Simulation of Ceramic DPF Media, Soot Deposition and Pressure Simulation of Ceramic DPF Media, Soot Deposition and Pressure Simulation of Ceramic DPF Media, Soot Deposition and Pressure Simulation of Ceramic DPF Media, Soot Deposition and Pressure

Drop Evolution UsingDrop Evolution UsingDrop Evolution UsingDrop Evolution Using

Dr. Stefan Rief

Dipl.-Math. Kilian Schmidt

Dr. Andreas Wiegmann

FraunhoferFraunhoferFraunhoferFraunhofer----Institut Institut Institut Institut

TechnoTechnoTechnoTechno---- und Wirtschaftsmathematik, Kaiserslauternund Wirtschaftsmathematik, Kaiserslauternund Wirtschaftsmathematik, Kaiserslauternund Wirtschaftsmathematik, Kaiserslautern

GermanyGermanyGermanyGermany

Page 2

Outline

1. Virtual Structure Generation of DPFs

• Fibrous Structures

• Sinter Materials

2. Simulation of DPF Properties

• Modeling and Simulation Approach

• Example 1: Design Study of DPF Filters

• Example 2: Micro Sieves

3. Summary

Page 3

1. Virtual Structure Generation of DPFs

Multilayer Virtual Fibrous Structure Multilayer Virtual Fibrous Structure Multilayer Virtual Fibrous Structure Multilayer Virtual Fibrous Structure

• Stochastic generation of the structure with

guaranteed adjustable properties, e.g.

– Distribution of fiber diameters and cross

sections

– Fiber orientation

– Porosity

– Layer thickness

– …

• Stacking of layers with different parameters

• Use of highly flexible voxel meshes

Page 4

1. Virtual Structure Generation of DPFs

Virtual Sinter StructureVirtual Sinter StructureVirtual Sinter StructureVirtual Sinter Structure

• Stochastic generation based on

– Packings of spheres

– Morphological operations (to generate

sinter necks)

• Packings of spheres selected to match the initial

grain size distribution of the sinter process

• Approach was applied in an industrial project

when no tomographies were available due to

– Difficult preprocessing of samples

– Too coarse resolution

Page 5

1. Virtual Structure Generation of DPFs

SEM image Virtual Reconstruction

Basic shapes I

Inversion

Erosion

“Sand”

Basic shapes II

Inversion

Page 6

1. Virtual Structure Generation of DPFs

Quality Measures for Virtual StructuresQuality Measures for Virtual StructuresQuality Measures for Virtual StructuresQuality Measures for Virtual Structures

• “The Eye”

• Porosity, specific surface area

• Chord length distribution

• Pore size analysis

• Flow properties, e.g. effective permeability or

flow resistivity

• Bubble point, capillary pressure curves

• Filtration properties

• Acoustic properties

Comparison of Effective Flow Properties

0

0,5

1

1,5

2

2,5

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200

Pressure Drop [Pa]

Ve

loc

ity

[m

/s]

Simulation

Experiment

Page 7

2. Simulation of DPF Properties

Flow

Single Fiber Simulation

1. Choose initial structural parameters

2. Generate / modify structure

3. Solve CFD problem

4. Compute particle transport and deposition

5. Compute filtration efficiency and pressure drop

6. Choose new material parameters

OPTIMIZATION

Simulation of Filtration ProcessesSimulation of Filtration ProcessesSimulation of Filtration ProcessesSimulation of Filtration Processes

LIFETIME

Page 8

2. Simulation of DPF Properties

Flow simulation is based on NavierFlow simulation is based on NavierFlow simulation is based on NavierFlow simulation is based on Navier----StokesStokesStokesStokes----Brinkmann equationsBrinkmann equationsBrinkmann equationsBrinkmann equations

Henry Darcy (1803-1858)

RemarkRemarkRemarkRemark

• convective term optional -> fast flow

• Brinkmann term optional ->subgrid particle deposition and effectiveporous media

Page 9

2. Simulation of DPF Properties

Lagrangian Particle TransportLagrangian Particle TransportLagrangian Particle TransportLagrangian Particle Transport

Particle DepositionParticle DepositionParticle DepositionParticle Deposition

• Collision handling

• Adhesion model

Modification of GeometryModification of GeometryModification of GeometryModification of Geometry

• Solid deposition model (particles resolved by voxels)

• Porous deposition model -> small particles are handled as porous media

• Porous deposition model + subvoxel collision handling

Page 10

2. Simulation of DPF Properties

Particulate filter medium with fiber layer

Design Study of a Diesel Particulate FilterDesign Study of a Diesel Particulate FilterDesign Study of a Diesel Particulate FilterDesign Study of a Diesel Particulate Filter

• What is the effect of an additional fibrous layer on top of a sintered substrate ?

• Soot particles (~80nm) are much smaller than

voxels (1µm) -> porous deposition model

• Navier-Stokes-Brinkmann model to handle free and porous flow

• Permeability and maximum degree of filling of porous voxels are determined by high resolution

single fiber experiments

• Several hundreds of millions of particles are needed for a lifetime computation

Page 11

2. Simulation of DPF Properties

Cut-Out of Soot LayerHigh Resolution Single Fiber Simulation

Page 12

2. Simulation of DPF Properties

Page 13

2. Simulation of DPF Properties

Micro Sieves Micro Sieves Micro Sieves Micro Sieves –––– Study of Different Deposition ModelsStudy of Different Deposition ModelsStudy of Different Deposition ModelsStudy of Different Deposition Models

• 20 µm x 20 µm holes in periodic arrangement

• Soot particles (~80nm) are much smaller than

voxels (1µm)

• Navier-Stokes-Brinkmann model to handle free and porous flow

• Comparison of porous voxel approach w/o subvoxel collision handling

Page 14

2. Simulation of DPF Properties

Porous Deposition Model Porous Deposition Model with Subvoxel Collision Handling

Page 15

2. Simulation of DPF Properties

Backpressure Evolution

Filter Efficiency Evolution

Page 16

2. Simulation of DPF Properties

Page 17

3. Summary (and more …)

---- FiberFiberFiberFiberGeoGeoGeoGeo,,,, SinterSinterSinterSinterGeoGeoGeoGeo, WeaveWeaveWeaveWeaveGeoGeoGeoGeo, GridGridGridGridGeoGeoGeoGeo, , , , PackPackPackPackGeoGeoGeoGeo (Structure generation)

---- ProcessProcessProcessProcessGeoGeoGeoGeo (3d image processing)

---- LayerLayerLayerLayerGeoGeoGeoGeo (building media stacks)

---- ImportImportImportImportGeoGeoGeoGeo (Tomography, STL, etc.)

---- PoroPoroPoroPoroDictDictDictDict (Pore size analysis)

---- FlowFlowFlowFlowDictDictDictDict (Flow properties)

---- FilterFilterFilterFilterDictDictDictDict (Filtration)

---- DiffuDiffuDiffuDiffuDictDictDictDict (Effective diffusion)

---- SatuSatuSatuSatuDictDictDictDict (Capillary pressure curves)

---- ElastoElastoElastoElastoDictDictDictDict (Effective elasticity)

---- ThermoThermoThermoThermoDictDictDictDict (Heat conductivity)

---- ExportExportExportExportGeoGeoGeoGeo (Fluent, Abaqus)

---- AcoustoAcoustoAcoustoAcoustoDictDictDictDict (acoustic absorption properties)

Page 18

GeoDict Development Teams

The FiberGeo TeamAndreas WiegmannKatja SchladitzJoachim Ohser

Hans-Karl HummelPetra Baumann

WeaveGeo & PleatGeoAndreas Wiegmann

The SinterGeo TeamKilian SchmidtNorman Ettrich

The ElastoDict TeamHeiko Andrä

Dimiter StoyanovAndreas Wiegmann

Vita RutkaDonatas Elvikis

GridGeo & PackGeoAndreas Wiegmann

FlowDict Lattice Boltzmann Team

Dirk KehrwaldPeter KleinDirk MertenKonrad SteinerIrina Ginzburg

Doris Reinel-Bitzer

FlowFlowFlowFlowDictDictDictDict EJ Solver TeamAndreas WiegmannLiping ChengAivars ZemitisDonatas Elvikis

Vita RutkaQing Zhang

The FilterDict TeamStefan RiefKilian SchmidtArnulf Latz

Andreas WiegmannChristian WagnerRolf Westerteiger

The SatuDict TeamJürgen BeckerVolker Schulz

Andreas WiegmannRolf Westerteiger

The RenderGeo TeamsCarsten LojewskiRolf WesterteigerMatthias Groß

The GeoDict TeamAndreas WiegmannJürgen BeckerKilian SchmidtHeiko Andrä

Ashok Kumar VaikuntamRolf WesterteigerChristian WagnerMohammed Alam

Jianping Shen

The PoroDict TeamAndreas WiegmannJürgen BeckerRolf Westerteiger

The PleatDict TeamAndreas Wiegmann

Oleg IlievStefan Rief

Page 19

Software for Generation, Simulation, Visualization:

www.geodict.comThank You Very Much for Your Attention !