U.S. patent application number 13/687176 was filed with the patent office on 2014-05-29 for graphite blending method for ceramic shrinkage control.
The applicant listed for this patent is Daniel Edward McCauley, Anthony Nicholas Rodbourn, Patrick David Tepesch. Invention is credited to Daniel Edward McCauley, Anthony Nicholas Rodbourn, Patrick David Tepesch.
Application Number | 20140145360 13/687176 |
Document ID | / |
Family ID | 49780336 |
Filed Date | 2014-05-29 |
United States Patent
Application |
20140145360 |
Kind Code |
A1 |
McCauley; Daniel Edward ; et
al. |
May 29, 2014 |
GRAPHITE BLENDING METHOD FOR CERAMIC SHRINKAGE CONTROL
Abstract
A method for green-to-fired shrinkage control in honeycomb
ceramic article manufacture, including: measuring, prior to mixing,
the particle size distribution properties of at least one fine
particle size graphite pore former ingredient of a provided ceramic
source batch mixture; calculating the expected shrinkage of the
green body to the fired ceramic article based on the measured
particle size distribution properties of the at least one fine
particle graphite pore former; making the honeycomb ceramic
article; measuring the shrinkage of the resulting fired honeycomb
ceramic article; and adjusting the ceramic source batch mixture in
a subsequent batch material schedule, as defined herein, wherein
the adjusted ceramic source batch mixture provides finished
honeycomb ceramic articles having controlled green-to-fired
shrinkage.
Inventors: |
McCauley; Daniel Edward;
(Montour Falls, NY) ; Rodbourn; Anthony Nicholas;
(Avoca, NY) ; Tepesch; Patrick David; (Corning,
NY) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
McCauley; Daniel Edward
Rodbourn; Anthony Nicholas
Tepesch; Patrick David |
Montour Falls
Avoca
Corning |
NY
NY
NY |
US
US
US |
|
|
Family ID: |
49780336 |
Appl. No.: |
13/687176 |
Filed: |
November 28, 2012 |
Current U.S.
Class: |
264/40.1 |
Current CPC
Class: |
C04B 38/0006 20130101;
C04B 2235/425 20130101; C04B 2111/00129 20130101; B28B 17/0072
20130101; C04B 35/478 20130101; C04B 38/06 20130101; C04B 2111/34
20130101; C04B 38/0006 20130101; C04B 35/478 20130101; C04B 38/068
20130101; C04B 2235/9615 20130101; B28C 7/0418 20130101; C04B
2235/6021 20130101 |
Class at
Publication: |
264/40.1 |
International
Class: |
B28B 17/00 20060101
B28B017/00 |
Claims
1. A method for green-to-fired shrinkage control in honeycomb
ceramic article manufacture, comprising: measuring, prior to
mixing, the particle size distribution properties of at least one
fine particle graphite pore former ingredient of a provided ceramic
source batch mixture; calculating the expected shrinkage of the
green body to the fired ceramic article based on the measured
particle size distribution properties of the at least one fine
particle graphite pore former; making the honeycomb ceramic article
comprising: preparing an extruded green body from the provided
ceramic source batch mixture having the at least one fine particle
graphite pore former having the measured particle size distribution
properties; and firing the extruded green body to form the
honeycomb ceramic article; measuring the shrinkage of the resulting
fired honeycomb ceramic article; and adjusting the ceramic source
batch mixture in a subsequent batch material schedule according to:
if the measured shrinkage is greater than about 0.15 relative %
positive deviation from the calculated expected shrinkage, then
subtract 0.5 wt % from the fine particle size graphite content; or
if the measured shrinkage is greater than about 0.15 relative %
negative deviation from the calculated expected shrinkage then add
0.5 wt % to the fine particle size graphite content.
2. The method of claim 1 further comprising: extruding the adjusted
batch mixture to form an extruded green body; and firing the
extruded green body to form the ceramic article having controlled
shrinkage.
3. The method of claim 2 wherein the controlled shrinkage is less
than about +0.15 relative % to about -0.15 relative % from the
expected shrinkage.
