What is it about?
Pulsejets are among the simplest jet engines ever built. Instead of the heavy compressors and turbines of a conventional jet, they rely on intermittent combustion driven by sound waves. Fuel and air burn in a rapid burst, and the hot gas rushes out of a long tailpipe. Pressure waves then bounce back and forth inside the engine, sucking in a fresh charge of fuel and air through small flapping reed valves, and the cycle repeats hundreds of times per second. The famous World War II "buzz bomb" was powered this way, and hobbyists still use pulsejets for model aircraft and boats. Their simplicity and low cost also make them attractive for powering high-speed drones. In this paper we took a standard 50 cm hobby pulsejet and studied it in unusual detail, using both laboratory measurements and computer simulations. We measured pressure, temperature and thrust along the engine. We used a laser and a high-speed camera to watch the reed valves open and close, a light sensor to detect exactly when the fuel burns, and laser velocimetry to measure how fast the exhaust moves. We then built a three-dimensional computer model of the engine that reproduces these measurements, letting us see the waves, flames and flows inside the engine throughout its cycle.
Featured Image
Photo by Hermeus on Unsplash
Why is it important?
Although pulsejets have been around for a century, they were largely abandoned in the 1950s, and much of how they actually work was never studied with modern tools. Our engine produced an average thrust of 25 N while running at about 240 times per second. Our simulation predicted this frequency to within a few percent and matched the measured peak pressure. We found that the valves stay open for only about 30% of each cycle and that combustion takes place mainly while the chamber pressure is above atmospheric. We also measured, for the first time to our knowledge, that the exhaust flow briefly reverses and is sucked back into the tailpipe during each cycle. This backflow produces negative thrust and points to a clear target for improving performance. Textbooks traditionally describe the pulsejet as a simple organ pipe closed at one end, a so-called quarter-wave tube. Our results show that the real engine behaves more like a one-sixth-wave tube, because the opening valves and the changing shape of the engine alter how the sound waves travel. This updated picture, together with a validated computer model, gives engineers a practical tool to redesign and scale these engines. That is an essential step toward simple, inexpensive propulsion for small high-speed drones, where conventional turbine engines become too costly and inefficient.
Perspectives
This paper was the starting point for a long line of pulsejet research in our group. I remember how striking it was to see the reed valves captured in high-speed images and to find that combustion, flow and sound all had to stay perfectly in step for the engine to keep running. Working with my co-authors Tao Geng, Andrew Kiker, Robert Ordon, Tiefeng Zeng and Will Roberts was a pleasure. I am also grateful to DARPA for supporting this work. Some lessons came the hard way: our first attempts to run the engine on propane destroyed the reed valves, and switching to liquid ethanol, which cools the valves as it evaporates, solved the problem. Many people think of pulsejets as noisy relics from the past. I see them as a remarkable example of how much useful work can be extracted from simple physics with almost no moving parts. I hope this paper shows that combining careful experiments with modern simulation can make an old technology new again, and that it encourages others to consider pulsejets as a serious option for propelling the next generation of small, fast and affordable drones.
Andrey V Kuznetsov
North Carolina State University
Read the Original
This page is a summary of: Combined Numerical and Experimental Investigation of a Hobby-Scale Pulsejet, Journal of Propulsion and Power, January 2007, American Institute of Aeronautics and Astronautics (AIAA),
DOI: 10.2514/1.18593.
You can read the full text:
Contributors
The following have contributed to this page
