Develop ents in
`Adv ced
`Cera ics
`d
`Co posites
`
`í .Editors
`Manuel E. Bato
`Peter Fi hp
`Charles Lewinsohn
`Ali Sayir
`Mark Opeka
`Witham M. Mullins
`
`ENGINEERING LIBRARY
`DISPLAY PERIODICAL
`Non -circulating until:
`SEP 1 8 200
`
`Proceedings of the 29th International Conference on Advanced
`Ceramics and Composites, Cocoa Beach, Florida, USA (2005)
`
`General Editors
`Dongmmg Zhu
`Waltraud M Krasen
`
`utiivERENGINEWASH. rò
`
`PERIODICALS
`AUG 1 G 2005
`
`UBRARIES
`
`BASF-2020.001
`
`

`
`Developments in Advanced
`Ceramics and Composites
`
`A collection of papers presented at the
`29th International Conference
`on Advanced Ceramics and Composites,
`January 23 -28, 2005,
`Cocoa Beach, Florida
`
`Editors
`Manuel E. Brito
`Peter Filip
`Charles Lewinsohn
`Ali Sayir
`Mark Opeka
`William M. Mullins
`
`General Editors
`Dongming Zhu
`Waltraud M. Kriven
`
`a.r
`
`crt`y"c
`
`,.o.,comnic.oto
`
`Published by
`The American Ceramic Society
`735 Ceramic Place
`Suite 100
`Westerville, Ohio 43081
`www.ceramics.org
`
`BASF-2020.002
`
`

`
`Developments in Advanced Ceramics and Composites
`
`Copyright 2005. The American Ceramic Society. All rights reserved.
`
`Statements of fact and opinion are the responsibility of the authors alone and do not
`imply an opinion on the part of the officers, staff or members of The American Ceramic
`Society. The American Ceramic Society assumes no responsibility for the statements
`and opinions advanced by the contributors to its publications or by the speakers at its
`programs; nor does The American Ceramic Society assume any liability for losses or
`injuries suffered by attendees at its meetings. Registered names and trademarks, etc.,
`used in this publication, even without specific indication thereof, are not to be consid-
`ered unprotected by the law. Mention of trade names of commercial products does not
`constitute endorsement or recommendation for use by the publishers, editors or
`authors. Final determination of the suitability of any information, procedure or product
`for use contemplated by any user, and the manner of that use, is the sole responsibility
`of the user. This book is intended for informational purposes only. Expert advice should
`be obtained at all times when implementation is being considered, particularly where
`hazardous materials or processes are encountered.
`No part of this book may be reproduced, stored in a retrieval system or transmitted in
`any form or by any means, electronic, mechanical, photocopying, microfilming, record-
`ing or otherwise, without written permission from the publisher.
`Authorization to photocopy for internal or personal use beyond the limits of Sections
`107 and 108 of the U.S. Copyright Law is granted by The American Ceramic Society,
`provided that the appropriate fee is paid directly to the Copyright Clearance Center,
`Inc., 222 Rosewood Drive, Danvers, MA01923 U.S.A., www.copyright.com. Prior to
`photocopying items for educational classroom use, please contact Copyright Clearance
`Center, Inc.
`
`This consent does not extend to copying items for general distribution or for advertising
`or promotional purposed or to republishing items in whole or in part in any work in any
`format.
`
`Please direct republication or special copying permission requests to Staff Director,
`Technical Publications, The American Ceramic Society, 735 Ceramics Place, Suite 100,
`Westerville, Ohio 43081, USA.
`
`For information on ordering titles published by The American Ceramic Society, or to
`request a publications catalog, please call 614 -794 -5890, or visit www.ceramics.org
`
`Printed in the United States of America.
`
`ISSN 0196 -6219
`
`ISBN 1-57498-261-3
`
`BASF-2020.003
`
`

`
`l
`
`Contents
`
`Preface
`Ceramics in Environmental Applications
`Characterization of MnO -Doped Lanthanum Hexaluminate (LaMnAttiO19) in
`Terns of Selective Catalytic Reduction of NOx by Addition of Hydrocarbon
`Reductant (HC-SCR) ...... . .... . . . .... . ............ .
