SAE TECHNICAL
`PAPER ssmss
`
`2oo2-o1-o322
`
`Sic and Cordierite Diesel Particulate Filters
`
`Designed for Low Pressure Drop and Catalyzed,
`Uncatalyzed Systems
`
`S. Hashimoto, Y. Illiyalri, T. Hamanaka, R. Matsubara, T. I-larada and S. Miwa
`NGK INSULATORS. LTD.
`
`Reprinted From: Diesel Exhaust Emission control 2002:
`Diesel Particulate Filters
`
`(sP—1s73)
`
`Q‘ g The Engineer!!! 50819?!
`
`GI-E;‘i‘fia‘§’:':£'}’J’§‘.l?é’.?’£‘IiL’,
`
`INTERNATIONAL
`
`SAE 2002 World Congress
`
`Detroit. Michigan
`
`March4-7,2002
`
`
`
`400 Commonwealth Drive, Warrendale. PA 15096-0001 U.S.A.
`
`Tel: (724) 776-4841 Fax: (‘[24] 7?6-5760
`
`1
`
`JM 1007
`
`1
`
`JM 1007
`
`

`
`The appearance of this ISSN code at tl1e bottom of this page indicates SAE's consent that copies of the
`paper may be made for personal or internal use of specific clients. This consent is given on the condition.
`however. that the copier pay a per article copy fee through the Copyright Clearance center. Inc. Operations
`Center, 222 Rosewood Drive. Danvers. MA 01923 lor copying beyond that pennitted by Sections 107 or
`108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for
`general distribution. for advertising or promotional purposes, for creating new collective works. or for
`resale.
`
`Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.
`
`To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
`other works. contact the SAE Publications Group.
`
`
`
`GLOBAL IIOBIIJTY DATABASE
`
`AH SJIE papers, standards, and‘ selected
`hooks are abstracted and indexed tn the
`Globe.‘ liability Database
`
`No part of this publication may lie reproduced in any form. in an electronic retrieval system or otherwise. without the prior wrlten
`permission or the publisher.
`
`ISSN 0148-7191
`Copyright 2002 Society of Automotive Engineers. Inc.
`
`Positions and opinions advanced In this paper are those of the sulhor{s] and not necessarily those of SAE. The author is solely
`responsible for the content oi the paper. A process is available by which discussions will be printed with the paper if it is published in
`SAE Transactions. For permission to publish this paper in full or in part. contact the SAE Publications Group.
`
`Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
`word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board. SAE.
`
`Printed in USA
`
`2
`
`

`
`Sic and Cordierite Diesel Particulate Filters Designed for Low
`Pressure Drop and Catalyzed, Uncatalyzed Systems
`
`8. I-iashlmoto. Y. Miyalrl, T. I-lamanaira, R. Matsubara. T. Harada and S. Miwa
`NGK INSULATORS. LTD.
`
`2002-01 -0322
`
`Copyi'ight02002.°ioc:ietyolAutcmotiveEngii1eer|.Inc.
`
`ABSTRACT
`
`DPFs (Diesel Particulate Filters) have been a primary
`technology utilized to purify diesel PM emissions. One of
`the major challenges of the DPF is to reduce pressure
`drop caused by PM and ash accurntlation.
`
`This paper reports on the definition and investigative
`results of several major parameters wh|cl1 determine the
`pressure-drop of Cordierite and Sic (Silicon Carbide)
`DPF’s. After which. the successful material development
`for low pressure-drop cordlerlte and SIC DPF's are
`presented.
`
`INTRODUCTION
`
`Diesel engines have predominance in heavy-duty trucks
`and buses. More recently, sales of diesel passenger
`vehicles have greatly increased in Europe due to the
`additional power and fuel efficiency offered by diesels.
`Diesel engines are also considered to be one of the
`most realistic solutions for decreasing CO2 emissions.
`
`A main concern of diesel vehicles is the high PM
`(particulate matter) emission. wi1icl1 has been identified
`as a potential carcinogenic ‘fighter PM emission
`regulations will be introduced in Europe. United States
`and Japan on passenger cars and trucks over the next 7
`years. DPF (diesel particulate filter) has been the
`primary technology considered in meeting these tight
`PM emission Iirnits. The Wall—l|ow Cordierite DPF's has
`been consistently used for folk-lit and mining equipment
`applications over the years and are now being applied to
`truck
`and
`bus
`retrofit
`applications.
`Furtliemiore.
`S1c(SlIicon Carbide) DPF's were introduced on a
`production. European vehicle In 2000.
`In this case. a
`Cerium fuel additive is used to reduce soot combustion
`temperature and forced regeneration is conducted
`periodically [1] [2]. The Ceriumlsic-DPF system purifies
`PM emissions by more than 90%. Japanese Truck
`manufacturers are also seriously developing DPF
`
`systems to meet new short-term emission regulation
`etfectlveinoctober 2003.
