자동차용 후처리장치의 유동특성 예측을 위한 수치모델들의 성능평가

Abstract

Computational Fluid Dynamics(CFD) has been widely used in the automotive industry to simulate the flow within the converter. Most of CFD models for predicting flow through converter assume a monolith resistance based on one-dimensional laminar flow and treat a monolith as porous medium. However, auto-catalyst is getting closer to engine to have fast warm-up and wide-angled diffusers are often employed to connect the upstream exhaust pipe to the monolith. Therefore, the question arises as to the accuracy of porous medium approach based on one-dimensional laminar flow and model validation studies are essential if it is to become a useful design tool. From these reasons, this paper evaluates the prediction performance of porous medium and three-dimensional multichannel model by comparing with experimental results for understanding the limitation of porous medium approach and suggesting guidelines for the development of more accurate flow model. For evaluating each models, this work calculates the flow mal distribution and pressure drop across the monolith of an axisymmetric catalyst assembly under steady state condition. The results shows that porous medium model under-predicts flow maldistributions in the monolith by 6.4% at maximum and three-dimensional multi-channel model also under-predicts by 10.3% at maximum. From the results of this work, it can be concluded that modification of porous medium approach is strongly required to account of additional entrance effect and detailed modeling of monolith channels can not be alternative for more accurate flow simulation of auto-catalyst due to the inaccuracy Averaged Navier Stokes k-ε type turbulence model under abrupt change of length scale condition.