The competition to deliver ultra low emitting vehicles at a reasonable cost is driving the automotive industry to invest significant manpower and test lab resources in the design optimization of increasingly complex exhaust aftertreatment systems. Optimization can no longer be based on traditional approaches, which are intensive in hardware use and lab testing. This paper discusses the extents and limitations of applicability of state-of-the-art mathematical models of catalytic converter performance. In-house software from the authors’ lab, already in use during the last decade in design optimization studies, updated with recent, important model improvements, is employed as a reference in this discussion. Emphasis is on the engineering methodology of the computational tools and their application, which covers quality assurance of input data, advanced parameter estimation procedures, and a suggested performance measure that drives the parameter estimation code to optimum results and also allows a less subjective assessment of model prediction accuracy. Extensive comparisons between measured and computed instantaneous emissions over full cycles are presented, aiming to give a good picture of the capabilities of state of the art engineering models of automotive catalytic converter systems.

This paper involves an experimental investigation of the role of the volatile organic fraction ( VOF ) adsorbed on the diesel particulate, in the initiation of regeneration of a SiC diesel lter installed on a modern diesel engine, run on catalytic additive-doped fuel. VOF adsorption-desorption and oxidation behaviour is mainly determined by performing a thermogravimetric analysis ( TGA) of samples collected directly from a SiC lter installed on the engine running under low-and mediumspeed and low-and medium-load conditions, as more representative of city driving. Based on the TGA analysis results, the percentage VOF content in soot was calculated and mapped as a function of engine speed and load in the range of investigation. The e ect of adsorbed hydrocarbons on the regeneration behaviour was assessed by comparing regeneration experiments with the stepwise load increase for a lter loaded with soot at di erent VOF concentration levels. The appearance of a number of incidents of stochastic regeneration behaviour during loading at low exhaust temperatures with a relative high VOF content was observed and discussed. An e ort was made to correlate regeneration rate with the VOF content in soot and the prevailing engine operation point during loading. This work aims at better understanding of diesel lter behaviour with modern diesel engines and also aims to support improved modelling of fuel-additive assisted regeneration by use of fuel additives at low temperatures (150-400°C ).

Tight requirements posed by the increasingly stringent legislation complicate the design procedure for exhaust aftertreatment devices and systems. Since design optimization relies heavily on experiments and tests, emissions test data acquisition should comply with strict quality standards. Time-varying exhaust emission measurements incorporate a wealth of information stemming from the engine type, its fuel injection and ignition management and valve timing and the exhaust gas treatment devices present. The objective of this paper is to present the preliminary development process of a test data quality assurance methodology that may be coded in the form of computer software. This paper is a first attempt in this direction and is based on 15 years of experimental and computational experience of the authors' laboratory in exhaust gas treatment testing and modelling. The methodology comprises three steps. The first step involves modal analysis of the driving cycles. This allows comparison between different types of engine and test procedure regarding engine management philosophy and exhaust aftertreatment characteristics. The second step involves a systematic data synchronization and preprocessing procedure, which significantly improves data quality. The third and most important part of the methodology involves systematic checking by means of molecular and elemental balance calculations.

The need to deliver ultra-low emitting vehicles at a reasonable cost is driving the automotive industry to invest significant manpower in computer aided design and optimization of exhaust treatment systems. The significant fluctuations in the stock exchange market values of the precious metals employed in the manufacture of automotive catalytic converters has increased interest in the possibility of assisting precious metal loading optimization by means of mathematical modelling. Currently employed models of real world performance of catalytic converters cannot predict this effect. Recent improvements in the core chemical reaction modelling of the CATRAN code, reported in this paper, allow a good correlation to be made of precious metal loading with apparent kinetics, at least in the case of Pt-Rh catalysts. This may open new frontiers to the use of mathematical modelling in automotive exhaust after-treatment system optimization.

O p t i m i z a t i o n o f A u t o m o t i v e E x h a u s t T r e a t m e n t S y s t e m sTHESIS submitted in partial fulfillment of the requirements for the degree of D o c t o r o f P h i l o s o p h y i n M e c h a n i c a l E n g i n e e r i n g of the D e p a r t m e n t o f M e c h a n i c a l a n d I n d u s t r i a l E n g i n e e r i n g -3 - Since the core of almost every automotive exhaust aftertreatment system is some kind of catalytic converter, intensive research activity is devoted to the various scientific and technical aspects of catalytic technology. These aspects vary from practical research on new materials and novel techniques to improve catalyst microstructure, to theoretical contributions of detailed descriptions of chemical phenomena inside the converters and proposals of mechanistic reaction schemes.Nevertheless, the extremely high efficiencies that are expected from any modern catalyst cannot be attained without simultaneously optimizing aftertreatment device performance, engine design and the control system that drives the whole system. Thus, from a mechanical engineer's point of view, the exhaust treatment complexity is expressed as a strong interaction between these three fields: powertrain operation, engine management characteristics and exhaust treatment device operation.In general, the composition of exhaust gas depends on the engine type, the operating point of the engine, the ambient temperature, the transmission system and miscellaneous characteristics of each system. Thus, the engine management holds a key role in the behavior of the catalytic converter, since it affects both its input as well as the temperature range under which it operates. Moreover, exhaust line design, involving exhaust manifold heat capacity, insulated pipes and positioning of the devices, is strongly connected with the catalytic converter operation. Furthermore, catalysts with the same chemical composition may behave quite differently, depending on the actual procedures that were employed during catalyst preparation.The complexity of the automotive exhaust treatment systems quickly became apparent in the industry. Mathematical modeling tools are continuously developed, in order to complement the experimental efforts and provide guidance for areas that needed improvement. In the last 30 years, many models have appeared in the literature [2] . Their approaches and their scope were diverse, but in general they were rather focused on the individual devices than on the whole aftertreatment system. -6 -Specifically for the catalytic converter modeling: most kinetic studies so far resulted in models that are valid for the specific combination of engine and catalyst status at that moment (usually fresh catalysts). Although valuable for a specific engineering problem, such modeling results could only partially be extended in different systems, and only with great caution. Indeed, successful extension of the results of a specific case to another one is rare in the literature [1] . As a characteristic example...