4. The method of claim 1 wherein the measured shrinkage of the
fired honeycomb ceramic article measures a change in at least one
of: the diameter of the article, the major axis of the article, or
a combination thereof.
5. The method of claim 1 wherein the at least one fine particle
size graphite comprises a particle size distribution having a d50
of 3 to 5 microns, and a d90 less than 8 microns.
6. The method of claim 1 wherein calculating the expected shrinkage
(dS) of the green body to the fired ceramic article is accomplished
according to the formula: dS=dFG/A where dS the expected shrinkage
if the fine graphite is held constant, dFG is the change in the
amount of the at least one fine particle size graphite in the batch
mixture, and A is a scale factor which depends on the other batch
components and the attributes of the fine particle size graphite
selected.
7. The method of claim 1 wherein the provided ceramic source batch
mixture has a shrinkage variability of from about +0.5% to about
-0.5% from a nominal value over a production period of about 3
months to 2 years.
8. The method of claim 1 wherein the ceramic source batch mixture
having the at least one fine particle size graphite pore former
further includes at least one of: an aluminum oxide source; a
silica source; a titanium oxide source; a lanthanum oxide source; a
starch containing pore former; a coarse graphite pore former; and
mixtures thereof.
9. The method of claim 1 further comprising: if the expected
shrinkage is neither greater than positive 0.15 relative % or
greater than negative 0.15 relative % of the penultimate calculated
expected shrinkage, then use the penultimate batch sheet for the
next batch material schedule.
10. A method for green-to-fired shrinkage control in honeycomb
ceramic article manufacture, comprising: determining the
green-to-fired shrinkage of a green body when fired to a fired
ceramic article, wherein the green body is prepared from a ceramic
source batch mixture having the at least one fine particle graphite
pore former having a measured particle size distribution; comparing
the determined green-to-fired shrinkage of the resulting fired
ceramic article with an expected shrinkage in the fired ceramic,
wherein the expected shrinkage of the fired ceramic has a positive
deviation or negative deviation of less than 0.15 relative %;
adjusting the weight % of the at least one fine particle graphite
pore former having the measured particle size distribution in the
ceramic source batch mixture if the determined green-to-fired
shrinkage is greater than the expected shrinkage; and firing the
green body to form the honeycomb ceramic article having controlled
shrinkage.
11. The method of claim 10 wherein the controlled shrinkage of the
honeycomb ceramic article is a shrinkage deviation of less than
about 0.15 relative %.
12. The method of claim 10 wherein adjusting the weight % of the at
least one fine particle graphite pore former comprises adding or
subtracting graphite pore former from the batch schedule in an
amount of about 0.5 weight % or more.
13. The method of claim 10 wherein the at least one fine particle
graphite pore former comprises a mixture of two or more fine
particle graphite pore formers having different measured particle
size distributions.
14. The method of claim 10 wherein adjusting the weight % of the at
least one fine particle graphite pore former in the source batch
mixture is according to: if the expected shrinkage is more negative
than about negative 0.15 relative %, then adding fine particle size
graphite to the batch schedule; or if the expected shrinkage is
more positive than about positive 0.15 relative %, then subtracting
fine particle size graphite from the batch schedule; and if the
expected shrinkage is less negative than about negative 0.15
relative % or less positive than about positive 0.15 relative %,
then maintaining the fine particle size graphite content in the
batch schedule at a constant level.
15. The method of claim 14 wherein the adjusting the weight % of
the at least one fine particle graphite pore former comprises
adding or subtracting graphite pore former from the batch schedule
in an amount of about 0.5 weight % to about 2.5 weight %.
16. The method of claim 10 wherein the ceramic source batch mixture
further comprises at least one coarse particle size graphite pore
former.
Description
[0001] The entire disclosure of any publication or patent document
mentioned herein is incorporated by reference.
BACKGROUND
[0002] The disclosure relates to a method for controlling shrinkage
during firing in ceramic article manufacture.
SUMMARY
[0003] The disclosure provides a method for controlling shrinkage
of ceramic articles, such as honeycomb filters, during firing of a
green body to the ceramic article.
