h a l f b a k e r y
The mutter of invention.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
[FOR PEOPLE IN A HURRY or diagnosed with less than a week to
live. Use a modest-sized pressure vessel between the main tanks
and the engine, so you can pressure-pump the fuel in gulps
without turbopumps or high-pressure main tanks.]
So, we want to build a rocket. It's going to be a liquid-fuelled
rocket (eg, kerosene + liquid oxygen), because these are the most
efficient, mass-to-thrust wise.
The first bit is easy. We build a spherical combustion
with inlets for fuel and oxidiser at the top, and big hole at the
bottom. The big hole leads to the rocket nozzle, which is where
flames come out.
So far so good.
Now, in order to work efficiently, the combustion chamber has to
operate at very high pressures (and temperatures), since this
pressure is what provides the thrust. So, we have to bite the
and make the combustion chamber out of something very strong
heat-resistant. This is expensive and heavy, but there's no way
The only remaining problem is that we have to get the fuel and
oxidiser into the combustion chamber itself, and to do so at a
healthy rate. Here we encounter a problem: if the combustion
chamber is at high pressure (typically 100 atmospheres), then we
have to be able to push the fuel and oxidiser into it against that
What we would normally do is to have a couple of pumps (one for
fuel, one for oxidiser). These pumps, however, have to deliver
and oxidiser at very high pressure and quite high flow rates, which
means they are going to be expensive and heavy. We can live with
for something like the space shuttle (where their weight and cost
relatively small compared to the total), but not for a smaller
since they become more expensive and more heavy (relatively
speaking) as the rocket gets smaller.
We decide to use a pressure-pumped design instead. All we do is
pressurize the fuel and oxidiser tanks, and connect them directly
the combustion chamber (maybe via an on/off valve). Great!! No
complex pumps - we just use gas pressure (eg, helium) to push the
liquids into the combustion chamber.
The problem with this is that now, *everything* has to be strong
enough to withstand 100bar of pressure. The fuel and oxidiser
have to be pressurised to 100bar, to push the stuff into the
combustion chamber. A regular Coke can will take, maybe,
A large fuel tank capable of taking 100bar will be very heavy.
The V1 flying bomb got around this problem by using a pulse-jet,
which solves a similar problem by pulsing the propellant. Fuel is
squirted into the combustion chamber at low pressure (easy), and
then ignites; a check valve stops the pressure wave from pushing
the fuel lines. After the ignition cycle, the pressure drops and
fuel is injected, and so on. The engine is then driven by a series
pulses, not so different from a normal internal combustion engine.
This type of pulsed rocket is not very efficient. It's driven by a
series of explosions, with wildly fluctuating pressures, so nothing
really ever optimal. Apart from anything else, the combustion is
only happening for a fraction of the time.
Enter the Gulper Rocket.
Here's* one standing in front of you, stripped down so you can see
At the bottom, we have the usual combustion chamber, with a
shaped nozzle underneath it.
Above that are two spherical pressure-tanks, side by side. One is
fuel, the other is for oxidiser. Each is about the same size as the
combustion chamber, and a pipe leads from the bottom of each
pressure tank into the combustion chamber. In each pipe is a
non-return valve, allowing fuel to flow into the combustion
but not back. This valve has a spring bias to it; so, liquid at low
pressure is held back; liquid at high pressure pushes open the valve
and flows; but a reverse pressure shuts the valve securely. No
actuators are needed.
Now, above the pressure tanks are the main fuel and oxidiser
These are huge (with perhaps a few hundred times the volume of
pressure tanks) and full of liquid fuel and oxidiser. However,
only strong enough to contain the liquid, and therefore are
extremely light (just as on a regular rocket with fuel pumps). Big
pipes connect each of these main tanks to the corresponding
tank, and a simple one-way valve prevents back-flow.
Now, here's how it works. First, the main tanks are pressurized,
only slightly - a few bar. In fact, the pressure from the boiloff of
the liquids can be used to provide this pressure, or it can come
a separate helium tank. This modest pressure quickly pushes fuel
and oxidiser into their respective pressure tanks, until these are
but doesn't have high enough pressure to open the valve between
pressure tanks and the combustion chamber.
