The first engine I built was a small 5 lbs force rocket designed to run off air and ethanol. The fuel is pressurized with gaseous nitrogen. It is called an igniter because it is meant to create a flame for a much larger engine.
I started work on the engine in the Fall of 2011 and finished construction in early April 2012. Air is a notoriously poor oxidizer, however, it is also very safe to work with. I chose Air in order to learn how to safely control the flow of gases and liquids at high pressure.
The engine was built with standard bar-stock aluminum and formed on a lathe/mill. It is designed to run at 300psi (~2.06 MPa) and generate 5 lbs of thrust. It has a throat diameter of .125 inches and an area ratio of .2815, which expands to ambient pressure It runs at an Oxidizer/Fuel ratio of 2/1 and burns ~2.2 lbs of propellant per minute. A jet of fuel and oxidizer impinge upon each other inside the combustion chamber, which mixes the propellants and they are then ignited by the spark plug/coil. Once started it is expected to run at 1051 degrees kelvin or about 1432 degrees Fahrenheit. This is above the melting point of aluminum so the engine can only be run for a short time without overheating.
Detailed Engine Specs
Detailed engine specifications can be found here
. Note the file is .xslx format and has been scanned for viruses prior to uploading. I don't believe there are any macro's in it either.
The engine is controlled remotely via an Arduino and Xbee, with signals sent over serial connection via a laptop from a safe operational distance. The engine has 4 electric components: 3 solenoid valves and 1 spark plug/coil, which are all controlled remotely via the Xbee/Arduino link over serial. The solenoid valves control the fuel pressure, fuel flow, and oxidizer flow.
Igniter Control Boards
The first tests were run in April and ran through early June.
A key parameter in a rocket engine is the mass flow rate for the fuel and oxidizer. Getting these values right is extremely important in order to ensure the engine runs as expected. Prior to using actual propellants I simulated these tests with nitrogen and water as substitutes for air / ethanol respectively.
I had calculated these flows using basic fluid dynamics calculations for compressible (eg. gas) and incompressible (eg. liquid) flows. In order to determine whether my calculations were correct I used a pressure transducer for the fuel/ox lines and an Arduino to spit out real time (estimated) mass flow rates. I then compared these estimates with before and after weights of the tanks to see how close my estimates were.
Generally speaking I found the deviation of liquid measurements between calculated and measured to be around 15% versus 30% for the gaseous measurements. The fact that the gaseous calculations differed more from the measured values is unsurprising as compressible flows are harder to calculate as they have more inputs. Some of the deviation I've attributed to the scale used, which did not have high precision. The rest of the deviation is likely due to bad assumptions in the calculations.
Here is an example graph produced by taking the measured pressure from the transducer and calculating liquid flow every 10 milliseconds.
In June I really started testing the engine in earnest. This phase of testing concluded towards the end of July.
The first engine tests produced some interesting noises and I was able to generate combustion although it was only for a split second. My initial theory was that the temperature was too low for the ethanol to form a gas so that it could ignite. This theory was based on the flash-point
of air/ethanol, which is 61.9 degrees Fahrenheit. However, that theory was busted after testing on a hot June weekend. After studying this for a bit I decided on 2 courses of action (1) increase the chamber diameter (2) add a glow plug.
Increasing the chamber diameter gives the propellant specimens more time to mix before being forced out through the nozzle. This helps aid in combustion and I noticed some, although not significant, improvement.
The next change I made was to add a glow plug, controlled via a relay, to the engine. Glow Plugs
are used to help aid ignition in diesel engines by heating up the chamber to around 800 degrees Celsius. here
is an image of the modified test stand to include the glow plug + corresponding relay circuit.
I was hopeful that using this to heat the chamber as well as the propellants coming into it would make it easier for the spark plug to ignite the mixture. Unfortunately, while it did help I was still unable to generate reliable combustion. Here
is a short video clip showing one of the more successful engine tests.
As a final attempt, I decided to try pulsing the engine by having the Arduino turn on/off the valves a few times per second. This provided a very interesting outcome whereby the engine would burn for a split second, stop burning, and then start again as the next cycle started. However, the combustion was unreliable and increasing the cycle time (eg. more cycles per second) eventually stopped any combustion. I didn't investigate this phenomenon too much but it's likely that the compressed air was preventing rather than aiding combustion much like a candle can be blown out with a quick gust of wind.
I eventually concluded that getting a reliable engine fire using air/ethanol would be too time-consuming to pursue. At this point I'd gained enough experience working with pressure, plumbing, valves, engines, that I felt comfortable moving onto a Gaseous Oxygen (GOX) / stronger oxidizer