Date of Graduation


Document Type


Degree Type



Statler College of Engineering and Mineral Resources


Mechanical and Aerospace Engineering

Committee Chair

Benjamin C. Shade.


A direct comparison has been made between a tapered element oscillating microbalance (TEOM) and the traditional gravimetric filter method designed per 40 CFR §1065.145. A repeatability test, consisting of multiple emission tests, was also performed for the TEOM system that shows that the TEOM was a consistent instrument. The TEOM was also tested with and without a diesel particular filter (DPF) during an 8-mode test on a 1997/1998 Kubota 2.2L engine to see if the concentration of the particular matter (PM) flowing through the TEOM has an effect on accuracy. The flow rate was varied to find the optimal setting using the Federal Test Procedure (FTP) run on a 1992 Detroit Diesel Series 60 engine. The comparison to the gravimetric experiment was conducted over 30 runs using the FTP for a 2004 Cummins ISX heavy-duty diesel engine. The TEOM system was found to consistently report a total particulate matter measurement of an average of 25.7% less than that of the gravimetric filter method. During the DPF experiment it was shown that the TEOM had a 3%, 7%, 13%, 18%, 4%, 14%, 15%, and 59% less percent difference for modes 1 through 8 respectively while the engine was without the DPF compared to the DPF equipped engine. When the flow rates were varied, it was found that 2.0Lpm was the optimum setup with an average percent error of 9.68% and a coefficient of variation of 1.43%. The problems that are present in the TEOM system include inaccurate real time data because the setup was susceptible to vibrations during heavy acceleration, an increased accuracy with an increased surface area of the filter due to a buildup of PM, and a deficiency in obtaining all of the particles in a sample for an accurate calculation of total PM. An examination of the increased accuracy because of the clogged filter, which decreases the flow rate and increases the filter surface area, is conducted to show an optimal value for either has yet to be achieved. It was found that as the flow across the filter decreased, the average percent error went from 27.9% while the flow across the filter was a constant 2.0Lpm to an average of 22.4% while the flow decreased from 2.0 to 1.5Lpm.