`. ............3
`
`ix
`
`M. Stranzenbach and B. Saiuhan
`High Porosity Cordierite Filter Development for NOx/PM Reduction ........11
`I. Melscoet -Chauvel, C. Remy, and T. Tao
`
`Thermal Stability of Cordierite Supported V205- W03 -11O2 SCR
`Catalyst for Diesel NOx Reduction
`Y. Xie, C. Remy, L Melscoet -Chauvel, and T. Tao
`
`A New Family of Uniformly Porous Composites with 3 -D Network Structure
`(UPC -3D): A Porous A1203/LaPO4 In Situ Composite
`Y. Suzuki, P.E.O. Morgan, and S. Yoshikawa
`
`Novel, Alkali- Bonded, Ceramic Filtration Membranes
`5. Mallicoat, P. Sarin, and W.M. Kriven
`
`Controlling Microstructural Anisotropy During Forming
`S.M. Nycz and R.A. Haber
`
`Characterization of LZSA Glass Ceramics Filters Obtained by the
`Replication Method
`C. Silveira, E. Sousa, E. Morses, A.P.N. Oliveira, D. Hotta, T. Fey, and P. Grell
`
`21
`
`31
`
`37
`
`45
`
`53
`
`Fracture Behavior and Microstructure of the Porous Alumina Tube
`C.-H. Chen, S. Honda, and H. Awaji
`Tensile Testing of SiC -Based Hot Gas Filters at 6009C Water Vapor .......69
`R. Paslila, A.-P. Nikkila, T. Mantylä, and E. Lara-Curzio
`
`61
`
`v
`
`BASF-2020.004
`
`

`
`Quasi -Ductile Behavior of Diesel Particulate Filter Axial Strength Test
`Bars with Ridges
`G.M. Crosbie and R.L. Allur
`
`Multifunctional Material Systems Based on Ceramics
`Multifunctional Electroceramic Composite Processing by
`Electrophoretic Deposi tan
`G. Falk, M. Bender, and R. Clasen
`
`Transparent Alumina Ceramics with Sub -Microstructure by Means of
`Electrophoretic Deposition
`A. Braun, M. Wein, G. Falk, and R. Clasen
`
`77
`
`87
`
`97
`
`Functional Nanoceramic Coatings on Microstructured Surfaces via
`Electrophoretic Deposition
`H. von Both, A. Plrengle, and J. Haueell.
`High Damping in Piezoelectric Reinforced Metal Matrix Composites ......113
`B. Poquede, J. Schullz, T. Asare, S. Kampe, and A. Aning
`
`105
`
`Carbon/Carbon and Ceramic Composite Materials in Friction
`Preparation Of Large -Scale Carbon Fiber Reinforced Carbon Matrix
`Composites (C -C) By Thermal Gradient Chemical Vapor
`Infiltration (TGCVI)
`J. Lee and J.H. Park
`
`Frictional Performance and Local Properties of C/C Composites
`S. Ozcan, M. Krkoska, and P. Fllip
`
`Humidity and Frictional Performance of C/C Composites
`M. Krkoska and P. Fllip
`
`Study of "Adsorption/Desorption" Phenomena on Friction Debris of
`Aircraft Brakes
`K. Peszynska -Bialczyk, M. Krkoska, A. Pawliczek, P. Fillet and K. Anderson
`
`Friction and Wear of Carbon Brake Materials
`J.A. Tanner and M. Travis
`
`.121
`
`127
`
`139
`
`157
`
`167
`
`Processing and Friction Properties of 3D -C /C -SiC Model Composites with a
`Multilayered C -SiC Matrix Engineered at the Nanometer Scale ..........179
`A. Fillion, R. Naslain, R. Pailler, X. Bourrat, C. Robin -Brosse, and M. Brendle,
`
`vi
`
`BASF-2020.005
`
`

`
`Carbon Fiber -Reinforced Boron Carbide Friction Materials
`R.J. Shinayski, K. -C. Wang, R Filip, and T. Polìcandriotes
`
`Thermal Shock Impact on C/C and Si Melt Infiltrated C/C Materials (SiMI)
`D.E. Widmer and P. Filip
`
`187
`
`...195
`
`Reliability of Ceramic and Composite Components
`
`Post Engine Test Characterization of Self Sealing Ceramic Matrix
`Composites for Nozzle Seals in Gas Turbine Engines
`207
`E. Bouillon, C. Louchet, P. Sprier, G. Olard, D. Felndel, C. Logan, K. Rogers, and T. Arnold
`
`Dimension Stability Analysis of NITE SiC /SiC Composite Using Ion
`Bombardments for the Investigation of Reliability as Fusion Materials
`H. Kishimoto, T. Hinoki, K. Ozawa, K.-H. Park, S. Kondo, and A. Kohyama
`Fracture Strength Simulation of SiC Microtensile Specimens - Accounting for
`Stochastic Variables
`223
`N.N. Nemeth, G.M. Behelm, O.M. Jadean, W.N. Sharpe, G.D. Quinn,
`L.J. Evans, and M.A. Trapp
`
`215
`
`Design and Reliability of Ceramics: Do Modelers, Designers, and
`Fractographers See the Same World?
