It’s always interesting when an outsider comes into the space community to try to educate us on why we don’t have low-cost space access, particularly when he’s clearly unfamiliar with the history and literature on the subject. It happened again this morning at The Space Review. I was going to write something more extensive about it, but Clark beat me to it.
19 thoughts on “Misguided”
Comments are closed.
We don’t have low cost space access because the Earth’s gravity is too high.
Fix that, and…
We don’t have low cost space access because the Earth’s gravity is too high.
Cue in Eugene Podkletnov and his anti-gravity device.
Or not.
One thing that bugs me is that, while I believe like Clark that it is better to sacrifice performance for costs, SpaceX has been steadily doing the opposite. They switched to regenerative cooling and Falcon 9 will have pump-fed engines on all stages.
They probably got some valuable experience with the pressure-fed engines that made them more confident with later designs, but it seems pump-fed regeneratively cooled engines are too good to dismiss.
I wonder if, like I heard the Russians say, a staged-combustion cycle is also something which should be standard, despite claims otherwise by several people in the business who deem it too complex.
Given the need for re-usability, I wonder what the most practical rocket size would be (in payload lbs/tons) for reaching orbit if said rocket could launch 2/day. Smaller payloads would require more orbital assembly, but maybe the right rocket would make it practical to launch a Mars mission one crew member at a time if average cost was low enough. Bigger payloads would allow the launch of complicated, single-part big stuff (for instance, a nuclear reactor core), but the rocket would fly less often.
I guess my question is: at what point would orbital assembly become too big of a headache? And how much could we improve orbital assembly to make small payloads more feasible for more types of missions (allowing more launches, lowering cost further)? If you imagine a Bigelow module with a couple arms snapping pieces together, and if you had ~6 rockets launching 2/day each, I bet you could get parts to the Bigelow module just as quickly as the work crew could put them together (assuming assembly possible at all, as I assume might not be the case with power reactors or large structural parts capable of withstanding acceleration).
I think the issue is that McGowan and Lindsey (and Rand presumably) are talking past each other because they start out with fundamentally different premises.
Lindsey assumes that RLVs are within the current state of the art and wants performance scaled back to find the optimum cost and/or reliability.
McGowan assumes that RLVs are not within the current state of the art and wants technological improvements to bring them within the state of the art.
Both positions are logical. Only one premise is correct.
I have a sneaking suspicion that “Lindsey” and Rand are more on top of what the state of the art is than McGowan. Do you have some reason to believe that he is more so than us?
One thing that bugs me is that, while I believe like Clark that it is better to sacrifice performance for costs, SpaceX has been steadily doing the opposite. They switched to regenerative cooling and Falcon 9 will have pump-fed engines on all stages.
Godzilla, why do you think regenerative cooling or pump-fed engines are costly? I don’t know about the latter, but the Merlin 1C engine with regenerative cooling probably is less costly per kilogram of payload delivered than their old ablative engines. Better performance and lower cost sometimes go together.
Hi Jim:
“I think the issue is that McGowan and Lindsey (and Rand presumably) are talking past each other …”
No, I’m talking directly to Mr. McGowan and telling him that the good idea that he is presenting – incremental, trial-and-error development – has long been appreciated in the NewSpace industry and is what many of them are implementing.
“Lindsey assumes that RLVs are within the current state of the art and wants performance scaled back to find the optimum cost and/or reliability.”
Not exactly. The Kistler K-1 sort of fly-twice-a-month RLV is not only within today’s state of the art, it was 10 years ago as well. If you know of a fatal flaw in the design, then you spotted what numerous design reviews failed to find.
On the other hand, I don’t think an orbital, fly once or twice a day type of RLV is in hand, at least not for what I consider reasonable development and operations costs. I think the best way to learn how to build that type of RLV is to start with suborbital and do incremental, trial-and-error development.
Karl Hallowell,
Care to explain why you think “the Merlin 1C engine with regenerative cooling probably is less costly per kilogram of payload delivered than their old ablative engines”?
It has seemed to me over the years that the growth in announced Falcon pricing (and presumably costs) has been growing faster than inflation, tankage materials, or payload growth can account for.
I’ve been assuming the increasing engine complexity was driving up costs, even at the risk of self inflicted failures like the third Falcon I flight.
I am only an “enthusiast”, of course. Please enlighten me.
Rand: “I have a sneaking suspicion that “Lindsey” and Rand are more on top of what the state of the art is than McGowan. Do you have some reason to believe that he is more so than us?”