[0004] In embodiments, the disclosure provides a method for
green-to-fired shrinkage control in honeycomb ceramic article
manufacture, including:
[0005] measuring, prior to mixing, the particle size distribution
properties of at least one fine particle size graphite pore former
ingredient of a provided ceramic source batch mixture;
[0006] calculating the expected shrinkage of the green body to the
fired ceramic article based on the measured particle size
distribution properties of the at least one fine particle graphite
pore former;
[0007] making the honeycomb ceramic article;
[0008] measuring the shrinkage of the resulting fired honeycomb
ceramic article; and
[0009] adjusting the ceramic source batch mixture in a subsequent
batch material schedule, as defined herein, wherein the adjusted
ceramic source batch mixture provides finished honeycomb ceramic
articles having controlled green-to-fired shrinkage.
BRIEF DESCRIPTION OF DRAWINGS
[0010] In embodiments of the disclosure:
[0011] FIG. 1 illustrates the effect of the fine end of the
graphite distribution on green-to-fired shrinkage.
[0012] FIG. 2 illustrates a time series of filter shrinkage
observed for numerous runs over an extended period of about 1.5
years.
[0013] FIG. 3 illustrates an exemplary decision tree for filter
shrinkage control via secondary fine graphite addition.
[0014] FIG. 4 illustrates the relationship between the graphite
surface area (SA in meters squared per gram), which is analogous to
particle size in graphite, to the co-efficient of thermal expansion
(CTE) of the fired ceramic.
[0015] FIG. 5 illustrates the relationship between surface area
(SA), which is analogous to particle size in graphite, to %
porosity tested at two different field locations indicated by the
respect squares and diamonds.
[0016] FIGS. 6A and 6B, graphically illustrates the relation of the
predicted diameter shrinkage of the fired ceramic article using
classical regression and neural network methods, respectively
compared to actual measured diameter shrinkage.
DETAILED DESCRIPTION
[0017] Various embodiments of the disclosure will be described in
detail with reference to drawings, if any. Reference to various
embodiments does not limit the scope of the invention, which is
limited only by the scope of the claims attached hereto.
Additionally, any examples set forth in this specification are not
limiting and merely set forth some of the many possible embodiments
of the claimed invention.
[0018] In embodiments, the disclosed apparatus, and the disclosed
method of making provide one or more advantageous features or
aspects, including for example as discussed below. Features or
aspects recited in any of the claims are generally applicable to
all facets of the invention. Any recited single or multiple feature
or aspect in any one claim can be combined or permuted with any
other recited feature or aspect in any other claim or claims.
[0019] "Include," "includes," or like terms means encompassing but
not limited to, that is, inclusive and not exclusive.
[0020] "About" modifying, for example, the quantity of an
ingredient in a composition, concentrations, volumes, process
temperature, process time, yields, flow rates, pressures,
viscosities, and like values, and ranges thereof, employed in
describing the embodiments of the disclosure, refers to variation
in the numerical quantity that can occur, for example: through
typical measuring and handling procedures used for preparing
materials, compositions, composites, concentrates, or use
formulations; through inadvertent error in these procedures;
through differences in the manufacture, source, or purity of
starting materials or ingredients used to carry out the methods;
and like considerations. The term "about" also encompasses amounts
that differ due to aging of a composition or formulation with a
particular initial concentration or mixture, and amounts that
differ due to mixing or processing a composition or formulation
with a particular initial concentration or mixture.
[0021] The indefinite article "a" or "an" and its corresponding
definite article "the" as used herein means at least one, or one or
more, unless specified otherwise.
[0022] Abbreviations, which are well known to one of ordinary skill
in the art, may be used (e.g., "h" or "hr" for hour or hours, "g"
or "gm" for gram(s), "mL" for milliliters, and "rt" for room
temperature, "nm" for nanometers, and like abbreviations).
[0023] Specific and preferred values disclosed for components,
ingredients, additives, and like aspects, and ranges thereof, are
for illustration only; they do not exclude other defined values or
other values within defined ranges. The methods of the disclosure
can include any value or any combination of the values, specific
values, more specific values, and preferred values, including
intermediate values and ranges, described herein.