`G.D. Quinn
`
`The Effects of Incorporating System Level Variability into the Reliability
`Analysis for Ceramic Components
`R. Carter and O. Jadaan
`
`Finite- Element -Based Electronic Structure Calculation in Metal/Ceramic
`Interface Problems
`Y. Shilhara, O. Kuwazuru, and N. Yoshikawa
`
`3D FEM Simulation of MLCC Thermal Shock
`Y.H. Moon and H.J. Youn
`
`Analysis of Firing and Fabrication Stresses and Failure in Ceramic -Lined
`Cannon Tubes
`J.H. Underwood, M.E, Todaro, M.D. Witherell, and A.P. Parker
`
`Characterization Tools for Materials Under Extreme Environments
`On the Comparison of Additive -Free HfB2 -SiC Ceramics Sintered by
`Reactive Hot -Pressing and Spark Plasma Sintering
`F. Monleverde and A. Bello&
`
`239
`
`253
`
`261
`
`269
`
`281
`
`295
`
`vii
`
`BASF-2020.006
`
`

`
`Dynamic Analyses of the Thermal Stability of Aluminum Titanate by
`Time -of- Flight Neutron Diffraction
`I.M. Low, D. Lawrence. A. Jones, and R.I. Smith
`
`Characterizing the Chemical Stability of High Temperature Materials for
`Application in Extreme Environments
`E. Opila
`
`Effect of Oxygen Partial Pressure on the Phase Stability of Ti3SiC2
`I.M Low, Z. Oo, B.H. O'Connor, and K.E. Prince
`
`Mechanical Behavior Characterization of a Thin Ceramic Substrate at
`Elevated Temperature Using a Stereo -Imaging Technique
`S. Widlaja, K.L. Geisinger, and S.C. Pollard
`
`Functional Nanomaterlal Systems Based on Ceramics
`Synthesis and Characterization of Cubic Silicon Carbide (B -SiC) and
`Trigonal Silicon Nitride (a- Si3N4) Nanowires
`K. Saulig- Wenger, M. Bechelany, D. Cornu, S. Bernard, F. Chassagneux,
`P. Miele, and T. Epicier
`
`High Energy Milling Behavior of Alpha Silicon Carbide
`M. Aparecida Plnheiro dos Santos and C. Albano da.Cosla Nato
`
`Synthesis of Boron Nitride Nanotubes for Engineering Applications
`J. Hurst, D. Hull, and D. Gorican
`
`Comparison of Electromagnetic Shielding in GFR -Nano Composites
`W: K. Jung, S.-H. Ahn, and M.-S. Won
`
`303
`
`311
`
`323
`
`331
`
`341
`
`349
`
`355
`
`363
`
`Densification Behavior of Zirconia Ceramics Sintered Using
`High -Frequency Microwaves ......... ............................373
`
`M. Wolff, G. Falk, R. Clasen, G. Link, S. Takayama, and M. Trumm
`
`Manufacturing of Doped Glasses Using Reactive Electrophoretic
`Deposition (REPO)
`D. Jung, J, Tabellion, and R. Clasen
`Shaping of Bulk Glasses and Ceramics with Nanosized Particles ........389
`J. Tabellion and R. Clasen
`
`381
`
`Author Index
`
`397
`
`viii
`
`BASF-2020.007
`
`

`
`HIGH POROSITY CORDIERITE FILTER DEVELOPMENT FOR Nox/PM REDUCTION
`
`Isabelle Melscoet- Chauvel
`Corning Incorporated
`SP- DV -02 -I
`Corning, NY 14831
`USA
`
`Christophe Remy
`Coming S.A.S
`7bìs, Avenue dc Valvins
`77210 Avon
`FRANCE
`
`Tinghong Tao
`Corning Incorporated
`SP- DV -02 -1
`Corning, NY 14831
`USA
`
`ABSTRACT
`This paper presents the latest progress in high porosity filter product development for the
`4 -way application as well as the corresponding property and performance attributes. Two new
`compositions, Dev -HP I and Dev -HP2, at 72% porosity with median porc sizes of 17 pm and 20
`are identified and under development From composition development, narrow pore size
`istnbution and good pore connectivity were achieved in comparison to the previous version of
`igh porosity filters, Dev -EC, and also the standard commercial product, Dural apt CO. As a
`result, the pressure drop performance has been significantly improved while a high filtration
`efficiency of more than 95% has been preserved (artificial soot lab test). The detailed physical
`property and performance data
`for these two new 72% porosity filters (vs. Dev -EC and
`DuraTmp CO) are discussed. It is anticipated that significantly improved porc microstructure
`and high wall permeability will allow high catalyst loading in the wall for the 4-way catalyst
`application.