`
`The addition of the DPF exhaust system increases back
`engine pressure. which reduces engine power and
`increases fuel consumption. DPF accumtlates not only
`PM but also ash from the engine oil and fuel. further
`increasing pressure-drop. Also. the available space for
`packing the DPF in the under-floor or toe-board position
`for passenger vehicles
`is
`limited. For heavy-duly
`vehicles
`the packaging envelope is also limited.
`Typically in this case. the mutller is modilied to package
`the DPF. The limited DPF-packaging envelopes make it
`difllcult to provide sulficient DPF cross sectional area or
`suflicient filtration area. Based on these conditions.
`optimizing
`pressure-drop
`and
`trapping
`efflciency
`simultaneously is a notable challenge for DPF systems.
`
`A catalyzed DPF system has been proposed as a future
`PM emission [3] control
`technology.
`in the case of
`catalyzed DPF system. the catalyst in the DPF further
`increases
`tlow restriction.
`In
`addition
`to
`the
`aforementioned conditions. a DPF with low pressure-
`drop is very important
`from the aspect of engine
`performance.
`
`EQUATIONS OF DPF PRESSURE LOSSES
`
`GENERAL DISCUSSION -
`
`Asthefirststepofthis study,ti1eequationstcdefine
`DPF pressure-drop were established. Pressure drop of a
`DPF consists of five different factors. 111ese factors are
`shown in Figure 1. The first is contraction and expansion
`losses at the inlet and outlet faces of the DPF created by
`the plugged-cells. The second losses are the inlet and
`outlet channel frictional losses. The third and fourth are
`tiow restriction when gas tiow though porous wall of DPF
`and the soot layer, respectively. Pressure losses at the
`inlet and outlet cones of
`the converter are also
`considered.
`
`3
`
`

`
`PresstreLou.kPa
`
`D
`
`0.2
`
`0.4
`1-OFA
`
`0.6
`
`0.8
`
`Figure 3. Effect of OFA on Plug loss
`
`1.
`
`Several equations of DPF presstre losses have already
`been proposed [4}. In this study, the equations of DPF
`pressure drop were developed tor each of the five-
`factors experimentally and Independently, and these
`equations were used to determine the DPF‘s optimum
`cell structure to minimize pressure drop.
`
` S
`
`3.4
`
`5
`
`1. Contraction and expansion Loss
`2. channel flow loss
`3. Wall flow loss without soot
`4. Wall flow loss with soot
`5. Duct loss
`
`figure 1. Factors of DPF Pressure Losses
`
`Thetotal pressureiossofeDPF (AP)is expressedasa
`summation of these factors:
`the plug loss: AF-‘m.
`the
`channel loss: APc. the restriction of clean well: APw,¢......
`the restriction through a soot layer: APw,.... and the cone
`losses: APd.
`
`AP=APm-I-APc+ aPw,.,....+ APw_,,,,-I-APd
`
`CONTRAC11ON AND EXPANSION LOSSES -
`
`Contraction and expansion losses are generated at
`the plugged cells (on the inlet and outlet faces). To
`measure these losses. DPF's wilt various cell structures
`were sloed near the inlet and the outlet and pressure
`drops through these sliced test pieces were measured
`under various flow rates. Measured results are shown in
`the Figure 2. The Figure 3 shows the relationship
`between the ratio of open frontal area (OFA) and the
`pressure drop under the fixed flow rate of T Nm°imln.
`Thegeneralforrnfortltispressurelosswaseniployed
`with an empirical pressure coetticlent § .
`
`I-\F'm=-‘§'.tW‘
`
`--(1)
`
`CHANhEL FRICTION LOSSES —
`
`The elndexwas setto 2.2063. sotheflowralse effect
`agreed with
`experimental
`results. The
`pressure
`coeliicient 5 was expressed as a function of the ratio of
`open frontal area (OFA). The following equation
`matched the experimental data of
`the various cell
`stmolures.
`
`i: = C1{1-OFA)”'"
`c, = 0.02759
`
`— (2)
`
`The Reynolds number of the channel flow in the DPF is
`in the range of laminar flow even at the rnaxlrnum engine
`flow rate oondition. The pressure loss of laminar flow
`through a channel is generally in proportion to viscosity.
`length and velocity, and inversely in proportion to a
`square of
`the hydraulic diameter. Pressure losses
`without the inlet and outlet plugs were measured for
`various cell structures. The measured resuits are shown
`in Figure 4. Here the channel friction loss was expressed
`by the following equation with the empirical index m and
`
`4
`
`

`
`coefiioient C2. The parameter. m, which is expected
`theoretically to be a value near 1.0 was setto1.o184 by
`curve-fitting.
`
`APc = c,»/" L p. I (OFA) r(oI-I)‘
`C2 = 0.012.