Once the pressure tanks are full, they are driven to a very high
pressure (say, 100bar); there are various ways to do this and it's
Now, the high pressure in the pressure tanks slams shut the non-
return valve connecting them to the main tanks. At the same
it overcomes the springy resistance in the valves connecting the
pressure tanks to the combustion chamber, and fuel and oxidiser
flood in and are ignited.
Quite a long whoomf, of course (several seconds), because it's
the reservoir of fuel and oxidiser in the pressure tanks.
When the pressure tanks are empty, the fire goes out. At the
time, the flow of pressurizing gas to the pressure tanks is shut off.
There's now no high pressure in the combustion chamber or
else, and the cycle starts again, with the modest pressure in the
main tanks refilling the pressure vessels ready for another burn.
So, in effect, we have created a pressure-pumped rocket, but one
which only a portion of the fuel/oxidiser needs to pressurized at
one time. This is better than a pulse-rocket, since each "pulse"
last for several seconds (thanks to the reservoir provided by the
pressure tanks), lighter and simpler than a pump-fuelled rocket (I
hope), and also lighter than a pure pressure-pumped rocket in
the entire fuel and oxidiser tanks must be capable of withstanding
*you are now 50% of the way through this. Bear with me.
||Can you explain how pumping pressure into the little tank uses less uggg than pumping fuel into the combustion chamber?
||It doesn't use any less uggggg.
||The point of the pressure-tank is that only it (and not the
main tank) needs to able to withstand the pressures which
prevail in the combustion chamber itself. You need the
same amount of uggggg to get fuel into the chamber,
whether it's directly from high-pressure main tanks (which
are heavy), via a high-pressure pump (which is costly) from
low-pressure tanks, or via an intermediate gas-pressurised
pressure tank, as here.
||Depends. Suppose the pressure tanks have 1% the volume of
the main tanks, then each "gulp" will be 100th of the total
burn, or on the order of 10 seconds, with a "coasting" phase
of maybe 1 second while the system recycles.
||//Would be lucky not to fall apart under the shock loads
no ? //
||I'm not sure how you mean. It's basically a regular rocket
engine which is fired for ten seconds, off for one, on for
ten.. so anything that can take a regular rocket thrust
would be fine.
||//use some jet engines and feed oxidiser in at high
altitudes// Jet engines are way less efficient than rocket
engines, and no good for going into spacington space,
even if you feed them oxidiser.
||//gas-pressurised// where does the gas come from?
||[+] like it.
Might use external pressure source for liftoff and thereafter use inertia of fuel mass for hydraulic ram effect through check valves to charge pressure vessels during each gulp, saving more weight.
||// where does the gas come from?//
||There are choices. Some rockets (mainly last-stages) do
use gas-pressure-driven fuels, in the usual way, with
pressure-tolerant fuel tanks. They normally use helium; a
small tank of liquid helium can provide all the gas
||Another option would be to cunningly use the boiloff
pressure of the fuel/oxidant itself. This would need some
extra plumbing, but should be doable. In the main tanks,
you'd bleed most of the boiloff as usual to keep the
pressure low; in the pressure tanks, you'd let the boiloff
accumulate to much higher pressures to drive the fuel into
the combustion chamber. You'd probably want to heat the
pressure tanks by having them close to the combustion
chamber, to provide rapid and high boiloff pressure.
||// use inertia of fuel mass for hydraulic ram effect through
check valves to charge pressure vessels//
||The problem is that the pressure tanks need to be filled
whilst the rocket is "coasting" between bursts, so there'll be
no inertial force of the type you need. You could have two
complete systems in parallel, I guess, so that one gulped
while the other one was firing; in that case, your system
||We like this idea. It may well be unworkable, but the failures during the prototype testing stage will be spectacular enough to more than justify the attempt.
||So the idea is that you still need the same work done (pressure X Volume) to deliver the fuel to the combustion chamber, but the fact that you have a much smaller high-pressure fuel vessel results in weight savings.
||Nice. Can you feed a small ammount of fuel to the oxidiser pressure tank, and a small ammount of oxidiser to the fuel pressure tank - in order to create the high pressure? If metered correctly, this might be very simple. Would work even better with hypergolic fuels.