`
`TRODUCTION
`Diesel engines are the most energy efficient powcrtrains among all types of internal
`ombustion engines known today. This high efficiency translates to very good fuel economy and
`ow greenhouse gas emissions (CO2) which helps reduce the global climate effect Other diesel
`engine advantages that have not been matched by competing energy conversion machines
`include durability, reliability and fuel safety. Nowadays, in Europe, about 50% of passenger cars
`diesel powered, and this trend is increasing, with a diesel sham of new passenger cars in
`estera Europe at 51.9% [I]. Due to the potentially harmful effects on health and on the
`vironment from both NO and PM (particulate matter) emissions, there is a need for a
`uction of these emissions, and the regulations are tightening in Europe, Japan, and the USA
`[2]. For example, US 2010 regulations require tailpipe emission to be 0.2g/bhp -hr NO, and
`0.O1g/bhp -hr PM for heavy duty applications.
`
`To the extent authorized under the taws of the United States of America, all copyright interests in this publication s
`or The AAmerhan Ceramic Society. Any duplication, reproduction, or republication of this publication or any part tt
`the express writtenwnsent of The American Cernait Society or fee paid to the Copyright Citaenace Center, ix pro
`
`y
`
`BASF-2020.008
`
`

`
`To meet US2010 regulations for both PM and NO, emissions,
`there are several
`technologies proposed including diesel particulate filter (DPF), NO, trap or selective catalytic
`reduction (SCR) for NO and their combination. The combination of PM and NO, is the most
`attractive due to expected cost and space saving, but it is also the most challenging technology.
`Toyota has introduced their first combined PM/NO, emission control technology in late 2003 in
`Europe (Toyota Avensis 2.01.) and in Japan (Nino truck 4AL), using the diesel particulate and
`NO reduction (DPNR) catalyst system with NO, trap on cordicrite filter developed jointly by
`Toyota, NCK, DENSO et al. [3,41. With the worldwide tightening regulations, it stems that all
`diesel vehicles sold in these areas (Europe and Japan) will ultimately have integrated NO and
`PM functions. Alternatively, selective catalytic reduction (SCR) catalysts can be used to replace
`NO, trap catalysts in a DPF to provide NO, reduction. Because such a system is also able to
`reduce CO and HC emissions through catalytic oxidation, the system is also called 4 -way
`catalyst system. Both NO, trap and SCR 4 -way systems have their own advantages and
`drawbacks in catalyst technology and system design. However, their NO, reduction performance
`(efficiency and capacity) is dependent upon the total amount of catalyst loading in DPP filter. A
`high porosity filter is the leading approach to achieve high catalyst storage. How to maintain the
`delicate balance between high porosity and thermo-mechanical durability is a challenging,
`problem, and this paper sheds some light on the latest progress in high porosity filter product
`development.
`
`BACKGROUND
`Catalytic converters based on cordieritc have been developed and widely used over the
`past 35 years for the automotive market thanks to a long history of discovery and development
`[5]. Cordierite is a refractory ceramic with a melting temperature around 1450 °C. Another key
`feature of cordicrite is its low coefficient of thermal expansion (typically lower than 4 x 104/°C),
`which explains its excellent thermal shock resistance, an attribute necessary for automotive
`applications. Similar attributes of cordieritc have also been found desirable for diesel particulate
`filter applications.
