[AR] Re: Falcon Heavy use cases
- From: Henry Vanderbilt <hvanderbilt@xxxxxxxxxxxxxx>
- To: arocket@xxxxxxxxxxxxx
- Date: Thu, 8 Feb 2018 12:52:34 -0700
On 2/7/2018 10:31 PM, John Schilling wrote:
On 2/7/2018 9:13 PM, Henry Vanderbilt wrote:
The most expensive single component of an F9 booster is likely the
Merlin engines. 28 in an F9H, 27 recovered, one expended in the upper
stage. So, 3.6% of the overall vehicle cost there.
Careful, the Merlin Vacuum isn't the same engine as the Merlin 1D. Even
accounting for the shared hardware, being produced (and more importantly
being /tested/) on a much smaller scale, is going to make the upper
stage engine significantly more expensive than one of the booster engines.
Just because a typical cost-plus vendor would cheerfully allow an upper
stage engine's mechanical and procedural details to diverge radically
and expensively from the sea-level version doesn't mean SpaceX would. A
difference in fundamental motivations: Cost-plus means higher cost
equals *higher* profits. Commercial fixed-price, the opposite. SpaceX
demonstrably understands this.
So I'd be seriously surprised if SpaceX allowed the vast majority of vac
Merlin parts and procedures to diverge at all from the sea level
version, for the precise reason you cite, cost.
There's the vac nozzle extension, of course - but that's a passive
radiator, simple formed sheet metal (relatively expensive metal, yeah)
and I'd be amazed if it's involved during test-fires rather than bolted
on after. Simple mechanical testing separate from the engine should
reveal any nozzle extension problems far cheaper than doing upper-stage
engine tests in (expensive!) separate vacuum facilities.
The mechanical attachment provisions for the extension on the cooled
nozzle are likely the biggest single physical part difference - assuming
they don't just put those on all nozzles for commonality.
And ignition provisions may or may not be slightly different between sea
level and altitude lightoff. I'd guess, not much mechanical difference
at all (maybe some software procedural?) given that they relight
core-stage sea-level engines in vac for the initial braking burns.
Bottom line, I doubt SpaceX allowed vac Merlin costs to grow very much.
Which all begs the real question, as far as I'm concerned: F9H cries
out for a high-energy upper stage.
In the best of all worlds, an ACES stage scaled for the F9H would
provide very, very interesting capabilities.
Yes, and now that you've suggested it, I'm going to have to model that
combination. I'll let you handle the politics of a ULA-SpaceX joint
venture.
I'd be fascinated to see what you might be able to share there.
Intuitively, it should make modest difference in F9H payload to LEO,
with increasing payload mass benefit the farther/faster the mission.
As for the politics, the obvious way to bypass that is if the government
is the too-high-volume-to-ignore customer for both SpaceX and ULA. Or
for SpaceX and Blue Origin - a BE-3U might actually be sized about right
for an F9H-lofted upper stage. And Blue also has current hydrogen-stage
experience. Not to mention an existing stage flying in roughly the
right size range - maybe check out performance of a stripped-down BE-3U
New Shepard on top of an F9H also?..
(All this is in the context of a government program aimed at significant
results in relatively few years, of course. In a context either purely
commercial, longer-term, or both, there's far less likelihood of such
multi-vendor booster cooperation - all three mentioned have their own
long-term high-performance integrated booster plans.)
Also, Schilling's three rules of space launch propulsion:
1. It is foolish to use anything but cheap, dense propellants in your
Earth launch stage. You need thrust against gravity, and you shouldn't
much care about weight when it's just cheap rocket fuel and sheet metal.
2. It is foolish to use anything but LOX/LH2 for your orbital insertion
stage. You need Isp to build delta-V, and every pound of "cheap"
propellant has to be lofted halfway to orbit by an expensive booster.
3. It is foolish to use different propellants on different stages of
your rocket, because that makes every bit of hardware and every
operational procedure a complete duplication of effort.
Now go design a not-foolish space launch vehicle. Elon has made his
choice, and in my experience most rocket scientists are fairly stubborn
about which of the three rules is "obviously" wrong or at least less
important than the other two. Rocket plumbers may be more pragmatic, of
course, but I don't take Elon to be a plumber.
Well, Delta 4 certainly illustrates Rule 1. With considerable help from
Aerojet doing a seriously piss-poor job of balancing sea-level and vac
performance in the RS-68, in my view... (Their answer: Neither!)
I tend to agree with Rule 2. Working with LH2 isn't trivial, but XCOR
served as proof it's far short of fluorinated devil-juice.
Rule 3 is I think overstated. I can see the advantages of embracing it
for SpaceX in getting this far, in that it reduced demands on their
finite engineering/development bandwidth to a (barely) sustainable level
while still allowing them to field a (very) commercially viable initial
low-orbit transport system with useful (albeit not optimal) GEO capability.
But I think SpaceX's apparent continuing embrace of carved-in-stone Rule
3 may well provide their currently-behind rivals with competitive
windows they can occupy and expand. We'll see.
Henry V
Other related posts: