Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using Hoon Lee Center for Transportation Research Argonne National Laboratory Orlando, Florida, USA March 19, 2013
Outline Objective Filtration Algorithm Cell Value Localization Background Built-in Function Utilization Recursive Operation Diesel Engine & DPF Two Approaches in Filtration Modeling Computing Environment Theoretical Analysis Model Results Pressure Drop Model Soot Filtration Model Channel-flow Profiles & Pressure Drop Wall-flow Rearrangements Experiment Local Soot Mass Deposited Local Collection Efficiency Soot Cake Layer Properties Model Setup Summary Domain Setup & Meshing Physical Assumptions & Boundary Conditions Future Work Acknowledgement Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Objective Background Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Objective To study quantitative analysis of soot filtration processes in DPF (diesel particulate filter) systems by developing a three dimensional model using a commercial CFD package, STAR-CCM+. To analyze the time evolution and spatial distributions of local filtration parameters – e.g. porosity, soot mass, collection efficiency, soot cake profile - for each filtration period, along with evaluations of flow properties and pressure drop characteristics across the DPF. Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Background Objective Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Background (1/2) Diesel Engine Diesel Particulate Filter (DPF) Highway diesel vehicles are required to meet Pros: High Thermal Efficiency, Fuel Economy, the stringent PM emission standards. Torque, Low Emission (CO, UHC) Cons: Noise, Vibration, $, Emission ( PM , NO X ) Physically trap ( Filtration ), and chemically oxidize PM ( Regeneration ) periodically. PM and NO X are the major emissions Uncontrolled regeneration may occur which regulated. (USA: EPA Tier 4, EU: Euro 5 / 6) causes system failure due to highly exothermic reaction. NO X reduction by SCR and/or EGR PM needs to be reduced in both mass and Prediction of particulate deposition number... ☞ Arising issue for GDI engines, too! in the porous filter wall is important. Porous wall Plug 4 Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Background Objective Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Background (2/2) Two Approaches in Filtration Modeling Lagrangian : Qualitative analysis by tracking Eulerian : Quantitative analysis of soot filtration particle trajectories with appropriate B.C.s. process by coupling specific filtration algorithms. S. Bensaid et al. Chem. Eng. J. , 2009. P. Tandon et al. Chem. Eng. Sci. , 2010. Chem. Eng. Sci. , 2010. H. Lee et al. 2012 DEER Conference H. Kato et al. Int. J. Engine. Res. , 2011. H. Lee et al. SAE 2013-01-1583 Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Objective Background Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Theoretical Analysis (1/2) Pressure Drop Model Δ𝑄 = Δ𝑄 𝑞𝑝𝑠𝑝𝑣𝑡 𝑥𝑏𝑚𝑚 + Δ𝑄 𝑡𝑝𝑝𝑢 𝑑𝑏𝑙𝑓 + Δ𝑄 𝑔𝑠𝑗𝑑𝑢𝑗𝑝𝑜 + Δ𝑄 𝑑𝑝𝑜𝑢/𝑓𝑦𝑞𝑏𝑜𝑡 𝑥 𝜍𝑣 2 𝜍𝑣 2 𝜈 𝜈 3𝑏 2 𝑉 𝑝,𝑗𝑜 𝑀𝜊 + 𝜈𝐺 𝜈𝐺 𝑣 𝑥 𝑥 𝑡 + 𝛾𝜍𝑣 𝑥2 𝑥 𝑡 = + 𝑣(𝑦) 𝑒𝑦 + 3𝑏 2 𝑉 𝑝,𝑝𝑣𝑢 𝑀𝜊 + ζ 𝑑𝑝𝑜𝑢 2 + ζ 𝑓𝑦𝑞 𝑙 𝑝 𝑙 𝑡𝑝𝑝𝑢 2 0 Darcy- Forchheimer’s Law w Each velocity term is defined as, a - 2w 4𝑏𝑀 = 𝑉 𝑝 𝑏 2 𝑣 𝑥 = 𝑅 𝑝 = 𝑉 𝑝 𝐵 𝑝 4𝑏𝑀 = 𝑉 𝑝 𝑏 w s 4𝑀 : Clean filter condition for A o 𝐵 𝑔𝑗𝑚𝑢 𝑥 𝑥 𝑥 a 𝑅 𝑝 𝑅 𝑝 𝑒𝑦 = 𝑅 𝑝 𝑏 𝑣(𝑦)𝑒𝑦 = 𝑒𝑦 = 8𝑀 ln 𝑏 − 2𝑥 𝐵 𝑔𝑗𝑚𝑢 (𝑦) 4( 𝑏 − 2(𝑥 − 𝑦 )𝑀 