||As for cross-feeding the fuel and oxidiser, I'd be wary of
that - flashbacks and all that. On the other hand, it's an
interesting suggestion: use very-non-stoichiometric
combustion in the pressure tanks to pump the liquids into
the combustion chamber. Nice!
||What I'm hoping to come up with would be a system
which, once started, would be entirely self-driving (ie, no
need for externally-controlled valves etc), and therefore
as simple and light as possible.
||Hypergolic fuels would, as you point out, be a nice way to
go, especially as there'd be no need to re-ignite the
combustion chamber on each cycle. However, as far as I
know, the hypergolics are all pretty evil chemicals. I
presume reignition can come from a spark- or glow-plug
type system? Or perhaps there's a catalyst which will
ignite kerosene/LOX mixes (just as platinum will ignite
hydrogen/LOX, I think)?
||Conventional designs do not to my knowledge require fuel tanks capable to combustion chamber pressures. The fuel pumps draw large volumes and low pressure in and eject them at high pressure into the chamber.
||why not use an automatic machine gun type approach and store pressure from each firing sequence in a storage tank to pressurize the feed tanks after they receive their fuel charge.
||if you used some sort of spring loaded piston arrangement you could in essence create an internal combustion rocket motor where the exhaust is what provides the motive force.
||//Conventional designs do not to my knowledge require
fuel tanks capable to combustion chamber pressures. The
fuel pumps draw large volumes and low pressure in and
eject them at high pressure into the chamber//
||I discussed, in the idea (above, left) the fact that most
rockets use turbopumps to pump fuel into the combustion
chamber. Some other rockets (see above, left) use high-
pressure fuel tanks instead. Both solutions involve
complexity, cost and/or weight.
||//if you used some sort of spring loaded piston
arrangement you could in essence create an internal
combustion rocket motor where the exhaust is what
provides the motive force.//
||I also discussed the use of "pulse jet" (or pulse rocket)
techniques, and explained why they're not very efficient
(see above, left).
||// store pressure from each firing sequence in a storage
tank to pressurize the feed tanks after they receive their
||I've been trying to think of a way to implement that, but
without success. Basically, you have to pressurize the
pressure tanks to a pressure higher than that of the
combustion chamber. This could be done with some sort
of hydraulic jack type of system, but then you're talking
about pistons which will be in contact with liquid oxygen
or with fuel, and with combustion gases at some ridiculous
temperature. I think that's a tall order.
||However, using liquid helium to provide the pressure isn't
so bad - a small amount of liquid helium will be enough to
pump (ie pressurize) many times its own volume.
||Also, I wonder if the heat of the combustion chamber
could be used to enhance boiloff in the pressure tanks,
and avoid the need for external pressurization?
||I'm moving towards a system which is totally self-driving.
Low-pressure boiloff pushes liquids from the main tanks
into the pressure
tanks, which are closed when full by a non-return valve.
Heating of the pressure tanks causes boiloff and raises
their pressure sufficiently to open the escape valve letting
liquid into the combustion chamber. When the pressure
tanks are empty, the whole cycle begins again.
||So, I think this can be done with an entirely passive
system and no need for additional pressurizing gas.
||//fuel gets cycled around the outside of the combustion
||Yes, that would be the sort of thing. I'm imagining some
sort of glorified steam engine, in which the heat of
combustion drives the pumping of the fuel. However, I
want to avoid all pistons (and anything bar the simplest
valve), since these will be tricky under harsh conditions.
||I could do this if I were sober.
||You said that about the genetically modified hedgehogs, and look how that ended up ...
||Hey, no, I said genetically modified HEDGES. Just because
they turned out to be able to walk doesn't mean they're
||We did warn you. "Triffid" and "Privet" are NOT interchangeable. Admittedly, you improved the security of your properties, but terms like "collateral damage" and "acceptable casualties" are not - as we understand it - normal in the context of herbacious borders in your society.
||Correct us if we're wrong (Like that could ever happen).
||That is a pleonasm. By definition, "casualties" are "casual",
and therefore acceptable.
||We defer to your superior legal representation.
||So in essence, it's a staged-pressure V1.
||Not advisable for human spaceflight due to the g-loading every 10 seconds or so, but interesting.
||How much time in between pulses would it take to refill the pressure tanks under a low-pressure fill system? And what is the rocket doing during this time period? Coasting off course? Falling?
||How about a pair of gulper-sets? One fires while ullage reloads the other; then they switch. Feed alternates between the two, and combustion is continuous.
No relight, no back-pressure recovery time, no fry/freeze temp oscillations of your combustion chamber, no whiplash for the astronauts, no (ok, less) pogo-stress for the payload...
||//So in essence, it's a staged-pressure V1. //
||//Not advisable for human spaceflight due to the g-loading
every 10 seconds or so//
||Probably true. However, the peak G will be the same as in
a continuous rocket, so it should be survivable, but
||//How much time in between pulses//...//what is the
rocket doing during this time//
||I'm guessing the refil could be done in 1/10th the burn
time, so 10 seconds on, 1 off... since the filling pipes can
be as wide as reasonable. During the 1 second coast, the
rocket is coasting and losing speed, but not much. Toss a
rock in the air and it's coasting from the moment it leaves
your hand. In fact, staged rockets such as the old Apollo
system often have "coasts" of a second or longer between
stages - it's not a big problem.
||Why not just use a graviton polarity generator powered from a forced quantum singularity ?
||Because it doesn't really exist, actually.
||I've bought one on eBay - never arrived. Once bitten.
||Fair enough. We blame FedEx, they lose stuff all the time. We are still waiting for our Ronco "Record Vacuum" ... we did the Western Union cash transfer thing just like they said, how long does shipment take from Nigeria to Low Earth Orbit ?
||So you had no luck getting a vacuum sent to space? Oddly
enough, my brother in Newcastle is still waiting for his
shipment of coals.
||I knew we'd agree on something.
||In this case, it's so easy ...
||Wow! Thanks, [bigsleep]. An animation is clearly worth
1,093 words. That's exactly the sort of thing I had in mind.
||//once the hot stuff gets out of the combustion chamber
there's a lot of heat energy there going to waste//
||Ideally not. The ideal rocket nozzle allows the
combustion gases to expand until their pressure as they
leave the end of the nozzle just matches that of the
outside atmosphere (which gets thinner, of course, as you
go up; hence no one nozzle is ideal at all altitudes) . This
expansion equates to cooling, so there's less "wasted"
heat. However, the energy needed to pump the fuel into
the combustion chamber is a tiny percentage of the
energy output of combustion, so siphoning off a tiny
amount of heat to pump the fuel will reduce the net
||Adjustable nozzle, shirley?
||What if the fuel reservoir were stored entirely inside the combustion chamber. The reservoir would not be especially robust or pressure worthy. The combustion chamber is. The chamber is gradually pressurized perhaps by incrementally adding fuel, or maybe cold and at the same time the reservoir is gradually pressurized.
|| The reservoir experiences the same pressure on the inside and outside so does not need to be robust as there is not a pressure difference across the wall. Pumps enclosed within the reservoir operate at low pressure to emit fuel and sustain combustion / pressure outside in the combustion chamber.
||A problem is that this would be a huge combustion chamber to enclose the reservoirs. I wonder if the reservoirs were surrounded with some fluid that would transmit the pressure up from the combustion chamber if that would work.
||//What if the fuel reservoir were stored entirely
inside the combustion chamber.// Interesting idea,
but in the end I don't think you gain. Ultimately,
you still have to enclose the same volume (ie, the
combustion volume plus the reservoir volume)
against high pressures, whether the latter is inside
the former or not.
||That won't help, because the pressure created
when they're mixed (and again when they react
with the fuel) will still propagate back up into the
||As far as I can see, if you have continuous flow
from the tanks into the combustion chamber, and
if that flow is pressure driven (rather than pump
driven), then the pressure in the tanks has to be
higher than that in the combustion chamber. But
an intermittent flow uncouples the tanks from the
||I guess that a "gulper rocket" approximates to a
machine gun, where a bullet is fed in under low
pressure, and then fired, creating high pressure,
||I like this idea. If it is difficult to inject fuel into a high
pressure vessel from a lower pressure storage chamber,
just wait till the pressure drops. Eminently sensible. [+]