47147-AC9
Modeling and Optimization of Diesel Particulate Trap Ignition Phenomena
Background and Motivation
Diesel engines emit large amounts of particulate matter. Diesel particulate traps, with a monolithic reactor structure, have been proposed to collect these emissions. These traps are periodically regenerated to remove the collected particulates. This regeneration mimics the ignition phenomena found in the classical reaction engineering literature regarding packed bed reactors and catalytic converters. Just as in these examples, a proper understanding of particulate loading and reactor control is required in order to lead to a safe and efficient regeneration.
The goals of this project are to increase the fundamental understanding of the coupled transport mechanisms underlying diesel particulate filter regeneration and to utilize this knowledge to develop design strategies to significantly reduce environmental pollution from trucks and buses and subsequently improve human health.
Summary of Annual Progress and Impact of Research
It is noted that this research work builds upon that of Dr. Haishan Zheng who completed his PhD in Chemical Engineering from Michigan Technological University in 2004. Dr. Zheng developed a reproduction of the computer model of the diesel particulate trap developed by Bissett (Chemical Engineering Science, vol. 39, pp.1233-1244 (1983)). This model solves several partial differential equations to predict diesel particulate trap performance. Zheng and Keith (AIChE Journal, vol. 53, pp.1316-1324 (2007)) developed a simplified model using averaging theory which provided excellent agreement with the full model. However, it was tested under very limited conditions.
During the first year of this project, research has focused on two key areas:
- Parametric study of diesel particulate trap ignition
- Sensitivity analysis
During the past year, two different parametric studies were performed. The first study focused on the structure of the diesel particulate filter, and varied the length of the diesel particulate trap and the thickness of the walls within the diesel particulate trap. For a wide range of operating temperatures, the ignition time (the time it takes to regenerate the diesel particulate trap) and ignition length (the position within the diesel particulate trap at which the reaction was initiated) were computed using the reproduced model of Bissett and the averaged model of Zheng and Keith. Excellent agreement was found between these models. It was found that reducing the length of the diesel particulate trap could lead to incomplete regeneration, while increasing the length of the diesel particulate trap does not have a significant impact on regeneration. Changing the wall thickness has a large impact on the regeneration (in general, decreasing thickness reduces the ignition time and length).
The second parametric study varied the exhaust gas flow rate and the initial thickness of the carbon deposit layer on the particulate trap. For a wide range of operating temperatures, the ignition time and ignition length were computed using the reproduced model of Bissett and the averaged model of Zheng and Keith. In most cases, excellent agreement was found between these models. However, in computing ignition lengths in particular, it was found that at low initial deposit thickness the averaged model under predicts the value obtained with the full model. Thus, these conditions violate a key assumption in the averaged model. It appears that this is likely due to inability of a thin carbon deposit to “ignite,” or rapidly react. Increasing the gas flow rate led to ignition occurring far downstream of the diesel particulate trap entrance. This may mean that high exhaust gas flow rates can prevent ignition, especially at lower temperatures. This could be important under certain driving conditions.
Finally, a sensitivity analysis was performed using the averaged model to compute ignition times and ignition lengths. The parameters varied included engine exhaust temperature, gas flow rate, and initial deposit thickness. An important finding is that the slope of the ignition time vs. deposit thickness is sensitive to changing the deposit thickness but is relatively insensitive to changes in the gas flow rate. However, the slope of the ignition length vs. deposit thickness is sensitive to changes in both of these parameters.
All of these results have an impact on the design of diesel particulate traps for real world driving conditions. Furthermore, this work affects the development of control algorithms to manage total accumulated carbon within a diesel particulate trap.
Impact on Career of PI and Students
Funding from the American Chemical Society has allowed the PI to pursue this important project in emissions abatement and to support a graduate and undergraduate student that otherwise could not be supported. Both students have benefitted from the ability to study a research problem. Furthermore, the doctoral student has strengthened his writing and speaking skills.
