Dale Skran has a review, that mirrors a lot of my own concerns with it:
There is no discussion at all that the prospect for increased traffic to LEO for all purposes, including tourism, might lead to significantly lower costs; or that it may lead to reusable spacecraft with superior operational characteristics relative to existing vehicles or the SLS. This glaring absence seems remarkable given the stated goal of SpaceX to develop just such lower-cost, reusable craft, as well as their considerable progress in this direction. Of course, the efforts of SpaceX, Virgin Galactic, Blue Origin, XCOR Aerospace, and others to greatly reduce launch costs may all fail. However, the NRC report is based on the unstated assumption that over the entire period considered, all the way out to 2054, there will be essentially no progress in rocketry other than that funded by NASA exploration programs, and that for the entire period the SLS as currently envisioned will remain the preferred method for Americans to reach space. It is difficult to imagine a more unlikely foundation for the planning of future space efforts than this.
It is extremely myopic, and therefore of little value, but it was probably doomed to be so by its charter.
probably doomed to be so by its charter
If you start with the assumption that your pet must be a cat, you will get a certain result. You should not expect that it will bark or behave like a dog.
If you start which the assumption that space exploration should be done by the government, you will also get a certain result. That result will look nothing like terrestrial exploration, which was mostly (though not entirely) a private enterprise.
To paraphrase Clarke’s Law: Any sufficiently advanced government space-exploration program is indistinguishable from Apollo.
Skran errs slightly in saying the “NRC report is based on the unstated assumption that … the SLS as currently envisioned will remain the preferred method for Americans to reach space.” The assumption is quite clearly laid out on page 4-8, section 4.2.2.1: “For simplicity and consistency of presentation, the analysis in this report of all of the DRMs has presumed the use of SLS as the launch vehicle.”
“Simplicity and consistency of presentation” are lousy excuses for ignoring alternatives to SLS, particularly given the grim conclusions reached with SLS.
Well, to be fair, even SpaceX’s plans don’t seem to take into account the possibility that SpaceX will be successful.
Reusable rockets that can fly again the same day, or behemoth rockets that take people to Mars without on-orbit assembly, choose. If you have the former operational, the latter doesn’t make a heck of a lot of sense.
SpaceX’s BFR seems to be intended to support not on an occasional Apollo-style Mars sortie but colonization. Musk speaks of tens of thousands of people per year. I can readily believe that at that rate, a rocket much bigger than Falcon Heavy saves money.
They do not need to choose when they can do both. Successful companies in highly innovative technological fields do this all the time it is called hedging your bets. Besides one thing is a long term plan, which can change, and the other is what they actually do have. Which is cheap expendable rockets. Nothing prevents using the MCT engines on a smaller single engine rocket and that is quite likely what will happen.
In the long term SpaceX needs the bigger rocket so they can convince the USAF and the NRO that they can supply ALL their needs. i.e. that they do not need to remain hostage to ULA.
They need a Delta IV Heavy class vehicle and I doubt Falcon 9 Heavy will be as cheap or reliable as they need in the long term. I mean I am convinced SpaceX will prove it can work well enough. I expect Falcon 9 Heavy will have similar reliability to Proton i.e. 90% rather than 99% like Soyuz. Someone will pine that is not good enough even though it is good enough for a lot of comsat operators with satellites as expensive as those the military puts up. So in the long run SpaceX needs something with less engines than F9H with Delta IV Heavy capacity. That 1 Mlbf engine proposed for the MCT is a necessity if they want to fight for 100% of the US institutional market.
Once they do have that engine what will happen for a Mars or Moon rocket, if they ever make one, may change dramatically. SpaceX has proven they are not afraid of attempting to use unconventional technology if they think it will make things cheaper or better. For example I doubt a cheap Mars rocket would use conventional chemical propulsion in the last stages.
Completely disagree. You still seem to be in thrall to the quite incorrect notion that one big engine is inherently more reliable than clustered smaller engines. Wrong. Having proven the value of fault-tolerant engine clustering, I expect SpaceX to stick with it. In terms of mission success likelihood, Falcon Heavy’s booster stages will not be less reliable than Falcon 9’s booster, but more reliable. Falcon 9 doesn’t get into mission failure territory until it loses three of its nine 1st stage engines. For FH, this threshold is seven out of 27. For any given failure probability for Merlin 1-D engines, it is vanishingly less likely that seven will fail than that three will. Even three failures on a single mission is a multiple-sigma probability.
