Modelling of non uniform washcoat in catalytic monolith reactors

Conference poster, 2019

For advances in the design of exhaust aftertreatment systems, modelling can be a valuable tool. There have been various efforts in modelling the diesel oxidation catalyst (DOC) with varying degree of complexity. The simplest 1D models discretize the monolithic channel axially and use an effectiveness factor to account for different washcoat geometries. The more complex 1+1D models also resolve the catalyst washcoat which makes them able to better predict efficiency of e.g. layered catalysts. However, the vast majority of these 1+1D models assume the washcoat to be a uniform slab with homogeneous properties. Thus, they cannot identify tangential washcoat variations which have been found to have substantial effects on washcoat diffusivity (1). In this work a new parallel 1+1D reactor model has been developed. Similar to the sectionalizing method presented by Papadias et al (2), the washcoat is sliced into multiple tangential sections based on an evenly distributed angle. The model then solves each section of the washcoat independently – assuming that no mass is transferred between each section. Preliminary results (see figure 1) compare NO light-off simulations using the original 1+1D model and the parallel model using 3 sections. If the washcoat properties for each section are kept constant, then the conversion efficiency of the parallel model is slightly decreased because of the increased diffusive resistance of the thicker corners. However, if including the local porosity of each section (where the corners show a 13% increase in local porosity based on SEM images), the conversion efficiency of the parallel model is significantly higher because of the increased pore diffusivity in the corners. The proposed methodology enables analysis of local washcoat properties. The results are important for high performance modelling towards zero emission vehicles.