constructive, please...
Btw, I do prefer Jupiter to Ares, but I dont like the way the programme would be implemented, which would be simillar to Ares (cost plus).
You do have an argument here. I have to agree. I can see multiple smaller reactors though, or at least larger pieces of a slightly smaller reactor. I do have to agree though, that this is a problem. But then, a Mars mission does not just consist of a nuclear reactor. Also this still does not solve the problem of the high cost of such a mission. Jupiter is IMHO not cost effective enough to make Mars missions more than a "one shot stunt" like Apollo was. That however is not what I (and others) want to see.Now, if you want large multi-MWe nuclear reactors in space, current EELVs will not be enough. Ad Astra's smallest Mars mission requires a 48 mT nuclear power system, and their largest (the famed 39-day transit) needs an 800 mT system... no, you cannot assemble an 800-ton nuclear power plant out of 20-ton pieces in space...
Jupiter is not proven either. I would say that both have at least the same chance of success. History shows though that "cost plus" programmes are more expensive than COTS like programmes. Like it, or not, Jupiter would just be another cost plus programme. That is my biggest problem with it. The other is that it is a big dumb booster with a very low flight rate. Low flight rate = expensive. Give me something to convince me otherwise.Falcon 9 Heavy could probably do the launching more cheaply, but then so could Jupiter (they're in roughly the same $/kg class, especially if Jupiter is operated commercially - and if SpaceX can hit their cost targets, which isn't a given).
Operate the Jupiter commercially and it could start looking even better...
Ask Arianespace. The market for heavy lift is small.
I have promised myself to never use smileys. Sometimes it gets me in trouble...Skipjack wrote:I dont know what exactly you are referring to. I am not the only one with this opinion, though. So shutting my off like that is not fair. If you dont want a discussion with people on the other side, then a forum like this might be wrong place for you.Skipjack? Stay out of it unless you can be constructive, please...
You sure about that? The current plan (assuming they can pull together in the face of Obama/Garver's scorched earth policy) seems to be for a privately owned and operated system. It might be cost plus, but since they're doing this on their own, with NASA as a (fervently) hoped-for customer, it strikes me as a very different way of doing things - more of a freely-competed (as if there's any competition for a 100 mT launcher) market-priced solution. I personally don't have enough information to say. Do you?Btw, I do prefer Jupiter to Ares, but I dont like the way the programme would be implemented, which would be simillar to Ares (cost plus).
Who says you have to throw away your 800-ton nuclear reactor after one mission?Also this still does not solve the problem of the high cost of such a mission. Jupiter is IMHO not cost effective enough to make Mars missions more than a "one shot stunt" like Apollo was. That however is not what I (and others) want to see.
Doing Mars missions, Jupiter's flight rate would be high enough that it would probably constitute a fairly small fraction of the total cost of the mission. It would be less than half in any case, especially if (as the DIRECT team advocates) propellant were freely competed on smaller launchers. Launch to LEO is only part of the cost. Advances have to come from elsewhere to make large-scale human presence on Mars economic, and I continue to believe that if we just sit here and do nothing (no, I don't trust Obama's new plan as far as I can kick it), it's less likely that these advances will occur. People don't care about something we aren't going to do, so why expend effort on it? Our civilization is turned too far inward as it stands...
In my huge post above, I posted a link to a cost comparison by Ross Tierney, to the effect that even at two flights per year, a cost-plus NASA-operated Jupiter comes in at around $10,000-12,000/kg, which is not bad by today's standards. Double the flight rate (to 4 flights per year, easily exceeded by the Shuttle) and you're approaching $7000/kg, which is pretty decent by today's standards. Mars missions would require a much higher launch rate.Low flight rate = expensive. Give me something to convince me otherwise.
Not if NASA gets going on the VSE. There's enough money for that and a robust R&D budget, now that Ares is gone. All we need to do is realize this.Operate the Jupiter commercially and it could start looking even better...
Ask Arianespace. The market for heavy lift is small.
-
kunkmiester
- Posts: 892
- Joined: Thu Mar 12, 2009 3:51 pm
- Contact:
You're probably going to have trouble making one out of 100 ton peices too, though it'll probably be easier.and their largest (the famed 39-day transit) needs an 800 mT system... no, you cannot assemble an 800-ton nuclear power plant out of 20-ton pieces in space...
I'd also have to point out(and agree) that teleoperation outside of LEO is not feasible. This would be the main reason to send men to Mars--no signal lag for the robots; which also brings up the point you could probably save some mass by going with robotic landers rather than putting humans on the surface, though that's the main point of a "flag planting" mission.
Evil is evil, no matter how small