I think this is a subject where reasonable, well informed people can disagree.
Clark: “Not exactly. The Kistler K-1 sort of fly-twice-a-month RLV is not only within today’s state of the art, it was 10 years ago as well.”
Whether the K-1 would have worked as advertised is not known.
“If you know of a fatal flaw in the design, then you spotted what numerous design reviews failed to find.”
How many design reviews has Ares I had?
I’ve been assuming the increasing engine complexity was driving up costs, even at the risk of self inflicted failures like the third Falcon I flight.
I wouldn’t assume that. It’s possible, but there are other explanations. I would attribute it to an increasing realism in early naive and optimistic cost estimates as they get more experience with actually building and flying hardware.
How many design reviews has Ares I had?
Depends on how you count. If you mean PDR, about a half.
Jim: “Whether the K-1 would have worked as advertised is not known.”
But it is quite safe to say it did not violate any physics and that none of the components were near the bleeding edge of technology. Perhaps in operation maybe it would have taken 3 weeks instead of 2 weeks to turn around or some other operational shortcoming. But there were no great fundamental unknowns about the design. I don’t hold the K-1 out as an ideal RLV, just that it is a proof of principle that even a bunch of conservative Apollo era NASA engineers can produce a viable design.
For the several billions thrown away on SLI and OSP, NASA could have funded K-1, even with large overruns, as well as a couple of competing TSTO designs.
“How many design reviews has Ares I had?”
As Rand indicates, it’s highly questionable as to whether the Ares I even passed its PDR. A lot of serious design issues have been raised about Ares I and need to be settled before they start building hardware. The K-1 project had moved long beyond its design reviews and reached the point where the hardware for the first was 70% built before they ran out of money.
“But it is quite safe to say it did not violate any physics and that none of the components were near the bleeding edge of technology.”
That could have been said about the Shuttle, Ares I, Roton, etc.
“I don’t hold the K-1 out as an ideal RLV, just that it is a proof of principle that even a bunch of conservative Apollo era NASA engineers can produce a viable design.”
It’s curious that no one holds the Shuttle up as such a proof of principle. I suspect the same scenario would have played out if the K-1 had progressed to operations. It would have worked – after a fashion – but it would have shown no improvement in cost or reliability over contemporary ELVs because refurbishment would have been more expensive than hoped for and the number of refurbishments would have been fewer than hoped for and the flight rate would have been less than hoped for. And the same enthusiasts who once held it up as a “proof of principle” would quickly disown it – “It’s not a reusable launch vehicle – it’s a rebuildable launch vehicle” – “You can’t draw any conclusions about RLVs from K-1” – “The K-1 was flawed from the start” – “It was a design by committee of former Apollo dinospace types” – etc, etc, etc.
That could have been said about the Shuttle, Ares I, Roton, etc.
Again, Kistler was much further along in the development cycle than Ares or Roton are/were, with major portions manufactured.
It’s curious that no one holds the Shuttle up as such a proof of principle.
Because it’s not reusable? It throws away the tank, and the SRBs aren’t reused, they are rebuilt.
It would have worked – after a fashion – but it would have shown no improvement in cost or reliability over contemporary ELVs because refurbishment would have been more expensive than hoped for and the number of refurbishments would have been fewer than hoped for and the flight rate would have been less than hoped for.
I’m not sure why you “suspect” that, though the flight rate probably would have been an issue. Kistler didn’t suffer the Shuttle’s problem of spending too little up front, sacrificing later operational costs. If anything, they had the opposite problem, spending too much up front to make sure that it was operable, and then running out of money to complete it after the market collapsed.
I don’t see any basis for your suspicion other than the single flawed (government-developed) example of the Shuttle. We’ll never know now, unless Rocketplane manages somehow to scrape up the money to resurrect it (which Chuck Lauer says that they’re still trying to do). More money in COTS might help.
“That could have been said about the Shuttle, Ares I, Roton, etc.”
No it could not be said about those projects. As Rand indicates, the K-1 was still holding to its goals by the time construction had started. When the Shuttle concept was first discussed there was the talk of once a week flights but long before metal started being bent it was clear that Shuttles would not come even close to that goal. It’s not as if NASA started flying and suddenly discovered a bunch of flaws that made turnarounds long and labor intensive. This was clear long before Columbia first flew.
Rotary Rocket never got to the point of finalizing an engine design for the Roton, much less building one. Kistler just went out and bought existing Russian engines that gave them the performance they needed. Can’t think of a sharper contrast in situations.