`
`MATERIALS AND PROPERTIES
`Specific amounts of selected raw materials with carefully controlled particle size
`distributions are batched together, mulled and then extruded to form monoliths, which are then
`dried and fired with specific schedules, before finally being plugged to make a filter (adjacent
`channels alternatively plugged at each end in order to force the diesel exhaust through the porous
`walls acting as a mechanical filter). Figure I. shows the different processing steps followed to
`obtain a cordieritc ceramic monolith.
`
`Mix Clay + Talc + SiOr + A17O3 Raw Materials
`With pore farmers and organic binders
`
`Mull, Extrude, Dry
`u
`Firing a 1380 -1450 °C
`Figure 1. Generic Processing of Cordierite Monolith
`
`The porosity is controlled during the solid state reaction taking place during firing of the
`monolith. Higher and more controllable levels of total porosity can be achieved by adding pore
`
`12
`
`BASF-2020.009
`
`

`
`fanners to the batch. By careful material composition development and subsequent drying and
`firing process optimization, we are able to engineer and control filter pore microstructures and
`their characteristics in terms of total porosity, pore size and pore size distribution.
`1 summarizes the key physical properties for two newly fabricated filters in
`Table
`development, Dev -HPI and Dev -1-1P2, in direct comparison to a previous version high porosity
`filter, Dcv -BC, as well as DuraTrap CO, a commercial cordierite filter product. The detailed
`methods used to measure physical properties of filter materials were described in a previous
`paper [6].
`
`Table 1. Physïcal Properties for High Porosity and Reference Filters
`DuraTrap
`Property /Composition
`Dev -HPI
`Dcv -HP2
`Dev -EC
`CO
`50
`0.4
`
`b4
`0.7
`
`72
`1.0
`
`72
`1.0
`
`Porosity (volume %)
`Intrusion Volume (ml/g)
`Fine Pore Size
`at 10 %pore filling
`dro (Pm)
`Mean Porc Size
`dso (pm)
`Coarse Pore Size
`at 90% pore filling
`dx, (pm)
`d- factor
`(den- dto)/dso
`Coefficient of Thermal
`Expansion (CTE) - (RT to
`1073 K)
`(x104)
`Modulus Of Rupture- MOR
`at RT, in MPa (cells per
`square inchhvall thickness
`in mil)
`E- Modulus
`at RT, in MPa
`Strain (x104)
`Thermal Shock Parameter
`(K)
`Bulk Volumetric Heat
`Capacity (JR K)
`
`5
`
`12
`
`35
`
`5
`
`11
`
`20
`
`9
`
`17
`
`32
`
`10
`
`20
`
`37
`
`0.58
`
`0.54
`
`0.47
`
`0.50
`
`3,5
`
`4.8
`
`8,7
`
`7.5
`
`2.76
`(200/12)
`
`5.52
`
`5.00
`
`1429
`
`500
`
`1.67
`(200/12)
`1.88
`(300/13)
`
`2.36
`795
`I657
`
`420
`
`1.19
`(300/13)
`
`1.25
`(300/13)
`
`1.08
`
`11,03
`
`1267
`
`346
`
`0.98
`
`12.82
`
`1709
`
`346
`
`The parameters dto, dw and (Ito are the pore diameters a respectively 10 %, 50% and 90%
`olumes filled by mercury. The d- factor, (de0- d10)idso, is used o describe the fine fraction of the
`size distribution (PSD). The coefficient of thermal expansion (CTE) of a material reflects
`much this material expands (or contracts) during heating (or cooling). The cell per square
`(cpsi) and wall thickness both indicate the cell geometry of the filter. The mechanical
`strength is measured by using a four -point bending method, which provides the corresponding
`modulus of rupture (MOR), or fracture stress. The elastic modulus (E-Mod, or Young's modulus)
`
`BASF-2020.010
`
`

`
`is a way to measure the degree of stiffness of a material. It is commonly known that for brittle
`materials such as ceramics, the lower the elastic modulus, the lower the stress for a given strain
`(deformationtelongation).