The single second stage engine is a far likelier cause of mission failure for both FH and F9 than are their heavily-clustered first stages. For reasons of both greater reliability and greater throw weight, I can certainly see SpaceX, at some point – probably not too far off – offering a high-performance upper stage based on two or three Merlin 1-D vacuum engines rather than just one. In combination with a stretch to accommodate more propellant, this would be especially helpful as a performance boost for the Falcon Heavy, likely making it a match or near match for the Block I SLS. It also seems that SpaceX could easily engineer and fly this upgrade years before SLS’s first mission.
Then there’s reusability to consider. Engine clustering also figures into SpaceX’s strategy here. It’s far easier to achieve deep throttleability on a booster stage by simply shutting down whole engines than it is to make a single big engine deep-throttleable. Going to a one-big-engine expendable model would be two steps backward for SpaceX. Not going to happen.
There’s also a time-line consideration here. SpaceX is going to continue its attempts to first “soft-splash” then to soft-land a Falcon 9 1st stage on most of its remaining 2014-manifested missions. I rate their chances of a feet-dry return to launch site by year’s end fairly good. The Raptor engine development program will take longer than this to complete. With practical reusability in-hand earlier, why would SpaceX later elect to build a completely different Falcon 9-class booster stage lacking multi-engine reliability, reusability potential or even propellant commonality with the existing F9 upper stage and which would also lack EELV certification? This isn’t a hedged bet, it’s just bonkers.
I also have no idea what you’re on about with your comments about SpaceX needing “a Delta IV Heavy class vehicle”. Falcon Heavy, when it flies next year, will roughly double Delta IV Heavy’s lift capacity – 53 vs. 27.6 tonnes to LEO, 21.2 vs. 13 tonnes to GTO. As soon as FH flies three successful missions and gets EELV certified, it’ll handle anything in the DoD/USAF/NRO stable – in pairs if necessary. ULA, arguably, might need a Falcon Heavy-class vehicle, but SpaceX will shortly have Delta IV Heavy very thoroughly covered. SpaceX is fully capable of eating 100% of ULA’s lunch without reference to Raptor or anything built around it.
Your doubts about Falcon Heavy’s reliability I have already dealt with. Your grasp of the statistical fundamentals of reliability engineering I simply find defective. Your expressed doubts about FH’s affordability, I simply find baffling. The base price of a Falcon Heavy launch is on the SpaceX website. It’s $85 million. This is less than 25% of what ULA charges for a Delta IV Heavy mission. Even doubling this price to account for government-specific rigamarole would be a huge bargain compared to ULA prices. Do you think SpaceX is lying about this posted price? Whence comes this unaccountable skepticism?
As to what SpaceX will use to get to Mars, that’ll be a BFR or RBFR (Really Big Frickin’ Rocket) based on – most likely – one or three reusable 1st stage cores each having nine Raptors pushing 8 to 10 meter diameter tankage uphill. Second stages will be multi-Raptor designs too. That allows for seriously impressive throw weights into LEO. It also allows one to fire up the Raptors again for Trans-Mars Injection. If SpaceX decides to pursue something with significantly better Isp to cut the transit time way down, the only two approaches that look feasible are a nuclear thermal engine based on a thorium reactor or a VASIMR engine powered by Polywell fusion. Both would require significant long-leadtime R&D. The latter alternative would probably involve acquiring both Ad Astra and EMC2 for their technology.
I think you drank too much of the koolaid. I know a fair bit about complex systems engineering. For something like a rocket, where a blown engine can imply loss of vehicle, the reliability statistics do not work the same way as on a system where something like that is impossible. The Russians had the same idea with the N1 and it was an utter failure. Granted they had a lot more issues than just the fact that they used a lot of engines. They never bothered building a test stand and had all sorts of quality control problems. If you look at SpaceX’s design approach they pre-emptively did a lot of things that the Russians did with the N1F to prevent some N1 failures modes like adding filters inside the propellant piping. Not to mention that they bothered building a test stand where they could test not only the first stage but they simulated the whole rocket on the test stand too. Still F9H will never be as reliable with 27 first stage engines as F9 with 9 first stage engines and you do not need to know college level statistics to figure out that much. Even if it had no propellant crossfeed system. Which it does.
Yes F9H will be more expensive than a system with Raptor because system costs grow with the number of parts on the system. More engines = more parts you need to build, assemble, and test.
Raptor should also have enough performance that it will make it easier to make a SSTO in the future, with engine upgrades, if SpaceX ever decides they want to do something like that. It may require switching to an aerospike or TAN nozzle. It may require going tripropellant LOX/LH2/Methane. But the possibility will be there.
As for the last stage of the MCT I think the most viable alternatives are solar-thermal and solar-electric. Neither of which is particularly expensive in terms of R&D.
The whole goal should be low-cost access to orbit, period. Government won’t be providing that, so we desperately need a free market in spaceflight.