0 0 0 𝑅 𝑅 16𝑅 𝑉 𝑝,𝑗𝑜 = = = 𝜌𝐸 2 σ(𝑏 − 2𝑥) 2 : Soot cake 𝜌𝐸 2 2 (𝑏 − 2𝑥) 2 𝐵 𝑝 4 1 2 1 𝑗𝑜𝑚𝑓𝑢 thickness ( w ) for A o (𝑏 + 𝑥 𝑡 ) 2 𝑅(𝑏 − 2𝑥) 2 = 16𝑅(𝑏 + 𝑥 𝑡 ) 2 𝑅𝐵 𝑝 𝑅 𝑝 = = 𝐵 𝑝 𝜌𝐸 2 4 1 2 1 (𝑏 − 2𝑥) 2 𝜌𝐸 2 𝑗𝑜𝑚𝑓𝑢 2 (𝑏 + 𝑥 𝑡 ) 2 𝑅 𝑣 = 𝑂𝑏 2 𝑅, 𝑉 𝑅 𝑝 , 𝑉 𝑝 , 𝑣 𝑥 , 𝑣(𝑦) Half-cut sample for experiment Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Objective Background Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Theoretical Analysis (2/2) Soot Filtration Model Unit Collector Mechanism Collection Efficiency Diffusional collector Deposition ( 𝜽 𝑬 ) Particulates Flow-line collector 𝒆 𝒅𝟏 = 𝟒(𝟐 − 𝜻 𝟏 ) 𝟒 𝒆 𝒅𝟏 Interception ( 𝜽 𝑺 ) 𝒆 𝒒𝒑𝒔𝒇 𝒄 𝟒 = 𝟐 − 𝜻 𝟏 𝟑𝜻 𝟏 −2 3 𝜁 𝑗, 𝑢 = 1 − 𝑒 𝑑 (𝑗, 𝑢) 𝑉 𝑗 𝑒 𝑑 3 𝜃 𝐸 = 3.5 𝜁 𝑄𝑓 −2 3 = 3.5 𝜁 (1 − 𝜁 0 ) 𝑒 𝑑0 𝐸 1 3 𝜁 3 3 3 𝑛 𝑚𝑝𝑑𝑏𝑚 (𝑗, 𝑢) + 𝑒 𝑑0 2 𝜃 𝑆 = 1.5 𝑂 𝑆 𝑒 𝑑 (𝑗, 𝑢) = 2 Key parameters Key parameters 3−2𝜁 4𝜌 𝜍 𝑡𝑝𝑝𝑢,𝑥𝑏𝑚𝑚 2 1 + 𝑂 𝑆 3𝜁 for User Code to for CFD code 2 𝑔(𝜁 𝑗, 𝑢 ) obtain local soot (UDF) to specify 𝑒 𝑑 (𝑗, 𝑢) 𝜃 𝐸𝑆 = 𝜃 𝐸 + 𝜃 𝑆 − 𝜃 𝐸 𝜃 𝑆 𝑙 𝑗, 𝑢 = 𝑙 0 𝑔(𝜁 0 ) mass ( m w ) region properties 𝑒 𝑑0 𝐹 𝑗, 𝑢 = 1 − 𝑓𝑦𝑞 − 3𝜃 𝐸𝑆 1 − 𝜁 𝑗, 𝑢 Δ𝑧 𝛸 𝑢 = 𝑒 𝑑 (𝑗, 𝑢) 2 − 𝑒 𝑑0 2 2𝜁(𝑗, 𝑢) 𝑒 𝑑 (𝑗, 𝑢) 2 𝛺 𝑐 2 − 𝑒 𝑑0 Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Theoretical Model Filtration Computing Model Future Experiment Objective Background Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Experiment 2” x 6” cordierite DPF Test Results Clean Filter Test Soot Loading Test 10.0 8.0E-13 Permeability, k o [m 2 ] Pressure Drop [kPa] ◆ 100 CPSI (w s =17 mils) 9.0 7.0E-13 ◆ 200 CPSI (w s =12 mils) 8.0 6.0E-13 k o =2.30E-13 k o =1.77E-13 7.0 5.0E-13 6.0 4.0E-13 5.0 3.0E-13 4.0 2.0E-13 3.0 1.0E-13 2.0 ● Low flow rate (7.4 SCFM): 200CPSI ● High flow rate (9.0 SCFM): 200CPSI 0.0E+00 1.0 0.0E+00 2.0E-03 4.0E-03 6.0E-03 0.0 0 10 20 30 40 50 60 2.4 Pressure Drop [kPa] Time [min.] 2.1 1.8 PM Mass Concentration PM Size Distribution 1.5 1.2 0.9 ● 100 CPSI (w s =17 mils) 0.6 ● 200 CPSI (w s =12 mils) 0.3 ◆ Analytical Solution 0.0 SMPS TEOM 0.0E+00 2.0E-03 4.0E-03 6.0E-03 Vol. Flow Rate [m 3 /s] Clean filter permeability ( k o ) and particle-laden flow properties are directly measured. Ready Soot cake permeability ( k s,cake ), particle density ( ρ s ), and soot cake porosity ( ε s,cake ) can be estimated. to model Packing densities ( ρ s,wall, ρ s,cake ) are assumed. Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Model Theoretical Filtration Computing Model Future Objective Background Experiment Summary Acknowledgement Analysis Setup Algorithm Environment Results Work Model Setup (1/2) Domain Setup Meshing Geometry is based on a 200CPSI, lab-scaled Volume meshes are generated by using (2”x 6”) cordierite filter with regions of upstream Trimmer for porous regions (filter wall, soot flow and soot cake formation. cake), and Polyhedral for fluid and solid regions (channels, plugs). Upstream= L 0.78”, Plug= L 0.39”, w s = 12.0 mils Total 2,013,762 cells CFD domain 10 9 w 8 7 6 5 4 3 w s 2 y 1 z a x Trimmer is exclusively used for filter wall, consisting of 10 separate porous regions , All cells in wall regions to represent soot filtration. (Growth rate = 1, Cell size = thickness of each region) are regular hexahedrons Upstream I/O Channels Plugs Filter wall Soot cake POROUS REGION FLUID REGION SOLID REGION Integrating Filtration Mechanism with a 3D Diesel Particulate Filter (DPF) Model using STAR-CCM+
Recommend
More recommend
