TOWARDS VALIDATED CATALYTIC REACTOR MODELS

Licentiatavhandling, 2020

The use of liquid fuels in the internal combustion engine inherently produces several toxic emissions that need to be removed. This is done through a series of catalytic converters, referred to as the exhaust aftertreatment system (EATS), each catalyst with its own purpose. To cope with increasingly stringent emission legislation for the automotive industry, the performance of these catalysts needs to improve. For their development, modelling is an indispensable tool.

This thesis presents a mathematical model, a so-called single-channel 1+1D reactor model, that describes the reactions that occur inside a diesel oxidation catalyst (DOC) - along with the heat and mass transport. Moreover, the purpose was to improve an already existing model using relevant experiments such as kinetic experiments in a synthetic catalyst activity test (SCAT) bench, gravimetric analysis (GA), temperature programmed desorption (TPD) as well as scanning electron microscopy (SEM). These experiments enabled better inputs for the modelling framework. Furthermore, emphasis has been put towards investigating boundary conditions for the experimental setup. The radial mixing in a SCAT bench was investigated using pair of cleverly designed DOCs. The investigation showed that there was a concentration maldistribution across the catalyst inlet and the problem was solved using a 3D printed -alumina mixer.

The original 1+1D model relies on some simplifications which are rarely fulfilled. The model assumes that the catalytic washcoat forms a uniform slab - an assumption which may lead to incorrectly estimated light-off temperatures. This limitation of the 1+1D model was circumvented through the use of a sectionalizing principle, where the washcoat was divided into multiple segments which were simulated independently. Different experiments allowed for estimation of local properties - such as external mass transfer coefficient, washcoat porosity and thickness. The new model showed increased NO conversion at elevated temperatures compared to the original model.

By improving these single-channel models, EATS modelling in general can become more predictive - leading the way to emission-free transportation.