[0024] The ability to produce extrude-to-shape diesel particulate
filter (DPF) articles can depend on the ability to minimize
variability in the amount the filter shrinks (or grows) during the
firing process. Dimensional tolerances being implemented in the
industry are narrowing and so is the acceptable shrinkage
variation. Graphite pore-former particle size distribution (PSD),
especially the fine end of that distribution, has been identified
as a significant factor in shrinkage and shrinkage variability.
Generally, the larger the population of small particles of graphite
pore-former present, for example in wt %, the more the fired
ceramic article or part will shrink. The very fine pores created by
the graphite pore-former eventually sinter away during the firing
process. Variability in the fine end of the distribution directly
leads to more shrinkage variation. Additional details on this
phenomenon and subsequent control strategy have been disclosed in
commonly owned and assigned U.S. patent application Ser. No.
12/395,005, now U.S. Patent Application Publication No.
2010/0052200. The commonly owned and assigned application disclosed
a single graphite pore former and did not mention an iterative
property control method.
[0025] Graphite has been either all or part of the pore former
package for commercial diesel particle filter (DPF) compositions as
a fugitive pore former in the past but has been limited to a single
grade within a given composition. This present disclosure provides
a method involving blending a fine particle size graphite pore
former with a coarse particle size pore forming graphite, in
varying amounts, so as to stabilize the naturally occurring
shrinkage variability due to upstream PSD variability of all of the
other batch raw materials that may impact shrinkage. The PSD of the
added fine graphite component must be a fine enough grade to
produce pores that will completely sinter away in the firing
process. The fine particle size graphite completely sinters away
and permits control of ceramic article shrinkage variability
without impacting other physical or chemical properties.
[0026] In embodiments, the present disclosure provides a method for
green-to-fired shrinkage control in honeycomb ceramic article
manufacture, comprising:
[0027] measuring, prior to mixing, the particle size distribution
properties of at least one fine particle graphite pore former
ingredient of a provided ceramic source batch mixture;
[0028] calculating the expected shrinkage of the green body to the
fired ceramic article based on the measured particle size
distribution properties of the at least one fine particle graphite
pore former;
[0029] making the honeycomb ceramic article comprising: [0030]
preparing an extruded green body from the provided ceramic source
batch mixture having the at least one fine particle graphite pore
former having the measured particle size distribution properties;
and [0031] firing the extruded green body to form the honeycomb
ceramic article; measuring the shrinkage of the resulting fired
honeycomb ceramic article; and
[0032] adjusting the ceramic source batch mixture in a subsequent
batch material schedule according to: [0033] if the measured
shrinkage is greater than about 0.15 relative % positive deviation
from the calculated expected shrinkage, then subtract 0.5 wt % from
the fine particle size graphite content; or [0034] if the measured
shrinkage is greater than about 0.15 relative % negative deviation
from the calculated expected shrinkage then add 0.5 wt % to the
fine particle size graphite content.
[0035] In embodiments, the method can include or further comprise,
for example,
[0036] extruding the adjusted batch mixture to form an extruded
green body; and
[0037] firing the extruded green body to form the ceramic article
having controlled shrinkage properties.
[0038] In embodiments, the controlled shrinkage can be, for
example, less than about +0.15 relative % to about -0.15 relative %
difference compared to the expected shrinkage.
[0039] In embodiments, the measured shrinkage of the fired
honeycomb ceramic article measures a change in at least one of: the
diameter of the article, the major axis of the article, or a
combination thereof.
[0040] In embodiments, the at least one fine particle size graphite
comprises a particle size distribution having a d50 of 3 to 5
microns, and a d90 less than 8 microns.
[0041] In embodiments, calculating the expected shrinkage (dS) of
the green body to the fired ceramic article can be accomplished,
for example, according to the formula:
dS=dFG/A
where dS the expected shrinkage if the fine graphite is held
constant, dFG is the change in the amount of the at least one fine
particle size graphite in the batch mixture, and A is a scale
factor which depends on the other batch components and the
attributes of the fine particle size graphite selected.