`m = 1.0814
`
`-— (3)
`
`u: wall flow velocity tmls)
`p: viscosity (Pa s )
`uo: standard velocity (ms)
`[191 viscosity under standard condition (Pa s)
`f(t,)'. Increase In the wall tlow pressure loss
`with soot loading under on and [Jo
`
`
`
`Pannellally 3
`
`(Air Velocity) {Sample Thlottnou) (Vii ol'Nr_}
`(Pressure Drop}
`
`Figure 5. Wall Permeability Measurement
`
`--The u is oonstanto (u=uu)
`--Compensated under Room Temperature
`
`3 §
`
`5
`n.
`g
`E
`£
`
`- -
`LI'rul'r-'I2Il- -
`flint-~ ittlnln
`Itfi Oanflhn
`TIII|Ii'dulI'l'|\.'l'- fl.IdI§- '
`
`_Z7'1Trrlt!10tlcpst
`
`Flow Ratetsmarmin)
`
`Figure 4. Channel Loss Measurement
`
`WALL AND SOOT LAYER PRESSURE LOSS-
`
`lt is assumed that Darcy's law is applicable to these
`losses. The permeability of a clean well can be
`measured and determined for each DPF material as the
`gradient of wall flow pressure plot as shown in the
`Figure 5. However. the permeability of the wall with soot
`loading will change with the soot loading-rate. Typical
`relationship between the soot loading-rate and the
`pressure loss is shown in the Figure 6. During the llrst
`stage of soot loading. the characteristics of the pressure
`drop with soot loading has a steep slope. shown as "I" In
`the fiQl.I'B. This is due to the soot being trapped inside
`the wall. as described later in this paper. The pressure
`drop curve has a transition-curve ‘ll’. then moves to the
`stable slope at a given soot-load, indicated as ''III' in
`Figure 6.
`
`Pressure drop of the conplete-DPF during soot
`loading was measured. and the resulting increase in how
`resistance was detennined as a function of the soot
`loading-rate for each material. Finally.
`the pressure
`losses through the wall and the soot layer are expressed
`as follows:
`
`APw=APw_....,+lB w_...,.
`=(te'ko)uu+f(t.)ulu'tuauoJ
`
`-—(4)
`
`1! (U9 pq)- 6.3ED6
`it: (well + soot} apparent permeability (m’)
`t: (well + soot) thickness = in + t, (m)
`t,: apparent soot layer thicltness (m)
`k.,: pemteahllity of clean wall (m’}
`in: wall thickness (m)
`
`
`
`
`Initial Pressure Drop
`tlniko) '
`‘-'o'Fo
`
`I
`
`II
`
`fits)!
`experimentally
`decided
`
`Soot Amount per Unit Flltraion Area
`
`Figure 6. Wall and Soot Layer Loss
`
`INLET AND OUTLET CONE LOSSES —
`
`The general form of oontraction and expansion was
`employed:
`
`APd =2pv.’(1-u.’ro‘)’
`V1! velocity in tube (mls)
`d.: Diameter of tube (In)
`D: diameter of DPF (m)
`
`-5)
`
`It has been continued that the estimated pressure drop
`by these equations matched measured values of DPF
`pressure drop under various temperatures. new rates.
`DPF dimensions. and soot loading rates.
`
`5
`
`

`
`INVESTIGATION OF THE MAIN CONTRIBUTOR
`FOR PRESSURE DROP
`
`PRESSURE DROP WITHOUT SOOT LOADING
`
`Thecontributionrateofeachparameteronbolhwith—
`and without-soot-loading was
`evaluated with
`the
`pressure-drop equation. This study was carried out in
`terms of airliow rate at room temperature. The results of
`this study are shown in the Figure 7.
`It was also
`concluded that the main contributors for pressure-drop
`without soot loading were the expansion and contraction
`loss due to the cell-plugs and ohannei flow loss.
`5
`
`PressureDrop.
`
`kPa
`
`Figure 7. The effects of pressure-drop parameters
`on pressure-drop without soot loading
`
`Tl1is study also concludes the percentage ¢flfIII'ibUtiOl'|
`ratioofwailllowlosswasrelativ-elyminor.Accordingto
`equation (4). the permeability of the DPF material is the
`key factor for wall
`flow
`The relationship
`between penneability and pressure-drop without soot
`loading condition was examined at a 9 Nm°lmln of gas
`flow rate. These results are shown in I-‘rgure B. Cordierlte
`DPF's with several different levels of permeability were
`evaluated.
`
`§
`-
`
`6
`
`-I.ti6"Dx8'L
`
`l2rnl.b'3D0cpsi
`
`0
`
`2:10“
`
`mo"
`Penneabillly. m”
`
`6x10"
`
`axle“
`
`Figure 8.
