Gulper rocket

[pauldear]

This is an idea which I just posted on the Halfbakery website. I suspect people here will be able to tell me if it's been tried.

The basic problem is how to deliver fuel and oxidiser to the combustion chamber of a liquid rocket. Because the combustion chamber is at high pressure (?100bar?), it takes a greater pressure than this to pump the fuels into the chamber. As I understand it, this is normally done either by high-pressure pumps (which are complex and heavy), or by pressurising the fuel and oxidiser tanks (which means they have be very strong and heavy, to take >100bar).

My idea is to do away with pumps, and to avoid the need for the entire fuel and oxidiser tanks to be highly pressurized. Instead, there's a small intermediate high-pressure reservoir (one for the fuel, one for oxidiser), which holds (say) 10 second's worth of each component. There are valves between the main tanks and the pressure tanks, and between the pressure tanks and the combustion chamber.

The gulper rocket fires in bursts of (say) 10 seconds. First, the pressure tanks are closed from the combustion chamber and opened to the large, low-pressure fuel and oxidiser tanks, to fill the high-pressure tanks. Then the h.p. tanks are closed to the main tanks, and opened to the combustion chamber, and driven to high pressure (eg by helium, as in a "standard" pressure-fed rocket). Combustion is ignited, and the rocket burns for 10 seconds. When the fuel in the pressure tank is exhausted, combustion stops, and the whole cycle repeats itself. Hopefully, the pressure tanks could be refilled in 1 second or so, giving a "10 seconds on/1 second coast" pattern.

Basically, it's a way of making a pressure-fed rocket where you only need to pressurize the small, intermediate pressure tanks instead of the whole main tanks.

Any thoughts?
Paul

[lavalamp]

An interesting idea, but I think difficulty lies in re-igniting the engines every 11 seconds, not to mention putting them through a lot of very high intensity heat cycles. You'd also lose a lot of velocity in the coast period when climbing vertically.

Perhaps in implementation there could be two high pressure tanks each for the fuel and oxidiser, one tank being filled and pressurised while the other is being emptied. If each tank was smaller it could take up the same volume in the rocket, and by sharing a common wall the mass needn't be that much higher than for a single tank. Best of all the rocket could burn continuously, providing a graceful switchover between tanks to feed the engines could be achieved.

The question is, would carrying around an extra tank (or two) for each feed line, plus a high pressure helium tank, and all the systems that go with it, weigh less than a couple of pumps? And even if it did weigh less, is it worth the extra complexity to pull it off?

I'm inclined to say no, but then a lot of people told Whittle no when he has this crazy idea for a turbo-jet engine.

[pauldear]

I'm not sure how much velocity you'd lose during the recycles - I'm assuming something like one second out of ten spent "coasting". At reasonable altitudes, where air resistance isn't too huge, you shouldn't lose much more than G (ie, 10m/s) during the coast; also, the total burn time will be the same as in a continuous rocket.

Re-ignition is a problem, but I'm hoping it's not too difficult (but then again, I know nothing about ignition systems for rockets). And I'm not sure if the cyclic stresses would be much more damaging than the continuous stresses (but again, I'm not sure).

Regarding complexity/weight: I agree, at least for larger rockets where pumps are a small part of the cost/weight budget. But I was thinking of smaller rockets, and I think it's possible to make this system almost "self running", in the sense that nothing more than some spring-biased one-way valves could control the cycling inherently. You still need the overpressure (eg, helium) to drive the system, but this needn't be a big volume (since a small amount of liquid helium expands to give a big volume of gas).

Alternatively, as someone on the HB suggested, you could bleed a little oxidiser into the fuel high-pressure tank and vice versa, and use a very-very-limited combustion inside those tanks to pump the liquid into the combustion chamber. Or you might be able to bleed some of the combustion pressure itself to help drive the system.

Ideally, I want to (and think I can) devise a system which involves only the combustion chamber, the pressure tanks, the lightweight main tanks and some very simple plumbing and one-way valves, and which would then be entirely self-running. The model I have in my mind is something like the little "putt-putt" engines for toy boats, which perform a reciprocating cycle in a very elegant and simple way.

