A 2-D Computational Model Describing the Heat Transfer, Reaction Kinetics and Regeneration Characteristics of a Ceramic Diesel Particulate Trap 980546

A 2-D CFD model was developed to describe the heat transfer, and reaction kinetics in a honeycomb structured ceramic diesel particulate trap. This model describes the steady state as well as the transient behavior of the flow and heat transfer during the trap regeneration processes.

The trap temperature profile was determined by numerically solving the 2-D unsteady energy equation including the convective, heat conduction and viscous dissipation terms. The convective terms were based on a 2-D analytical flow field solution derived from the conservation of mass and momentum equations (Opris, 1997). The reaction kinetics were described using a discretized first order Arrhenius function. The 2-D term describing the reaction kinetics and particulate matter conservation of mass was added to the energy equation as a source term in order to represent the particulate matter oxidation. The filtration model describes the particulate matter accumulation in the trap. This model includes the diffusion, direct interception, and inertia mechanisms, and are linked together to capture the overall filtration characteristics of the trap (Opris and Johnson, 1998).

The effect of trap material proper-ties on trap regeneration behavior was studied with and without a Cu fuel additive for controlled and uncontrolled regeneration tests. The theoretical results compared well to the experimental data. The regeneration model indicated that the fundamental difference between SiC and Cordierite is to be found in the trap design (wall thickness = 0.8 mm for the tested SiC vs. 0.43 mm with Cordierite), and trap material properties (thermal conductivity, density and specific heat where 11 W/m/K, 1600 kg/m³, and 750 J/kgK with SiC, and 0. 1 W/m/K, 1000 kg/m³ and 600 J/ kgK with Cordierite). These differences are largely responsible for the regeneration behavior difference.

Using the regeneration model, it was determined that the temperature gradients (dT/dt) for similar conditions, were up to five times larger with Cordierite, when compared to SiC. The regeneration times with SiC were more than twice as long for the tested conditions. These results were also confirmed experimentally (Gantawar et al., 1996 and Awara et al., 1996), and are consistent with the difference in trap design and material properties. For all tested conditions (SiC and Cordierite, with and without the fuel additive, controlled and uncontrolled regeneration tests), the computed results indicated that the regeneration process is driven by the local exhaust gas temperature. Since all regeneration processes were initiated by ramping the engine (flow rates and temperatures) from a low reaction rate condition (∼300°C) to a higher reaction rate condition (>450-500°C), the regeneration process was always initiated at the trap inlet. However, as the local reactions increased the local gas temperatures, and as some of the released energy was convected downstream in the channel, the maximum temperatures were always calculated (and measured) towards the end of the inlet channel indicating a local energy storage. Consequently, the reaction rates increased locally even more, yielding a local particulate oxidation rate higher than the rest of the inlet channel. For all tested cases, the particulate matter was first oxidized towards the end of the inlet channel. As regions of the inlet cell were partially or completely oxidized, the reaction front moved to upstream regions where there was availability of particulate matter. Consequently, the reaction front moved from downstream to upstream locations.

The parametric study performed in this research indicated that an optimum trap configuration can be achieved by changing the material properties (e.g. thermal conductivity similar to SiC's and density and specific heat similar to Cordierite's). From a filtration/trap pressure drop point of view, the trap pore size could be increased to approximately 30 μm and a trap porosity of approximately 45%.