Ejector Jet Engine

Ejector Jet, Testing and principal sketch

Bell Rocket Belt
Hydrogen peroxide rockets are easy to control, light weighted and still powerful. This made them the natural choice in applications like the Bell Rocket Belt. They are not ideal for this application though, because they consume a lot of hydrogen peroxide. The time of operation for a Rocket Belt is only about 30 seconds. The weight of the filled hydrogen peroxide storage tanks will be too heavy at take-off, if trying to make the time of operation longer.

Bell Jet Belt
The short time of operation was probably the most important reason why the Rocket Belt was never introduced in bigger scale by the US military. Bell tried to improve the system and started to develop a Jet Belt, powered by hydrogen peroxide. A simple jet engine was replacing the rocket engine. Jet engines consume less fuel, because the surrounding air is "pumped" with the help of the motive force of a rocket nozzle or a turbine and ads thrust. The Jet Belt project was never successfully completed. I believe partly because the system became too complicated and heavy.

Peroxide Propulsion Ejector Jet
When reading about the Bell Jet Belt, I could not help start thinking about how to make an even simpler jet engine. The idea developing in my mind I called the Ejector Jet. You find a principal sketch of it at the upper part of this page. As you can see, the engine is extremely simple. There are no moving parts. It is a normal hydrogen peroxide rocket, surrounded by a light air ejector duct. The surrounding air is sucked in to the ejector pipe by the motive gas flow from the rocket nozzle. According my calculations the power can be close to doubled, compared to a "naked" rocket engine.

Ejector Jet Thrust Calculations
To learn more about ejectors I studied the literature and found a few usable references. The one I have used the most is:

A Simple Air Ejector, J.H. Keenan and E.P. Neumann, Cambridge, Mass., Journal of Applied Mechanics, June 1942, page A75 to A81

Based on the references found I prepared an Excel calculation program, which you find here.

The program is not very user friendly at this stage. It was made just for my own convenience and I did not think of publishing it at the time. I will try to clean it up when I have some time over. It is an iterative calculation procedure. The pressure at the inlet is adjusted until the momentum at the inlet equals the momentum at the outlet.

Calculation Results
As an example, the calculations show that the thrust can be almost doubled when the area of the ejector pipe is 600 times the area of the rocket nozzle throat.

Some recommendations from literature are that the length of the ejector pipe should be 6 times the diameter. The nozzle should be positioned 0.5 D in front of the ejector pipe.

For a person with some technical skills it sounds peculiar at first glance that the thrust can be increased in this way, because energy can not be created from nothing. The reason to the increase is that the mass flow of gas is increasing, but at the same time the gas velocity is decreasing compared to the naked rocket nozzle. One can simply say that the engine is geared down. This is no problem for slow moving vehicles like Jet Belts.

Ejector Jet Prototype
After having convinced myself that the ejector principle is working perfectly well in theory, I was eager to find out if I could make it work in practice.

I made a simple prototype from standard light weighted pipes, normally used for ventilation ducts in buildings. I have not found the English name for this type of pipes, but in Swedish they are called SPIRO pipes. A hydrogen peroxide rocket was connected to the ejector pipe, as can be seen on the sketch on the upper part of this page. There you can also see a photo from the testing of the prototype.

D, rocket nozzle throat = 6 mm
D, ejector = 150 mm
L, ejector = 1000 mm
The rocket is mounted inside a bigger pipe in front of the ejector pipe. D = 250 mm

At the expected test conditions, the prototype ejector was calculated to increase the total mass flow (air flow + motive flow) with a factor of almost 18 times the motive flow from the rocket. At the same time the flow velocity is reduced from almost 1400 m/s at the rocket nozzle exit, down to 140 m/s at the ejector exit.

Prototype Testing
I have reported from this test in the April/May Development Report. In this report I did not mention any details. The reason was that we had some thoughts of filing a patent application for this engine. Later on I have understood that the Ejector Jet idea is not new. One can read about it in the literature. It is normally called "Augmentor". As an example you can find an augmentor test made by NASA here.

Test Results
So far I have done only one test run. The result was poor. The first problem was that I could not run with the throttling valve fully open, because than part of the hydrogen peroxide was not decomposed. The exhaust gas became "wet" and visible. This is what you can see on the photo. The reason was that the catalyst had lost too much activity. The rocket was not running clean until the pressure in the feed line before the rocket was throttled down to below 20 bar. The measured thrust of the ejector jet was than 7 kg. This is just slightly more than the thrust that a naked rocket without an ejector would have. The expected calculated thrust value for the ejector at the test conditions was about 12 kg.

The noise level was 100 dB, measured 15 meters from the engine. This is 5 to 10 dB less than for a naked rocket. This does not seem like a big difference, but you actually experience the difference to be quite big. This is because the dB scale is logarithmic. The engine was insulated at the test.

The temperature on the exhaust gases at the exit of the ejector pipe was typically 60 oC, This is much higher than the calculated value of about 25 oC. The reason to the difference is probably that the ejector has sucked much less air than expected.

Discussion and Conclusions
According to the calculations the ejector was expected to increase the thrust from the rocket substantially. This did not happen at this first test. I would do some corrective adjustments and make more tests if I could just figure out WHAT to do. If anyone of the readers can come up with something that can lead forward I would be most grateful!

Even if the thrust did not increase much, the ejector is still an interesting possibility to lower the noise of a rocket and to cool down the exhaust gases. This is of special interest if hot rocket systems are used, when a pilot is running a risk to get burned. The ejector also helps to cool down the rocket, which may be necessary for hot systems.

This article was updated on November 30th, 2006