[0042] In embodiments, the provided ceramic source batch mixture
can have a shrinkage variability of from about +0.5% to about -0.5%
from a nominal value over a short term, intermediate term, or long
term production period, for example, from about 1 month to 5 years,
from about 2 months to 4 years, from about 3 months to 2 years, and
like intervals, including intermediate values and ranges.
[0043] In embodiments, aluminum titanate based compositions are
provided that can have a porosity of, for example, from 43% to 65%,
and can have a web thickness of, for example, from 9 mil to 14
mil.
[0044] In embodiments, the ceramic source batch mixture having the
at least one fine particle size graphite pore former can further
include, for example, at least one of:
[0045] an aluminum oxide source;
[0046] a silica source;
[0047] a titanium oxide source;
[0048] a lanthanum oxide source;
[0049] a starch containing pore former;
[0050] a coarse graphite pore former;
[0051] and mixtures thereof.
[0052] In embodiments, the method can further comprise, for
example, verifying the stability of the method of making the
ceramic article over time, for example, from batch-to-batch, from
day-to-day, week-to-week, month-to-month, year-to-year, and like
useful time intervals. If the method of making is satisfactorily
stable with respect to shrinkage control or tolerable article
shrinkage, and the raw materials are satisfactorily consistent,
then relying on the previously determined batch sheet or
penultimate batch sheet for use in the next or subsequent batch
material schedule. In one specific illustrative example:
[0053] if the expected shrinkage, for example based on measuring
the fine particle size graphite pore former ingredient, is neither
greater than positive 0.15 relative % or greater than negative 0.15
relative % of the penultimate (i.e., previous or one before the
last) calculated expected shrinkage, then use the penultimate batch
sheet for the next batch material schedule. (i.e., null adjustment
or "in spec" production, when, for example, there is insufficient
variability in the shrinkage property to merit an adjustment in the
fine particle size graphite pore former content.)
[0054] In embodiments, the present disclosure provides a method for
green-to-fired shrinkage control in honeycomb ceramic article
manufacture, comprising:
[0055] determining the green-to-fired shrinkage of a green body
when fired to a fired ceramic article, wherein the green body is
prepared from a ceramic source batch mixture having the at least
one fine particle graphite pore former having a measured particle
size distribution;
[0056] comparing the determined green-to-fired shrinkage of the
resulting fired ceramic article with an expected shrinkage in the
fired ceramic, wherein the expected shrinkage of the fired ceramic
has a positive deviation or negative deviation of less than 0.15
relative %;
[0057] adjusting the weight % of the at least one fine particle
graphite pore former having the measured particle size distribution
in the ceramic source batch mixture if the determined
green-to-fired shrinkage is greater than the expected shrinkage;
and
[0058] firing the green body to form the honeycomb ceramic article
having controlled shrinkage.
[0059] In embodiments, the controlled shrinkage of the honeycomb
ceramic article can be, for example, a shrinkage deviation of less
than about 0.15 relative %.
[0060] In embodiments, adjusting the weight % of the at least one
fine particle graphite pore former comprises adding or subtracting
graphite pore former from the batch schedule in an amount of about
0.5 weight % to about 5 wt % or more, such as 0.6, 0.7, 0.8, 0.9,
1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0 wt %, or
more, including intermediate values and ranges.
[0061] In embodiments, the at least one fine particle graphite pore
former can comprise a mixture of two or more fine particle graphite
pore formers having different measured particle size
distributions.
[0062] In embodiments, the ceramic source batch mixture can further
comprise at least one of a coarse particle size graphite pore
former, alone or in combination with a fine particle size graphite
pore former or with another non-graphite pore former.
[0063] In embodiments, the pore former can further comprise a
mixture of two or more graphite pore formers having different
measured particle size distributions, such as a mixture of a fine
graphite pore former and coarse graphite pore former.
[0064] In embodiments, the pore former can further comprise a
mixture of one or more graphite pore formers and one or more
non-graphite pore formers, such as pea starch or like non-graphite
pore formers. The pore formers can have, for example, the same or
different measured particle size distributions.