[ZiraldoAerospace]

I was reading a book about the X Racer rocket engines, and it mentioned that it uses a slow piston driven pump to pump the low pressure fuel into an intermediate tank that is high pressure, then it is ignited. The reason for this is the piston doesn't provide a steady stream, but it can keep up with the pressure needs if you find a way to "average" it in a way. That is what the intermediate tank is for. This way, the rocket doesn't have to shut off at all, yet doesn't need thick tanks or a turbopump.

[pauldear]

That's a nice idea. However, that's basically a sort of equivalent of the turbopump, but with a pressure reservoir to even out the flow from the piston. Also, I wonder how easy it is to make a piston system that will handle liquid oxygen...seals would be an issue.

I think the "gulper" can be made to work with one pressure supply of helium, one "active" valve (ie, that is actively controlled), and four fairly simple one-way valves. In other words, it should be mechanically simple.

[Sage]

Similar concepts have been around since the 60s (at least) (so Paul, you have good company on this one). You don't need to turn off the engines and coast; the system can be made in such a way that there is a (relatively) constant flow to the engines; you can do this via buffering the system so you are filling one small reservoir while the other is emptying. Steve Harrington (Flometrics) has been working on implementing such designs for years; there are several drawbacks, and it really only makes sense for boosters (and even then it's questionable). I can go into more detail if anyone is interested.

Sage
http://www.littlemonsterrocket.com


EDIT:
I see that lavalamp has anticipated the general buffering solution (from his post above):
"Perhaps in implementation there could be two high pressure tanks each for the fuel and oxidiser, one tank being filled and pressurised while the other is being emptied. If each tank was smaller it could take up the same volume in the rocket, and by sharing a common wall the mass needn't be that much higher than for a single tank. Best of all the rocket could burn continuously, providing a graceful switchover between tanks to feed the engines could be achieved."

Yes, but there are issues. As I mentioned earlier, I can go into more in detail if anyone is interested...

[pauldear]

Ah well. Glad to know I'm treading a known path!

I have the benefit of knowing almost nothing serious about rocket design, which is a great advantage in liberating the thought processes. I've got this fantasy that I could build a spacefaring rocket out of seventeen coke cans and a couple of hoover parts - so it's just as well I'm not one of the entrants!

[lavalamp]

[cpooley]

The idea has been anticipated by some years by a San Diego CA company called Flowmetrics.

See: http://www-rohan.sdsu.edu/~sharring/Pis ... or_CEV.pdf

Steve Harrington has called it the "pistonless piston pump" and he has been playing around with it for years.

It has the complexity of a turbopump and the pressurized gas requirement of a pressurized system.

It's fundamentally silly, based on fallacies.

The mass of a turbopump can be very small if its structure is integrated as shown on the last slide of
http://www.microlaunchers.com/7816/L3/sa05/sa05.html .

In 1958 a German company build what may have been the first closed cycle engine called the P-111. It used a turbopump built into the head of the engine, LOX cooling, other things not done even yet.

The Russians followed a year or two later with the NK-15, NK-33 for their moon program rocket, the N-1. That method is the standard Russian to today. It works well.

My calculations for the design on my site could have a thrust/weight ratio over 200 (for NK-33 it was 135). This will be due to a higher degree
of mechanical integration. There will be only one seal in it: to separate the fuel, LOX at the inputs to the pumps.

A general plan for developing, testing has been worked out, and it will be undertaken after the development of the N Prize and (now2) the new NASA Nanosatellite Challenge.

Or, if someone were to come forward with interest in it sooner, it might be done sooner and be used in the "ML-1" entry level launcher.

A closed cycle turbopump design is definitely the lightest of any of the systems proposed.

Charles Pooley Microalunchers, team #5

[cpooley]

Paul: see http://en.wikipedia.org/wiki/Pistonless_pump , and Flowmetrics. Old idea and I think it will not work well.

ML-1 plan is for a pressurized system, with chamber pressure about 10-15 atmosphere for 1st stage and about 5 for stage 2 and 2-3 for stage 3.

If a turbopump is to be used it would be for stage 1 only and the chamber pressure would be about 20 atm.