`
`Influence of Permeability on pressure
`drop without soot loading
`
`The penneabllity of the DPF material has a relationship
`with pressure drop characteristics when the penneability
`of the material is less than 4:: 10"’ (m‘}. It means that
`the permeability (accordingly, wall flow less) has less
`impact on pressure drop without soot loading condition.
`if the material has sufficient permeability. This explains
`the reason why the contribution of the wall flow loss
`(without soot loading) was small.
`
`PRESSURE DROP WITH SOOT LOADING
`
`The same pressure drop study as shown in Figure 8 was
`conducted at a soot-loading condition of 4 g.i‘L. Figure 9
`shows the evaluation results of the contribution of each
`pressure-drop parameter with the soot-loading condition.
`Contrary to the “no soot loading“ condition. the wall flow
`loss was the key contributor of pressure-drop. while the
`plug and the channel
`losses had less impact. Also.
`Figure 9 shows the pressure-drop with soot- loading is
`much higher than that without soot loading. Accordingly.
`reducing wall flow loss is the key to a low pressure-drop
`DPF.
`
`
`
`All’ Flow Rate. Nm'lrrin
`
`Figure 9. The effects at‘ pressure-drop parameters
`on pressure-drop with soot loading
`
`PRESSURE DROP MEASUREMENT RESULTS
`
`111s pressure-drop measurement of several Cordlerlte
`DPF's with dilferent cell structures were conducted
`under without- and with-soot-loading conditions. Table 1
`shows the material properties and cell structures used
`for this study.
`
`Figure 10 outlines the pressure-drop evaluation test
`results without soot loading. The DPF sample size used
`for this study was Q 5.66‘ {143.8mm) and 6" (152.4mm)
`in length. Pressure-drop was measured in terms of gas
`flow rate at room temperature.
`
`The pressure-drop of material A with 1?mil I 100cpsi
`was approximately 10% higher than material B with the
`same cell structure. Also. the higherthe cell density. the
`higher the pressure-drop. These tendencies were the
`same for all the air-flow rates used for this study. As the
`
`6
`
`

`
`initial pressure-drop is primarily intluenced by channel
`flow loss. small hydraulic diameter due to high cell
`density increases chamel flow losses.
`
`Table 1. Properties of Cordierlte DPF
`for pressure drop evaluation
`
`DPFSW RE
`l'JHC-
`DHC-
`l'JHC-E
`P°~siv<*> 3553
`Mean Pore Size
`
`
`
`PresstmeDrop.kPa
`
`(urn)
`DPF Size : 5.66"Dx6"L
`
`Air Temp. : Room Temp.
`
`o
`
`2
`
`5
`4
`Air Flow Rate, Nmalmh.
`
`a
`
`10
`
`Figure 10.
`
`initial pressure drop of Cordlerite DPF
`
`Alter measuring the initial pressure drop. the pressure-
`drop characteristics for the soot-loaded condition were
`evaluated. 1'he same samples were used for this study.
`
`performance with-soot was
`drop
`pressure
`The
`siyiilicantly different than the initial. no-soot pressure-
`drop (shown in I-‘tgure11).ln the case ofmaterial-A. the
`
`initial pressure-drop without soot at 2.27Nm3imin. of
`airflow rate was approximately 0.4 kPa. On the other
`hand. the pressure-drop with 5g of soot loading was
`15lrPa. approximately 3?.5 times higher.
`
`The difference between material-A and -B increased
`under the soot-loading condition. Specifically. material-A
`had roughly 30% higher pressure-drop than the material-
`B above 2g of amount of soot loading. This is because
`the wall flow loss has less impact on the Initial pressure-
`dropbi.rtisal:eyccntI'il::t.Itoruvithsootloading.assho\nrn
`in Figure 7 and 9. In order to reduce the pressure-drop
`of the DPF. to analyze the material parameters on well
`flow loss is important.
`
`5.6B“D:t6'L
`
`NC-IK Soot Beneraor
`Eqtipmont :
`2(1) deg.C
`Gas Term. :
`Gas Flow Rae : 2.27 N'I'I3‘rlin
`DPFSZB:
`
`0
`
`2
`
`4
`
`6
`
`8
`
`10
`
`Amount ofsoot Loadlrg.g
`
`Figure 11. Pressure-drop of Cordierite DPF with
`soot loading condition
`
`Also Figure 11 exhibits cell density dependency on
`pressure drop under
`soot-loading conditions was
`contrary to the without soot loading condition: the higher
`the cell density. the lower the pressure-drop. In the case
`of initial pressure-drop.
`the high cell density reduces
`hydraulic diameter.
`resulting in high pressure~drop.
`because the channel llow loss is the key contributor. On
`the other hand. pressure-drop with soot loading was
`mainly influenced by the wall flow loss. As the filtration
`ereaincreaseswith cel density. the iowerthewall flow
`loss. This is the reason the high cell density DPF has
`low pressure-drop under soot loading conditions.
