Test Report 3, May 14, 2000
I'm writing this almost a month after the test. This has given me time to review the problems, causes, and develop corrective actions.
During testing, several problems arose, some were corrected, others were not. Sustained combustion was accomplished, though at much less efficiency than desired. Buzz was encountered during combustion which varied in intensity with fuel flow but not in frequency. Analysis of the testing resulted in several recommendations.
I moved the test setup to the corner of a shed that channeled the wind and increased its speed. On the test day the wind was clipping along at a mild (for desert springtime) 40 mph. This was increased to perhaps 50 as it turned the corner of the shed. The shop-vac added maybe another 20-30 mph.
The goal of this test was:
These pictures show the principle parts and assembly of the Near Flight Weight Engine Number 1 (NFWE-1). It weighs just under 2 lbm and I hope to get about 8 lbf thrust at 120 mph airspeed.
Six parts now make up this engine. The flame holder is flanked by the aft-body and the fore-body. In front are the inlet cone (subsonic) and the two fuel lines.
The fuel lines entered the engine aft of the flame holder at a right angle to the aft-body. They then made an s-curve (to avoid pinching) and ran up the inner edge of the flame holder. At the fuel injection end the fuel lines were formed into annular passages. This would allow an accelerator rod to be inserted through the middle without interference.
This method of routing and mounting proved to be very problematic.
The first problem that occurred was an inability to light the pilot. Upon disassembly I found that in the second iteration of forming the pilot fuel line I forgot to drill the holes so no fuel was entering the pilot can. This was corrected.
The second problem was poor control of the pilot fuel flow due to use of a valve intended for water. Following the test I found a potentially better valve sold for use in distributing air to fish tanks. I have yet to try this one.
The third problem was buzz, an interaction between the acoustical resonance of the engine and the variability in combustion characteristics with pressure. While the sound volume varied with fuel flow I did not consider it a major problem. In fact, some of the literature I've read indicated that, at low speeds, this may even increase the performance of an engine.
The fourth problem was gross instability in the combustion. This was due to the long fuel lines which arced up into the flame and caused boiling of the fuel in the lines. This resulted in pulsing of the flame. The correction for this is re-routing of the fuel lines.
The fifth problem was poor distribution of the flame. In the past I had been increasing the swirl to improve mixing the flame and fuel. This time I had too much swirl causing the flame to concentrate in the center of the airflow and not mix well with the fuel. Again I disassembled the engine and made some modifications which improved the situation but didn't correct it to my satisfaction. The true correction for this is to change the pilot can design to be more like that used in TR-2.
Below is an image of the improved flame. While this is much better I really want the entire cavity, wall to wall, filled with flame.
Other than that, the engine did operate, though thrust was not measured. The annular configuration for the afterburner fuel line did not work very well. The flame could not be coaxed out along the inner walls of the afterburner. I could get it to burn but it was farther aft than desired.
Note to readers. The onset of spring/summer is causing me to work on other activities that are seasonally sensitive. This means that further testing will be delayed for a while longer.