`
`Pore Characteristics and Microstructures
`As shown
`in Table 1, two new filter compositions, Dcv -HPI and. Dev -HP2, were
`identified that achieve ultra -high porosity of over 70 %. This is a substantial increase in porosity
`compared to previous versions of high porosity coniìerite filters which have porosities around
`60% [3,4,7] and n commercial cordicrite filter product of 50% porosity, Duralrap® CO. In
`addition to increased porosity, two distinct median pore sizes at 17 tun and 20 pm were
`developed for these new filter compositions [7]. The reason why these two median pore sizes
`were selected was to gain a better understanding as well as a technical evaluation of the catalyst
`coating response. Moreover, they also enable a deeper understanding of the delicate trade off
`between pore size and mechanical properties, especially for porosity levels as high as 72 %.
`In order to refine the filter pressure drop and catalyst coatability, an effort has been made
`to improve the pore size distribution of these two new filter compositions since the material
`mechanical strength is limited by the largest pore size, not the median pore size. The d- factor is
`used to describe the fine fraction of the pore size distribution, The lower the d- factor, the lower
`the pressure drop and also the better the catalyst coatability expected due to a reduction in the
`fine fraction of pores. The fine pores are always coated first due to the capillary effect in catalyst
`coating process and are expected to have a negative impact on back pressure. As listed in Table 1,
`Dev -HP I and Dev -í1P2 filters have d- factor values of 0A7 and 0.50 respectively, compared to
`0.58 for DuraTrap® CO and 0.54 for Dev -EC.
`Another important aspect of these two new high porosity filters is the significant
`improvement in their mercury intrusion volumes. The intrusion volume in a filter is a direct
`measure of physical space or volume available in the wall for catalyst storage space. There is
`about a 40% increase in intrusion volume for both new compositions compared to previous Dev-
`EC, and about 150% increase vs. DuraTrapt CO.
`The pore characteristics described above can also be seen in their corresponding SEM
`pictures shown in Figure 2. It is obvious that both new high porosity filters show better
`uniformity in pore microstructure and connectivity compared to Dcv -EC and. DuraTrapm CO
`filters. Furthermore, the increased median pore size from 17 um to 20 µm can be observed in the
`microstructures as well, which will allow higher catalyst loading while preserving a reasonably
`low back pressure.
`
`Thermal and Mechanical Properties
`Both Dev -HPI and Dev -HP2 arc early stage experi mental compositions as product
`concepts. From the preliminary process development to make these two new high porosity
`compositions into filters, their coefficients of thermal expansion are at 83 x 10 -7PC and 7.5 x 10-
`71°C respectively, which are higher than the reference filters Day-EC at 4M x 10 "f°C and
`DuraTrap® CO at 3.5 x I0''PC. Clearly, the coefficients of thermal expansion for both Dev -HP I
`and Dev-HP2 preferably need to be decreased to be comparable to DuraTrap® CO.
`As the porosity continues to increase, the filter mechanical strength is impacted M listed
`in Table I, the moduli of rupture for the two new high porosity filters are significantly reduced
`
`14
`
`BASF-2020.011
`
`

`
`DuraTrap® CO
`(50% porosity; 12pm MPS)
`
`Dev-EC
`(64% porosity; 1Ipm MPS)
`
`Dev -HPI
`Dev -HP2
`(72% porosity; 17µm MPS)
`(72% porosity; 20µm MPS)
`Figure 2. Porc Microstructures of DurTrap® CO, Dev -EC, Dev -HP1 and Dev -HP2
`
`(at 172 and 182 psi respectively, even for the 300/13 geometry) compared to more than 240 psi
`for Dev -EC and 400 psi for DuraTrap CO filters (200/12 geometry).
`It is interesting to note
`t Dev -HP1 at 17 pm median pore size has nearly identical strength to Dev -1-1P2 at 20 pm
`median pore size, presumably duc to contributions of its pore size distribution. There is an
`expected strength gain from catalyst coating, especially for heavy washcoat loadings in the 4-
`way catalyst application.
`The material elastic moduli for the two new filters arc also directly related to their
`porosity and porc microstructures. With an 8% porosity increase from Dev -EC to both Dev -HP I
`and Dev -HP2, the elastic moduli have been respectively decreased by 54% and 58 %. A similar
`trend is observed from DuraTrapr CO to Dev -EC with a 57% elastic modulus reduction by
`increasing the porosity from 50% to 64 %. These results show significant lower elastic modulus
`or reduction in rigidity in the ceramic body for our new high porosity filters and hence provide a
`better capability to withstand thermal and mechanical stress for the filter application.
