It is a simple rotary rocket, with an annular combustion chamber of square cross-section. Drilled holes closely spaced parallel to the surfaces of the combustion chamber provide fuel feeds. Regeneratively, the fuels, nominally LOX and propane, each cool one face of the combustion chamber. Concentric fuel feeds separate, each goes up one side of the engine, then through a small injector into the engine.
The engine is designed to rotate at 600 m/s, giving a propane pressure
of 16 ksi and a lox pressure of 26 ksi, more than enough to give 10 ksi
combustion chamber pressure.
The engine is 2 or 3 pieces assembled with thermal shrink fit and a small number of nozzles, also thermal shrink fit. The next prototype will have fluid dynamic bearings, fed with propane from a tap on the engine. lnteral ignition will be provided by internal spark.
In prototype operation, a starter engine will spin the engine up to some fraction of operational speed.
The result is an engine with combustion chamber pressure of 10 ksi (or
more) and a thrust/weight of over 400.
Unobvious aspects of this design are:
1) Corilois effects cause
the fluid flow in the combustion chamber to have a whirl, this will tend
to damp standing waves and combustion instabilities.
2) Injected cold fuels will be confined to the outer edge by the 10^6 g forces. Thus, they will mix and burn before they will travel to the nozzles, which are near the center.
3) Conventional designs depend on fuel velocity to diminish the boundary layers in the fuel feed tubes, analysis shows this is insufficient by a factor of 100 in this design. However, normal "free" convection forces exceed this effect by a factor of 1000, and are quite sufficient to cool the engine.
4) Pressure is cheap, so we will use as much as we can by spinning the engine to near its limits within some safety factor, simply by building it with the correct angle of nozzles.
5) The rotational velocity is exactly that which causes the exhaust to emerge with zero spin. This is the velocity that the nozzles rotate at, thus the angle of the nozzles is such that sin(angle) * exhaust velocity = rotational velocity of the nozzles. This result is simply explained due to conservation of angular momentum. Since some people find this difficult to see I will try to explain. Picture an engine rotating so that the exhaust comes out straight, with no spin. Now change the rotational velocity. The exhaust has a spin, and thus angular momentum. Where does this come from? Since the fuels enter with zero angular momentum this delta of angular momentum goes into the rotating engine. If the engine is rotating too slowly, it speeds up, if it rotates too fast it slows down. This is linear with the delta of rotational velocity, so on examination of the differential equation, it is found to be stable with no oscillation. The equation is of similar form to that of friction, which is also stable.
6) Less importantly, a non-rotating engine can be ignited by external ignition. The flame enters one nozzle and the pressure exhausts through the other. I've melted a few engines, so I know the combustion was internal. However a rotating engine will have external pressure waves due to the rotation, and this looks enough to keep ignition from entering properly.
7) Since the injectors are sized for 10ksi or more fuel pressures, in the non-rotating state there is very little thrust, less than 3 lbf in the prototype with 150 psi fuel feeds, this isn't (wasn't in the 1997 prototype) enough to spin the engine up due to friction mostly from the seals. Thus we need a starter motor, an external one in the prototype.
8) The engine isn't throttleable. The obvious tactic of fuel starving the engine looks very dubious, when one realizes that the fuel in the engine is enough for only a few milliseconds.
9) The engine can scale from 100 lbf to over 250,000 lbf with little effort. The first prototype was designed for 400 lbf, the next will probably under 1000 lbf. I want it small for logistic reasons. The first was as small as we could make and still have it simple to manufacture.
The result is we have an engine with combustion chamber pressure of 10ksi (or more) and thrust/weight of over 400.
It will be simple to manufacture, just a lathe and a drill press could be sufficient, though prototypes will be done with more precise drilling tools, to ensure uniformity, and give more validity to test results. No or minimal welding required!
So what ISP will this give? Probably over 350, perhaps approaching 400. Due to the fact that I don't have the right modeling tools for the 3400C that the right fuel rich mix gives at 10,000ksi, I'm not exactly sure.
I'm looking for technical criticism, engineering and financial assistance,
and a test crew. Most of the calculations are available on the website,
though not in a easily grasped format. Most of the calculations give
results for a particular design which way vary from time to time.
If you have Mathematica 4.1 you can download the notebook itself and try
different numbers. If you do this be sure to communicate with me,
I can use your results, and insights. Thanks for wading through this.
email@example.com (Roger Gregory)