`
`Based on these results. both material optimization and
`cell structure optimization are important to minimize
`pressure-drop of DPF.
`
`7
`
`

`
`DESIGN OPTIMILKTION
`
`_
`
`DESIGN PARAMETERS ON BACK-PRESSURE
`
`These two effects seem compensate and wall thickness.
`which also has impact on wall flow loss. dominantly
`influences total DPF pressure-drop.
`
`According to the pressure drop equations mentioned
`above. the followings can be listed up as the design
`parameters for DPF pressure losses:
`
`1) DPF Shape
`' Contour! Flow surface area
`' DPF Length
`
`2) Cell Structure
`' Wall thickness
`* Cell Density
`
`3) DPF Material characteristics
`* Pore Size
`* Mean pore size! Pore size dislrbutton
`
`‘the available packaging envelope of the particular car
`and truck manufacturers inttuences the DPF shape
`{Contour and length}. Therefore.
`the development
`focused on controllable parameters.
`such as cell
`structure and DPF material characteristics. to reduce
`DPF pressure-drop.
`
`CELL STRUCTURE OPTlMlfl\T|0N
`
`As shown in Figure 9, wall flow less is a major influence
`on pressure-drop with-soot-loacing. Greater filtration
`area by increasing cell density is an effective approach,
`as shown in Figure 11.
`
`The impact of cell structure on pressure-drop with soot
`loading was calculated by using the pressure-drop
`equations. The conditions for this study are listed in
`Table - 2 and the calctlated results are shown in Figure
`12 as isobars.
`
`Table 2. Conditions of pressure-drop calculation
`in terms of cell structure effect
`
`Material:
`Porosity:
`Mean Pore Size:
`DPF Size:
`
`Cordierite
`59 ‘It.
`25 pm
`1: 143.amm x 152.4mm
`(¢ 5.66" x 6"}
`3gr'Ll;|'otaI 7.59}
`Amountofsootz
`2.27m I min
`Gas flow rate:
`Gas Temperature: 200 degree C.
`
`when the cell density is less than zoocpsi. the DPF
`pressure-drop with 3g!L of soot
`loading depended
`mainly on cell density. This is because higher cell
`densityhasalarger filtration area.
`
`On the other hand. when the cel density is greater than
`300cpsi. DPF pressure drop is primarily dependent on
`wall thickness and much less on cell density. Above 300
`cpei.
`the cell structure increases filtration efficiency
`resulting in low wal ttow drop. however channel friction
`loss becomes high due to small hydraulic diameter.
`
`Based on these results and considering the potential
`soot plugging of a higher cell density DPF. 3t.‘rDcpsi was
`selected as the optimum cell structure for the filter in
`order to have low prusure-drop.
`
`thiclmess is the other design consideration to
`Wall
`reduce pressure-drop. Thinner wal reduces wall flow
`loss, resulting in low presstre-drop. But. thinner wall
`also reduces thennal mass and mechanical strength of
`the filter. Lower thermal mass increases the DPF
`temperature during forced and uncontrolled soot
`regeneration. From these points of view.
`it
`is not
`advantageous to significantly reduce wall thickness.
`
`Z Ilmmm
`
`,_ ,... _llIlIlIIt\V.K\‘t
`E... IIIIIIIIM
`§...
`
`it 300
`
`
`
`0 200
`
`150
`
`100
`
`.
`
`_
`
`_
`
`__
`
`3
`
`10 W1§bThir1:l:ness1?nil
`
`15
`
`20
`
`Figure 12. Pressure-drop of Cordierite DPF
`withsootloading condition
`
`DPF MATERIAL OPTIMIZATION
`
`KEY MATERIAL PARAMETERS ON BACK-PRESSURE
`
`In order to identify the key DPF material characteristics
`which affect pressure drop under soot loading conditions.
`21 different Cordlerlte and 4 different Sic DPF's were
`prepared for the evaluation.
`
`Basilly. porosity and mean pore size were determined
`to be the key material characteristics of DPF materials.
`To
`investigate
`the
`influences
`of
`pore
`sizes
`characteristics fi.rrther. several parameters were used for
`this study besides porosity and mean pore size.
`
`In ordertoinvestigatethe efiectofsmall pores.the
`parameter of "-10pm porosity" was established. This
`parameter is defined as the integrated porosity of pores
`less than 10pm diameter. The parameter of "10-70pm
`porosity" and "+70prn porosity‘ are defined as the
`
`8
`
`

`
`"it
`
`the results of pressure-drop and filtration efficiency, the
`optimum pore size is between 10pm and 7U|.tt'l't.
`
`MECHANISMS OF PRESSURE DROP OF DPF
`
`Further evaluation was conducted to confirm the soot
`trapping mechanism. The cross sect.iona| area of the
`DPF wall was observed at 0.1. 0.3 and 0.5 git. levels of
`soot loading. As shown in Figure 8. the characteristics of
`the pressure-drop vs. soot loading are: a steep slope
`under the low soot loading (<o.1 glL) condition (region
`'f'): than transition to a moderate slope between a soot
`loading of 0.3 g.tL and 0.5 glL (region ‘tt') : then. above
`0.5 gl‘L. the pressure-drop has a linear dependency on
`amount of soot loading (as shown “IIi" in the Figure 6).
`
`Photograph1showsil1esootdistribution.111etop
`photographs are observation results on the surface of
`DPF watt. Letter “A” shows the surface of the inlet
`channel, letter ‘B’ shows the DPF wall. and letter ‘C’
`shows the surface of the outlet channel. The surfaces of
`the inlet channel had dark or black colors. which is the
`trapped soot.
`
`The lower photographs are the cross sectional area of
`the walls. Arrows in the photographs show the direction
`of gas llow. Photograph 1 (1) shows the cross sectional
`view of DPF material under 0.1 g!L of soot loading
`condition. Pores inside the wall trapped soot. Near the
`entranoeottheg ltow.thedensityoftrapped sootis
`high compared with near exit side. In Photograph 1 (2).
`this phenomenon becomes clear. Finally in Photograph
`3. the soot layer developed on the DPF well on the inlet
`channel. 111s top photograph in Photograph 1 (3) also
`shows the uniform black color. which is the developed
`soot layer.
`
`Based on these results. as the pores inside the DPF wall
`trap the soot. the pressure-drop rapidly increases. and
`the rate ot increasing pressure-drop diminishes atter the
`sootiayerls developed. Thisrateseemstodepend on
`the porosity of soot layer. It appears that porosity ofthe
`soot layer is relatively high.
`
`These observations did not prove it the smaller pores
`trap soot {less than 10pm in diameter]. due to difliculty
`to identify unplugged small pores. However, according to
`these observation results.
`it
`is clear that pore-size
`characteristics are very important for pressure drop with
`soot loading. In addition to the influence of pores. soot-
`slze influence and pore connectivity are items for further
`evaluations.
`
`integrated porosity from 10pm to 70pm and more than
`70pm respectively. Figure 13 is a schernatlc illustration
`of these parameters. The relationship between these
`parameters and the pressure drop of DPF's with 1.5 gtL
`of soot loading were investigated in order to assess the
`impact ofthe rnid— and large size pores.
`
`10-T0prnPoro.sity:
`lntegraleporoeity
`flIIm10urntoT0|.rm
`
`olporediarnetar
`
`Figure 13.
`
`Schematic iiluslralion of parameters’
`definition for the study of pressure
`drop mechanisms
`
`Figure 14 shows that generally the higher the porosity.
`the lower the pressure-drop, with the exception of an
`average mean pore size below 10pm. The total porosity
`is the key contributor for pressure drop under soot
`loading conditions. Also, DPF materials with a high
`concentration of mean pore size below 10].t|'t'| have low
`porosity. Thlsseemslobeoneofthe reasonswhythe
`relationship between the porosity of below 10|.tl'I'I of
`poresand pressure-dropshowedtheopposite behavior.
`
`From these results. it was concluded that a material with
`high porosity and with high pore sizes is desirable to
`have low pressure drop performance.
`
`KEY MATERIAL PARAMETERS ON FILTRATION
`EFFICIENCY
`
`Filtration efficiency is an equaly key function ofthe DPF.
`Thelmpaotoftheseporesizeand porosiiyonlhe
`filtration efliciencywere also evaluated. Figure 15 shows
`these results.
`In this case.
`the relationship between
`tilbation efliciency without soot loading condition was
`investigated.
`
`Figure 15 outlines very interesting results. In this case.
`theporesizeofgreaterlhan 70p.mI'|fl.B artadverseeffect
`on filtration eiliciency. This tendency is not as obvious
`forthe10ru11to70p.mporerange.0n thecontrary, the
`porosity of less than 10pm diameter pores had a positive
`relationship with filtration etficiency. Soot can easily pass
`through large pores {more than 70pm) and can be easily
`trapped by the small pores (with less than 10pm
`diameter) at the initial soot ioaotng condition. Based on
`
`9
`
`

`
`1) The porosity ofpores below 10pm ofpores
`
`4] The totat porosity
`
`'-«I
`
`G
`
`IO-5U!
`
`(40
`
`
`
`0atuna:cannon
`
`'-I
`
`03
`
`
`
`Pressuredrop(kPa)no45-UI
`
`2
`
`20
`10
`-10pm por'osity(%)
`
`0)D
`
`30
`
`40
`
`60
`50
`Total porosity ($6)
`
`70
`
`80
`
`2)Thaporosttyofporesfrom10to70pmofpores
`
`5) The permeability
`
`
`
`7
`
`'56
`
`E 3
`
`5‘U
`
`E4
`
`n, 3
`
`2
`
`0
`
`60
`40
`20
`10-Toprn porosiIy(%)
`
`B0
`
`0
`
`2
`
`4
`
`--H 6
`
`8
`
`Par 2 MP3‘ (x1o"° m’)
`
`3) The porosity of pores above ‘(Burn of pores
`
`
`
`Pressuredrop(kPa)inor03"-I
`
`00
`
`I0
`
`O
`
`20
`10
`-I-70pm porosity(%)
`
`8
`
`Figure 14. The correlation study batman pore
`characteristics and pressure drop
`
`10
`
`10
`
`

`
`1) The porosity of pores below 10pm of pores
`
`4) The total porosity
`
`
`
`Fl|‘H'8HOll8fficI8l'ICy(96)
`
`3
`
`ct
`
`,,
`
`B83§
`
`
`Filtrationefliclancy(96) 3
`
`C
`
`0
`
`20
`10
`-10iim porosflv(%)
`
`30
`
`30
`
`40
`
`so
`50
`Total porosity (93)
`
`70
`
`so
`
`2) The porosity of pores from 10 to 70pm of pores
`
`5) The Perrneabmtv
`
`100
`
`
`
`
`
`Fltrationeffictency(95)__888
`
`
`
`
`
`Fiitrationefioiency(96)
`
`3388
`
`60
`40
`20
`10-70pm poros|ty(%)
`
`8
`
`O
`
`I'D
`
`9
`
`O!
`
`-h
`
`O
`
`-.8. O
`
`For 1: MP3’ (x1O'1° in‘)
`
`3) 1'he porosity of pores above 100u.m of pores
`
`383§
`
`
`Filtrationefficiency(Wu)
`
`0B
`
`O
`
`20
`10
`+70,"-n ppmsltypx.)
`
`39
`
`Figure 15. The oorrelation study between pore
`characteristics and filtration efflciency
`
`11
`
`11
`
`

`
`1) lJ.1g!L ofsoct loading
`
`[ 0
`
`APPROACHES FOR LOW PRESSURE DROP
`DPF
`
`LOW BACK-PRESSURE TYPE CDRDIERITE DPF
`MATERIAL
`
`the most
`Based on the aforementioned findings,
`etfeotive material
`improvement
`strategy to reduce
`pressure drop is to increase the material porosity and
`maintain the pore size between 10p.l'I'I to ‘mum. in this
`section. new Cordlerite and SIC material developments
`are described.
`
`NGK Insulators. LTD has newty developed low pressure-
`drop type Cordierite DPF's. Material development has
`been completed which Increases the porosity from the
`current mass production level (53%) to 85%.
`
`The current production material. DHC-553. has a
`porosity and mean pore size of 53% and15;un,
`respectively. From this material. porosity and mean pore
`were increased as the initial step. This material. DHC-
`611. has a 59% porosity and 25pm mean pore size.
`
`approaches were
`different
`two
`From DHC-611.
`conducted. The first was to increase porosity and
`maintain the 20pm mean pore size. This material is Cd-1
`with 65% porosity. The other di-ectlon was to increase
`the mean pore size to 35pm. This material. Cd-2. has
`59% of porosity. the same as DHC-811.
`
`these
`The Table 3 summaizes the properties of
`materials.
`and
`Figure
`16
`shows
`the material
`development approach schematically.
`
`Table 3. Properties of Low Pressure-drop
`type Cordlerite DPF
`
`DPFNE“!
`
`W’
`M@
`
`cpsi
`
`Deli
`Density
`
`100 300
`cpsi
`cpsl
`
`300
`cpsi
`
`300
`cpsi
`
`300
`
`Photograph 1.
`
`cross sectional observation results
`of DPF well
`
`The pressure drop perfonnance of these materials was
`evaluated. and is graphed in Figure 17. The measuring
`conditions areshowninTabie4.The hlgherthe porosity.
`
`12
`
`12
`
`

`
`it
`
` 0
`
`20
`
`15
`10
`5
`Amomt of Soot Loading. 9
`
`Filtration efiiciency evaluation results
`Figure 18.
`of low pressure-drop type Cordierite DPFs
`
`these Cordierite
`size distributions of
`The pore
`materials was determined and are plotted in Figure 19.
`The pore diameter plot of all the materials. except Cd-
`2. exhibited a statisticaly normal distribution. The
`peaks of these curves are between 20pm to 30pm.
`On the other hand. Cd-2 had unique pore distribution
`with two peaks. One is around 20pm and the other is
`100nm.
`
`the lower the pressure-drop. The pressure-drop of Cd-1
`with 12ml! 1 300cpsi are approximately 80% lower than
`DHc—55B with 17ml I 100 cpsi. and 25% lower than
`DHC-558 with 12rnil I 300t:psi.
`
`
`
`Mean Pore Size (um)
`
`Figure 18
`
`Schematic illustration of cordierite DPF
`material development approaches
`
`
`
`
`-infirm
`
`jam
`I.'lrlH|$_l—
`?"‘1n\\—
`Iain-a*.~.-vaugj
`
`T1
` 0
`
`
`
`
`jzqrnmgj
`
`1
`
`
`
`
`
`1000
`
`15
`10
`5
`Amount of Soot Loading, 9
`
`20
`
`100
`10
`Average Pore Diameter. pm
`
`Figure 1?. Pressure-drop evaluation results of low
`pressure-drop type Cordierlte DPFs
`
`Table 4. Measuring condition of pressure drop
`
`DPF Size:
`
`iii 143.8rnrn X 152.4n1m
`(ii 5.6?‘ x 6"}
`2.27m I min
`Gas flow rate:
`Gas Temperature: 200 degree C
`Generating rate of Soot: 5.5g i hour
`
`The filtration efficiency of these materials are also
`evaluated {Figure 18). The Cordierite materials with less
`than a 25u.rnrnean pore sizehad rnorethan90% of
`tiltration eiticiency with soot loading. while car: with
`35umofmeanporesizehadlowerfiltrationeiflolenoy.
`
`Figure 19.
`
`Pore size distribution of low pressure-
`drop type Cordierite DPFs
`
`According to the Figure 14 and 15. the optimum pore
`size is frornt 0pm to 70pm. DHC-611 and Cd-1 have
`sharp pore size distribution with increasing porosity.
`Almost all pores of these materials are within this
`range. This explains the superior pressure-drop and
`filtration efficiency of these materials. Also. Figure 14
`concluded that DPF material with a porosity of mom
`or more had low filtration effioiency. This is because
`particulate matter can easily pass through large pores.
`This explains why Cd-2 exhibited lower
`filtration
`efficiency.
`
`13
`
`13
`
`

`
`I L
`
`The pressure drop performance of Sic-1 and Sic-2 were
`measured and compared with Sic-3. The measuring
`conditions are the same as shown in Table 4.
`
`figure 20 graphs the evaluation results. Sic-2 with 12ml
`1 300cpsi has roughly 18% lower pressure-drop than
`Sic-1. Compared with the rs-crystallized SIC-DPF with
`13miI of wall thickness and ‘lB0cpsi of cell density. SiCa2
`with ‘l2mil and 300cpsi showed roughly 30% Imver back-
`pressure.
`
` O
`
`10
`
`15
`
`5
`
`Amount of Soot Loading {9}
`
`Figure 20. Pressure-drop evaluation results of low
`pressure-drop type SIC DPF
`
`Based on these resuls. Sic-2 and Cd-1 are excellent
`candidates for a low pressure—drop type DPF.
`
`LOW PRESSURE TYPE DPF FOR CATALYZED
`SYSTEM
`
`Recently. catalyzed DPF systems have been proposed.
`One example was presented in reference [3]. In the case
`of catalyzed DPF system. the wash cost, which contains
`catalyst for soot oxidation. is implied inside the pores. As
`aresuluheporosltyofthe DPF meteriallsreduoed and
`the post-coated pore size distribution is changed.
`
`Asihetotalporcsityistflekeyparameterforlow
`pressure-drop with soot
`loading.
`low pressure-drop
`Cordierite and Sic DPF are good candidates for the
`catalysed DPF. Some catalyzed systems are also
`considering NOX adsorption catalyst. For this system.
`larger porosity is preferable.
`
`the
`on
`pressure-drop measttements
`Preliminary
`catalyzed DPF's were conducted. For this evaluation,
`DHC-550 and DHC-611 were used. Cell structure for
`this evaluation was 12miU300cpsl. and ‘l00giL of wash
`coating was loaded onto the fitters. Figure 21 shows the
`presstre-drop alter coating. The pressure drop before
`wash coating is shown in Figure 17.
`
`14
`
`
`
`
`
`LOW BACK-PRESSURE TYPE SIC DPF MATERIAL
`
`A new type of Silicon Carbide (SIC) OFF [5] is now
`available. This material is a composite material of SIC
`and Si. The metal Silicon bonds the Sic particles.
`
`The re-crystallized Sic is another well—|-rnown DPF
`material. This material contains two different shes of
`particles. One is around 0.1 micrometer and the other is
`from 10 to 20 micrometer. The re-crystallized Sic
`requires vaporization of small particles and the surface
`of large particles for sintering. The tiring temperature will
`exceed 2200 degree 0. As pores in the green body
`prevent SIC particles from this sirrtering mechanism. re-
`crystallized Sic is limited in porosity. The pore forming
`material. which is currently used for Cordierite material
`for porosity control. can not be applied.
`
`sintering
`the Si-bonded Sic utili