`
`Thermal Shock Parameter
`In diesel particulate filter applications, the filter needs to be regenerated periodically by
`burning soot in presence of oxygen to reduce pressure drop resulting from soot accumulation.
`This regeneration event generates an exotherm temperature and related thermal gradient.
`Depending on the resulting stress level, this could cause cracking of the filter. Similar rapid
`beating and cooling cycles also occur in gasoline automotive catalytic converter applications
`when, it is common to use the thermal shock parameter (shown below) to describe the tendency
`of the product to crack upon such thermal gradients:
`
`15
`
`BASF-2020.012
`
`

`
`7SP - ModulusOJRupture
`Elastic-Modulus CTE
`From this definition, the
`thermal shock capability of the filter depends upon the
`combination of mechanical strength, elastic modulus, and CTE. The lower the CTE and clastic
`modulus, the higher the thermal shock capability. As depicted in Table 1, both Dev -IiPI and
`Dev -11P2 show adequate thermal shock capability comparable to DuraTrapt CO and Dev -EC,
`despite higher CTE and lower mechanical strength. Dev -I{P2 shows an even better thermal
`shock parameter than both DuraTrapt CO and .Dev -EC filters. Clearly, this results from the
`extremely low elastic modulus for this filter. The thermal shock parameter for the Dev -HP I filter
`is lower than that of the reference filters DuraTrapt CO and Dev -EC due to its higher CTE.
`Further improvement in mechanical strength and CTE in composition and process development
`benefits the thermal shock capability of both Dev -HP 1 and Dcv -HP2.
`
`FILTER PERFORMANCE
`In DPF applications, the most
`important performance attributes are pressure drop,
`filtration efficiency, and filter durability. Since the filter durability is a function of service life,
`and also the two new high porosity filters are in early stage of product concept feasibility
`demonstration, the durability for these two new filters is not included in this study.
`Filter pressure drop performance was conducted in lab test device at ambient temperature
`with air and artificial soot (Printer -U from Degussa Corporation) [8]. All filters were 5.66" or
`0.143 m in diameter and 6" or 0.152 mlong. The filter was loaded with artificial soot by aerating
`the fine soot powder into a compressed sir stream. After a specific and controlled amount of
`artificial soot is loaded inside the filter, the filter is removed, weighed, installed on a cold flow
`pressure drop rig and tested for pressure drop as a function of flow rate. The filter is then
`incrementally loaded with greater amounts of soot and the overall process is iterated until
`reaching the desired soot loading level (generally up to 5g/1). Pressure drop is typically plotted as
`a function of soot loading for the highest measured flow rate (356mr/hr). From pressure drop vs.
`gas flow rate curves, the filter wall permeability at clean and soot -loaded conditions can be
`determined
`Filtration efficiency was also measured on the same soot -loading device. Filters 2" or
`0.05 min diameter by 6" or 0.152 m long were used in this measurement A microdiluter was
`coupled with a condensation particle counter. These instruments are used in conjunction to dilute
`the aerosol flow upstream and measure the particles present at upstream and downstream of e
`particulate filter (8]. The filter is weighed before and after the test. The filtration efficiency of the
`entire system is determined by comparing the upstream and downstream numbers of artificial
`soot particles while the filter is in a quasi stationary test operation. The ratio of soot particles that
`arc trapped in the filler over the total soot particules injected leads to the filtration efficiency of
`the system. For a DPF, it is desired to have such number equal to or greater than 90%.
`
`(
`
`Soot Loaded Pressure Drop
`Soot loaded pressure drop performance for two high porosity filters, Dev -HP1 and Dev -
`HP2, is shown in Figure 3 in direct comparison to DmaTrae CO and Dcv -EC filters (all of theme
`in the bare state). It is obvious that both new high porosity filters have substantially improved
`soot loaded pressure drop performance, about 44% reduction vs. Dev -EC and about 60%
`reduction vs. DuraTrap'n CO. This excellent pressure drop performance can be ascribed to the
`improvement of total porosity, pore characteristics and microstructures as discussed above,
`
`16
`
`BASF-2020.013
`
`

`
`As pare characteristics (total porosity, pore size and distribution, and pore connectivity)
`continue to evolve, it i
`interesting to observe that the deep bed filtration mechanism during
`initial soot loading (< 0 -5 g!1) becomes less pronounced. This is counter- intuitive, especially for
`increased median pore size of Dey -HP I and Dev -HP2, since enlarged pore size will increase the
`probability of soot penetrating into the wall. One possible explanation is that overall soot laden
`gas flow is well distributed through the wall porosity due to good pore connectivity and pore
`uniformity and hence leading to slower gas velocity. As a small amount of soot penetrates into
`the wall porosity, it will increase soot loaded wall flow resistance similar to that through surface
`soot cake layers. As a result, deep bed filtration becomes less pronounced.
