Fully Transient One-Dimensional Based Diesel Particulate Filter Modeling Including Catalytic Surface Reactions

Christopher  Depcik, University of Michigan

Classical one-dimensional based Diesel Particulate Filter (DPF) modeling has been predicting pressure drops, the oxidation of diesel particulate and the exothermic reactions that result with good accuracy for over the last 20 years. To advance the state of the art and to account for the possibility of catalyzed filters, this classical model was enhanced to include the propagation of chemical species and particulates into the formulation of the model. In addition, full transient capability was incorporated into the model to account for regulatory tests where the inlet conditions to the DPF can change instantly. This transient capability allows the model to capture surface intermediate chemistry effects where the timescales of the process are much smaller compared to surface temperature timescales. While a compressible formulation of the equations of motion for the model will yield the most accurate results, it is rather impractical for multiple simulation runs because of its long computational time. As a result, an incompressible formulation is described and compared to full variable-property reacting-gas dynamics to illustrate differences between the models. It was found that the source terms used in the formulation of the DPF model are large enough to offset most of the effects of compressibility. As with previous catalyst modeling efforts, all pertinent assumptions and numerical solvers are presented.

Download oral presentation: