On Sun, Sep 25, 2016 at 10:07 PM, Henry Spencer <hspencer@xxxxxxxxxxxxx>
On Wed, 21 Sep 2016, Keith Henson wrote:
Reaction Engines ran their cost estimates out to a million flights per
year, which is where you need to be with a serious power satellite
project. I have never seen anyone do the same with VTVL rockets.
You don't see it done very often with HTHL airbreathing launchers either.
The reason is that such estimates are sales tools, and people don't think
they can defend such huge market models as realistic. (Max Hunter used to
urge that if you're going to cut your cost-vs-volume plots off at some
arbitrary modest volume, you should at least show the marginal cost of an
extra flight as well, to give some idea of the cost at very high volumes.)
It's not that there's anything wrong with the technology at high volume,
it's just that promoters fear nobody will take such numbers seriously.
I know 4-5 runways will accommodate this traffic. Any guess on how many
pads it would take and what separation between the pads is needed?
Much depends on assumptions about operations. With optimization for high
volume, there would be none of this nonsense of assembling the vehicle on
the pad; if the vehicle does need reassembly, rollout would happen only
when ready to fly. (Note that Ariane 5 already operates that way, and
both KSC and the Atlas V launch facilities were designed to do that.)
You might still want to do the main fueling on the pad/stand rather than
before rollout, just to minimize the mass being moved, but that's not a
lengthy process if done right by automated machinery -- fueling an Atlas
ICBM took only a few minutes.
If pad/stand design puts a high priority on long life and low-maintenance
operation, there's no reason not to operate it at a pretty brisk tempo.
Baikonur's Zenit facilities were designed for a salvo capability with each
pad launching every 90 minutes, and that's by no means the limit of what's
possible.
I'd pencil in a 10-minute launch cycle, and the Mk. 2 would probably
improve on that considerably. Given that a million launches a year equals
about two per minute, that still might mean ten or twenty pads/stands, but
they'll be much cheaper to build than Skylon's special runways.
Separation will be about the same as the separation needed for the Skylon
runways. A well-debugged VTVL will be no more prone to blowing up than a
Skylon, i.e. the possibility must be considered but it should be rare.
If you're going to require (say) that the VTVL assembly buildings should
be miles from the pad/stand for safety, you should impose the same
requirement on the Skylon facilities.
(At that launch pace, with either type of vehicle, every day you get more
operations experience than the entire history of spaceflight up to now.
As with high-performance aircraft, development would include hundreds or
thousands of test flights before the first valuable payload goes aboard,
and the ramp-up to full-speed operation would include a million or more
proving flights at gradually increasing tempo. Early test flights, the
first few flights of each new flight vehicle, and any flight with hardware
changes or major maintenance or other novelties, would use separate test
facilities; the high-throughput operational facilities would be used only
by operational vehicles.)