`Whereas these filters have significantly different soot loaded pressured drop, one can also
`observe that the difference in clean pressure drop is in comparison relatively small. This can be
`explained by the fact that the difference in pressure drop across the wall between these different
`filters in
`the clean state is somehow masked by other factors also contributing to the total
`pressure drop (such as channel friction or entrance and exit gas compression/expansion).
`
`.
`
`10. _.__._........._._-----_.._
`
`........__._
`
`s)
`ai
`
`duraTrap CO
`
`0
`
`1
`
`2
`3
`Soot loading (guy
`Figure 3: Soot Loaded Pressure Drop for Cordierite Filters
`
`4
`
`Filtration Efficiency
`When both filter porosity and pore size are increased for better catalyst coatability and
`pressure drop performance, it is always necessary to verify that the filtration efficiency is not
`compromised. Figure 4 shows filtration efficiency as a function of time for both new high
`porosity filters vs. Dev -EC. An internal standard filter with 35 pm mean porc size was used as a
`reference to calibrate and qualify the measurement and test setup. As expected, the standard
`reference starts at 70-80% efficiency due to its initial large pore size of 35 pm. As soon as soot
`builds up, the filtration efficiency increases since soot shrinks porc opening. It is obvious that the
`two new high porosity filters along with Dev -EC filter show excellent filtration efficiency
`greater than 97 %.
`
`17
`
`BASF-2020.014
`
`

`
`rt-nwHP1 (97%)
`-0-. WaHP2 (98%)
`_g_ Rot (92%1
`
`20
`
`25
`
`20
`
`0
`
`5
`
`10
`
`15
`nn. (mina)
`Figure 4. Filtration Efficiency for Different Families of Cordierite Fitt
`
`Filter Wall Permeability
`Front the pressure drop measurement, wall permeability can be detennined for both clean
`and soot -loaded conditions. The pressure drop as a function of wall permeability is plotted in
`Figure 5. As discussed above, clean pressure drop for the filters is dominated by other factors
`than the filter wall permeability after the permeability reaches a threshold value of about I x 10'
`12 m2. DuraTraps CO is right at that threshold for clean wall permeability. Therefore although
`Dev -EC and Dev -1-1P2 are two - and six -times better in wall permeability, the overall clean
`pressure drop among these filters is not so different (as shown in Figure 4). On the contrary, as
`the wall permeability is significantly reduced by soot as shown in Figure 5 (data point from same
`set of filters for soot loaded and clean values), the pressure drop then becomes more sensitive to
`the difference in Wall permeability among these different filters.
`Wall permeability is one of the most relevant parameters used to predict pressure drop
`after catalyst coating. With about a factor of three improvement in wall permeability of Dev -HP2
`compared to Dev -EC as well as additional improvement in porosity and pore connectivity, it is
`highly anticipated that these two new high porosity filters should be able to have high catalyst
`loadings with coated pressure drops comparable to DuraTrap® CO.
`
`CONCLUSION
`The new high porosity filters, Dev -HP1 and more particularly Dcv -HP2, show a
`significant improvement over the existing eordierite product, DuraTrapa CO, and also the first
`generation of high porosity cordierite, Dev -EC. These filters have a total porosity of 72% with
`narrow pore size distribution and mean pore sizes of 17 pm (Dev -HP1) and 20 gm (Dev -HP2).
`These basic product attributes are highly desirable for high washcoat loading necessitated for the
`4-way application while retaining low soot loaded pressure drop as well as a reasonably low
`CIE and an acceptable mechanical strength.
`
`ACKNOWLEDGMENTS
`We wish to thank our colleagues Adam Collier, Benjamin Stevens and Stanley Solsky for
`their continued help on this project. We would also like to recognize the help from our pressure
`drop testing team, our plugging team as well as our internal characterization team.
`
`18
`
`BAS