“It’s curious that no one holds the Shuttle up as such a proof of principle.”
I didn’t mention the Shuttle because I thought the big differences mentioned above between it and the K-1 were obvious.
Current aerospace engineering isn’t at the Star Trek stage but give it some credit. Sure, a vehicle project that reaches the hardware stage might miss a given operations goal by some percentage but it is extremely unlikely they will miss by a mile and even less likely that there will be some terrifically fatal shortcoming.
“It would have worked – after a fashion – but it would have shown no improvement in cost or reliability…”
That is exactly the sort of self-fulfilling attitude, rampant in NASA/DoD and in mainstream aerospace, that has made funding of projects like K-1 so difficult. If vehicles with improved designs never fly, one can always claim that they will fail. Furthermore, we do not gain any trial and error learning that contributes to progress towards the goal of low cost space transport.
There never would have been a DC-3 if everyone was sure at the time that any attempt to improve on the Boeing 247 might have worked – after a fashion – but would have shown no improvement in cost or reliability…
I honestly cannot fathom how K-1 would have been that much of an improvement. Lets face it, those Russian engines were state of the art when they were designed, and still are pretty impressive today, but there were improvements in performance and reusability since then. The RD-171 and RD-180 are better engines using the same combustion cycle. They are used in two superb rockets: Zenit and Atlas V. However, neither of those rockets have provided order of magnitude reductions in launch costs. I also share the thoughts echoed here that it would have been a rebuildable, rather than reusable rocket. You can hardly have a reusable with low turnaround times when you are using parachute recovery.
You can hardly have a reusable with low turnaround times when you are using parachute recovery.
I don’t think that parachute recovery impacts GTAT that much (unless one is looking for mere hours). The problem with it for me is comfortable landing and reliability.
“I honestly cannot fathom how K-1 would have been that much of an improvement.”
If they achieved two week turnaround, that’s obviously a big improvement. If total development cost up to first flight had been in the $1.5B range (including the $800M Kistler spent to get to the 70% constructed level) as RpK expected, that’s again a big improvement over the various govt. rocket project costs.
“Lets face it, those Russian engines were state of the art when they were designed, …”
The engines were good enough and were cheap. The point is not to try to maximize performance but to minimize cost. I’m sure later vehicles could take advantage of improved engines.
“neither of those rockets have provided order of magnitude reductions in launch costs.”
I would never expect those sort of ELVs to provide more than minimal cost reductions. Elon says that even with SpaceX’s much lower development costs they can only reach his order of magnitude lower cost goal for the F9 if they achieve reusability.
“You can hardly have a reusable with low turnaround times when you are using parachute recovery.”
Depends on whether you consider ~2 weeks to be low turnaround time or not.
Godzilla – SpaceX hasn’t really done anything yet, their cost keep creeping up and their schedule keeps creeping back. Systemic problems indeed. Flight rate is the key. If you want to fly things to orbit cheaply, you need to start small and launch a lot of them. SpaceX’s attempt to reinvent the wheel is driven in large part by a mission of interplanetary travel in the nearish future; which just doesn’t (seem to) jive with smallsat launches in the much nearer future. In effect, Musk is trying to circumvent the learning and market buildup that one might experience by launching lots of <1 to 50 kg payloads by throwing money at the problem. I wish him luck, but so far I haven’t seen a whole lot of difference between SpaceX and the other current players. It’s not even the technology that worries me – from what I have seen the Merlin engine is a terrific piece of hardware, and reasonably inexpensive too. It’s more the fact that the company can’t turn out rockets when they say they will. You can’t get the benefits of high flight rates – the opening of new markets using other carrots besides price elasticity – if you can’t fly on schedule. Companies and governments that need launch ASAP for things like battlefield surveillance, pre-bid and pre-exploration mineral extraction surveys, disaster reconnaissance – are not going to trust a company that can’t even launch within a published quarter until they demonstrate that ability. I wish them luck, but they have a ways to go.
It’s more the fact that the company can’t turn out rockets when they say they will.
Schedule slippage is not the same thing. Add a demo flight may slip the schedule but doesn’t mean they ‘can’t turn out rockets’… actually I’m pretty impressed with how they’ve turned out rockets, but more impressed by how they’ve geared up to do more in the future.
This also explains the increase in price. It’s not because of naive pricing before, it’s because they’ve radically increased their capacity so the increased price covers that. Down the road this will lead to much greater efficiency. I’m still waiting for a chance to own stock in what looks to be a survivor in the sector.