Sound and Vibration REU
Analysis of Gas Dynamics
Conducted
by:
Adam Rosenbaum
Colorado State
University
John Gillis
Northern Arizona
University
Project Overview:
We spent our eight weeks working at the Engines and Energy Conversions Lab (EECL) at Colorado State University. The focus of our work was on a new program called Virtual Engines (a product of Optimum Power Technologies). Virtual Engines allows the user to create an engine entirely within the computer and predicts its outputs.
Our goal was to model a number of engines in order to determine the validity of the program as well as provide some working data for the engines. Specifically related to sound and vibrations, we analyzed the sound spectrum outputs and the pressure waves created by the engine.
The three engines that we analyzed were:
A 594 cc Suzuki ZR600 snowmobile engine
A 1650 hp Dresser-Rand Clark K5X natural gas
engine
A 100 cc Kawasaki HD-III motorbike engine
Snowmobile:
The program also
incorporates animations to show how different gas properties change in time
with the cycle of the engine. For the
snowmobile, we looked at the pressure waves in the exhaust pipe. There are both left and right traveling
pressure waves.
One of the major concepts behind 2-stroke exhaust
tuning is to time the return of the pressure waves accordingly to provide
a plugging pulse to the cylinder ports, therefore trapping a larger mass of
air in the cylinder to optimize the power produced. The opening of the exhaust ports and the escape
of the high-pressure exhaust provide the pulses.
We were able
to optimize some of the dimensions of the exhaust to provide additional horsepower
though the desired range.
The
following graphs are a sequence of the animation. Shown are the cylinders,
ports, and exhaust system
The below graph represents the pressure waves
moving through the engine at about 80 degrees from bottom dead center
Red represents the right moving wave, blue represents
the left and purple represents the superposition wave (sum) on the graph
Dynamic
Wave Graph:
Summary:
We were able to analyze the dynamics of the pressure
waves for the snowmobile and tune the exhaust pipes to produce additional
horsepower. Instead of building the
actual exhaust system and varying the parameters we were able to optimize the
piping using the computer.
The data that we
came up with corresponded fairly well, in both the sensitivity and the overall
values, with the test data that the Clean Snowmobile Challenge team took over
the last year.
Dresser-Rand Clark KLX5:
Sound
Spectrum Analysis:
For the
Clark engine, exhaust tuning was not applicable to the project but we were able
to do some analysis of the sound produced by the engine. The following graph shows the sound levels
that would be registered by probes placed in the end of the intake and exhaust
pipes. The x-axis is the varying
frequencies and the y-axis is the dBA level (loudness).
Notice that the
exhaust is louder and there is an overlap of the sound output for the exhaust
and intake at around 250 Hz.
The red plot represents the measurements in
the exhaust and the black represent the measurements in the intake.
Sound Spectrum Graph:
Summary:
Again we were able to look at different parts of an engine without having to physically put sensors inside of the pipes. The model turned out to have a problem with the port placement, therefore our rougher model had to be used and gave us values that were significantly higher, on the order of 2200 hp instead of 1650 hp.
We also were unable to model the turbocharger due to a lack of data from the manufacturer and inability to measure it due to time constraints. We compensated by using an appropriately high inlet pressure and temperature (setting the atmospheric conditions to match that of the turbo output).
Dynamic
Pressure Wave Analysis:
We again analyzed the dynamics of the airflow out of the
cylinder and through the exhaust system. The following are graphs of the
pressure waves through the exhaust and the relative position of the port and
disc valve.
Dynamic Wave Graph:
Summary:
We were able to model the motorbike engine and get values
that corresponded very well with both the WAVE model and the published
values. A working model will hopefully
provide a basis for improving these engines to reduce the emissions that are
polluting the environment in many other countries.
Conclusion:
This REU provided us with an opportunity to work on a
real research project for the summer.
We learned not only about sound and vibrations but also a lot about engines
and computer simulations. This program
provided us with a chance to actually be involved in the research process and
learn how to conduct a project with the hopes of one day being able to conduct
our own research.
Acknowledgements:
We would like to thank the following people and organizations for without their help none of this would have been possible:
The National Science Foundation and Army Research
Office for funding this REU.
Colorado State University for providing the
facilities.
The EECL at CSU for making room for us to work.
All of the helpful people at the EECL.
Dr. Bryan Willson- our faculty advisor
Dr. Allan Kirkpatrick
Jessica Adair and Horizon Briggs:
-Providing
the WAVE data for the Clark and HD-III respectively and allowing us to work
with them on their research projects.
Walt Hull- for helping us out with the data
for the snowmobile.
Dr. Willson and Tim Bauer for use of the motorbike
and Manila pictures.
Optimum Technologies- for their easy to use
software and for the great tech support.