Modeling, reaction schemes and kinetic parameter estimation in automotive catalytic converters and diesel particulate filters

Abstract

This thesis is concerned with the modeling of catalytic converters (CC) and diesel particulate ?lters (DPF) employed in automotive exhaust lines for the reduction of gasoline and diesel engines emissions. Two engineering models are developed in this work, for the CC and DPF re- spectively. The models incorporate phenomenological reaction schemes and semi- empirical rate expressions. The objective of the models is to incorporate all necessary degrees of freedom for the accurate prediction of the CC and DPF operation and, at the same time, to eliminate super°uous complexity. The most important part of such models is the formulation of an appropriate reaction scheme that includes the primary chemical phenomena occuring in the wash- coat of the catalytic converters and the soot deposit of the DPF. A new reaction scheme is developed for each aftertreatment device to align with this fundamen- tal modeling requirement. These reaction schemes and the corresponding balance equations that are developed form the modeling cores for the respective devices. Analytical solutions of the balance equations are obtained to enhance modeling ac- curacy and computational e±ciency. Additionally, interfaces to two commercial FEM software suites are developed for the DPF model, in order to enable the 3D computation of temperature and stress ?elds. This allows the detailed study of the DPF operation as regards partial regenerations and cracks formation which endanger the integrity of the ?lter. A computer-aided optimization procedure is also developed for the estimation of the kinetic parameters introduced by the rate expressions of the models. The optimization procedure is based on a genetic algorithm and a properly de?ned per- formance measure. The validity of the models and the kinetic parameter estimation procedure is checked by applying them to real world case studies. It is demonstrated that the models are capable of predicting the operating behaviour of the CC and the DPF un- der realistic conditions and are thus well suited for use within the frame of computer- aided engineering design and optimization of automotive exhaust lines. Finally, the modeling levels in the ?eld of automotive powertrain applications is discussed and it is attempted to identify the requirements of the industry. It is claimed that the current modeling paradigm of the industry may not be directly applied to the exhaust aftertreatment devices modeling while, in the same time, mod- eling e®orts often do not recognize industrial needs. It is argued that the emergence of a new modeling paradigm, accompanied by models aligned with the industry's re- quirements, will eventually transform the current models from computational cores into robust engineering tools. This evolution is a long-term goal of the research ac- tivities of the Laboratory of Thermodynamics and Thermal Engines, among which this work is placed.