Modeling of gasoline particulate filters with combined use of membrane and catalyst coatings

Grigorios Koltsakis (Aristrotle University Thessaloniki)･Elias Mylonidis･Dimitrios Karamitros (Emisia SA)

The combination of a non-active membrane with a catalytically active coating could maximize the filtration efficiency potential of gasoline particulate filters. The paper addresses the challenges for optimum design and application of such a concept. A physics-based model is initially calibrated vs engine bench experimental data from both membrane and catalytically coated filters. The internal morphology of each filter is obtained by optical techniques and used as input to the model. The validated model is then applied to virtually benchmark alternative designs considering PM/PN filtration efficiency, soot oxidation, ash impact and emissions during filter regeneration.

Development of optimized GPF performances based on GPF Model

Sungmu Choi･Jiho Cho･Hyungjun Kim･Sangmin Lee (Hyundai Motor)

Models of GPF were developed and used to find optimal GPF. At first, coated GPF and membrane GPF were considered. Coated GPF has lower PN filteration efficiency and higher soot regenerability. To increase PN filteration efficiency, removing crack and reducing mean pore size of washcoat on GPF are proposed based on simulation results. PN filteration efficiency of coated GPF was improved to meet with higher soot regenerability. In case of membrane GPF, it has higher PN filteration efficiency and lower soot regenerability. To enhance its soot regenerability, additional coating on membrane GPF was considered, it is called hybrid GPF. The amount of coating on membrane GPF must be optimized in view of soot regenerability and pressure drop. Based on hybrid GPF model, 25g/L~50g/L coating amount was proposed to get effective soot regenerability with allowable pressure drop increase.

2-D Numerical Modeling on the Urea SCR Catalyst Installed in a Diesel Engine

Fuka Yoshida･Jin Kusaka (Waseda University)

Urea selective catalytic reduction (SCR) system is a promising technique to reduce NOx emission from Diesel engines by injecting urea water in front of the catalyst. For numerical modeling on the Cu-CHA catalyst, 31 surface reactions including 5 active sites coupled with the reaction of urea thermal- and hydro-decompositions have been constructed. Moreover, to estimate boundary conditions for simulations, the temperature distributions are measured in the radius direction in front of the SCR catalyst. The results of the numerical simulation constructed have reasonable agreement with the engine test results with consideration of reaction paths and surface chemical species.

An Optimal Development of dual SCR system for Heavy duty Diesel Engine using Genetic Algorithm

Jiho Cho･Sung Moo Choi･Hyung Jun Kim･Sang Min Lee･Dong Min Hwang (Hyundai Motor)

In this study, a predictive model-based SCR control optimization technique using a genetic algorithm is suggested and developed to minimize the development effort of DCU (Dosing Control Unit) mapping especially for cc_SCR system. This technique is an optimization method that automatically finds the cc_SCR control mapping that satisfies the target NOx emission and NH3 slip level in a short time using a genetic algorithm for an integrated prediction model consisting of a DCU model and a catalytic reaction model. By using this technique, it is possible to reduce calibration effort of the cc_SCR system and ensure the robustness of emission controls.

Simulation and Performance Prediction of Spray Droplet Behavior in a Urea-SCR System Using PIV-DDM

Shotaro Nara･Nachi Takeuchi･Joe Ono･YukiYuki Kawamoto･Naoya Fukushima･Masayuki Ochiai･Tetsuo Nohara (Tokai University)

Understanding the flow behavior inside the urea-SCR system is important for predicting exhaust gas purification performance. This study attempts to reproduce the internal flow behavior without performing flow field calculations via CFD (Computational Fluid Dynamics), using a method combining PIV (Particle Image Velocimetry) and DDM (Discrete Droplet Model). A comparison with experiments simulating the inside of the exhaust pipe confirmed the same tendency. This report presents the findings.