Re: Skylon SSTO
- From: Pat Flannery <flanner@xxxxxxxxxx>
- Date: Tue, 23 Jan 2007 17:04:03 -0600
OM wrote:
...This begs the question: which should come first - consulting Google
or asking Henry? :-)
Here's the discussion in question: *http://tinyurl.com/ynmzvh
"I wrote:
>Incidentally, for Hotol fans, note that the May 1993 issue of Spaceflight
>(the BIS's other journal) broke the story on how the original Hotol engine
>worked -- it has been declassified...
A couple of people have asked me to summarize this, since they don't have
easy access to BIS journals, so herewith a brief overview...
The article is sort of half writeup and half interview with Alan Bond (the
engine man) and Bob Parkinson (HOTOL project manager).
They wanted to build an SSTO, and they thought rocket SSTO was too hard.
Airbreathing SSTOs split into three main categories:
1. turboramjet to Mach 5, rocket thereafter
2. liquid-air engines, aka LACE, again moving to rocket circa Mach 6
3. scramjets
Turboramjets are fairly mundane technology, but they are impossibly heavy
for practical launch systems.
LACE systems use on-board liquid hydrogen to liquefy air as the initial
oxidizer. Thrust:weight ratio is better, but liquefying air requires
much more hydrogen than you can then burn with the liquid air, and doing
anything with the extra is difficult.
Scramjets look promising, but... First, they only start working at
hypersonic speeds, so you need something else to get them up to that
speed, and then of course you need rockets again later: "the Smithsonian
in orbit -- one example of everything". It turns out that the penalties
incurred make scramjets unattractive if you need a major rocket phase
afterward, which is why NASP wanted to air-breathe almost all the way.
Unfortunately, any particular scramjet shape is optimal for only one
speed, so a scramjet becomes a complex variable-geometry device, and
it is very difficult to make the geometry variable enough to function
from Mach 2-3 up to Mach 20-25. The HOTOL people concluded that
Mach 13-15 is about tops for a scramjet that will also start at a
reasonably low speed, and this makes them relatively unattractive.
Second, "scramjets have an inherent contradiction -- they demand good
and rapid mixing between the air and the hydrogen, but without disturbing
the supersonic airflow at all!".
Third, intake area must vary over a huge range if the engine is to run
over a wide range of speeds. To keep airframe loads bearable, altitude
must rise in a way that thins the air out faster than the speed increases,
so the required intake area gets larger and larger.
Finally, scramjet research is extremely difficult. Confidence in computer
modelling is not high, wind tunnels simply are not up to the job, and
developing the technology with research aircraft would be lengthy and
costly.
"Scramjets do not have an intrinsic performance benefit and even if they
did the cost of engineering the hardware required would be much greater
than the other options. The demise of the X-30 NASP may be a belated
realisation of this fact." They may be workable for high-speed *cruise*,
but they are not suited to rapid acceleration, i.e. a space launcher.
So they went back and took another look at LACE cycles. Bond came up
with three new wrinkles.
First and most important, you don't have to liquefy the air. Merely
cooling it to a very low temperature (just above liquefaction point)
and compressing it there gets you most of the benefit and requires
much less cooling.
Second, heat exchangers can be made much lighter and more efficient
than previously thought. (No details given.)
Third, you still need more hydrogen for the cooling than for the
burning, and you *dump the extra hydrogen overboard unburnt*. Any
attempt to use it (with one exception, see below) turns out to show
a net loss.
"Any hybrid engine must end up being a very efficient rocket for most
of the flight. I began with a good rocket engine and made it a bad
air-breather. Everybody previously had done the reverse."
The cycle is a bit complex -- the article has a diagram and detailed
explanation -- but basically hydrogen cools the incoming air, which
is then compressed, used to cool the chamber, and burned. 2/3 of the
hot hydrogen drives the compressor turbine and is then dumped overboard;
the remainder goes to the chamber to be burned. Airbreathing thrust
is about half rocket thrust. At Mach 5-6, when the momentum loss from
decelerating the air to near-zero relative speed is starting to wipe
out the gain from airbreathing, and the intake is starting to get
awfully hot, the system switches to rocket operation using LOX.
Since the system starts out acquiring air using a compressor, it can
run even at zero airspeed, eliminating any need for other engines and
allowing much testing to be done in normal jet-engine test facilities.
During detailed intake design, it was realized that the problem of
varying intake area could be simplified if there were provisions to
"spill" some of the incoming air into an exhaust duct, away from the
compressor, at low speeds where a big high-speed intake takes in too
much air. The spill duct is shaped as a nozzle, to expand and accelerate
the air and recover some of the momentum lost in acquiring it. At this
point, a small trick was added: some (not all) of the hydrogen being
dumped is injected into the spill duct, and it runs as a simple ramjet.
This is basically just meant to overcome the losses in the intake and
spill duct; any useful thrust is a bonus. "The idea was that the
spill ramjet did not have to be very effective, and was not something
requiring extensive ground test development facilities. Instead we
hoped that it would get 'tuned' during flight development, and generally
give us a bit of extra performance and hence margin." It would run
between perhaps Mach 0.5 and Mach 4.5.
Hence HOTOL's appearance from the back. Below the fuselage is an intake,
with the spill ducts at its rear end. On the back of the fuselage itself
are the main exhaust nozzles, with small dump ducts flanking them.
The bottom line, alas, is that after the British government lost interest,
Rolls-Royce assessed the potential market and concluded that even under
optimistic assumptions it would not sell enough engines to pay the costs
of development. So the RB545 is dead. Alan Bond is still pursuing
variants of the idea, and in fact there's an article in the same issue
talking (a little bit) about his newer design concept, Skylon. I'm not
going to summarize that one; this has already taken enough time!
--
Belief is no substitute | Henry Spencer @ U of Toronto Zoology
for arithmetic. | h...@xxxxxxxxxxxxxxx utzoo!henry "
Pat
*
.
- Follow-Ups:
- Re: Skylon SSTO
- From: Neil Gerace
- Re: Skylon SSTO
- References:
- Re: Skylon SSTO
- From: OM
- Re: Skylon SSTO
- Prev by Date: Re: Skylon SSTO
- Next by Date: Michael Griffin on Why Explore Space?
- Previous by thread: Re: Skylon SSTO
- Next by thread: Re: Skylon SSTO
- Index(es):
Relevant Pages
|