[0065] In embodiments, adjusting the weight % of the at least one
fine particle graphite pore former in the above mentioned source
batch mixture can be accomplished is according to, for example:
[0066] if the expected shrinkage is more negative than about
negative 0.15 relative %, then adding fine particle size graphite
to the batch schedule; or
[0067] if the expected shrinkage is more positive than about
positive 0.15 relative %, then subtracting fine particle size
graphite from the batch schedule; and
[0068] if the expected shrinkage is less negative than about
negative 0.15 relative % or less positive than about positive 0.15
relative %, then maintaining the fine particle size graphite
content in the batch schedule at a constant level.
[0069] In embodiments, adjusting the weight % of the at least one
fine particle graphite pore former can comprise, for example,
adding or subtracting graphite pore former from the batch schedule
in an amount of about 0.5 weight % to about 2.5 weight %.
[0070] In embodiments, "consisting essentially of" can refer to,
for example:
[0071] a method for shrinkage property control in ceramic article
manufacture, such as honeycomb filters, including:
[0072] characterizing at least one of a mixture of ceramic source
batch materials, including a fine particle size graphite pore
former, such as the particle size distribution of the graphite pore
former;
[0073] calculating the expected shrinkage (i.e., to determine, for
example, the predicted shrinkage) in the fired ceramic made from
the ceramic source batch materials; and
[0074] adjusting the source batch materials according to: [0075] if
the expected shrinkage is greater than about a positive 0.15
relative percent ("rel %"), then subtract 0.5 wt % of the fine
particle size graphite content from the batch material schedule; or
[0076] if the expected shrinkage is more negative than about a
negative 0.15 rel % then add 0.5 wt % of the fine particle size
graphite content to the batch material schedule; and
[0077] extruding and firing the batch materials to obtain a
honeycomb ceramic article having controlled shrinkage of about 0.15
rel % shrinkage or less shrinkage.
[0078] In embodiments, the disclosure provides a method for
controlling shrinkage and controlling shrinkage variability in
graphite containing diesel particle filter (DPF) batch compositions
by adding or removing specified amounts of the fine particle size
(e.g., d50 3 less than or equal to 5 microns, such as 3 to 5
microns, d90 less than 8 microns), graphite component from the
batch materials.
[0079] In embodiments, the disclosed method, in combination with
the knowledge and understanding of the shrinkage of the fired
article, can provide a manufacturing process that can continuously
extrude a constant size green body honeycomb structure.
[0080] Shrinkage variability over the time, such as many days or
weeks of production, can be greatly reduced using the disclosed
iterative graphite blending method along with the knowledge of the
shrinkage shifts caused by other raw materials. When shrinkage
variability is relatively constant over the time, the shrinkage
property can be purposely altered by, for example, changing the
weight percent of one or more batch ingredients, such as the fine
particle size graphite content, or other ingredients of the
batch.
[0081] In embodiments, the disclosed method provides a significant
reduction in the shrink mask skin former hardware called for to
produce satisfactory product and provides significant tooling cost
savings.
[0082] In embodiments, the disclosed method permits honeycomb skin
finishing to move from a shrink mask skin former type system to a
die cut skin former system or other skin finishing methods. A die
cut skin former system has numerous advantages including, for
example, reducing skin related defects such as collapsed cells, air
checks, skin fissures, and enables a variety of other known skin
forming technologies to be used. By facilitating the use of more
advanced skin technologies, skin finish defects can be reduced.
[0083] In embodiments, the disclosed shrinkage control and blending
method minimizes the impact of variations in the physical
properties of batch materials on the performance of batch extrusion
and the resulting extruded, dried, and fired honeycomb
structures.
[0084] Referring to the Figures, FIG. 1 shows the fine end of the
graphite pore-former particle size distribution curve (d.sub.10)
and the fine end can have a significant impact on the fired
shrinkage of, for example, an aluminum titanate (AT) body. FIG. 1
illustrates the effect of the fine end of the graphite distribution
on green-to-fired shrinkage.
[0085] FIG. 2 illustrates a time series of filter shrinkage (or
growth) in relative % (y-axis) observed for numerous runs over an
extended extrusion interval time in hours.times.1000 (x-axis) of
about 13,000 hours or about 1.5 years. Previous work investigating
the graphite/shrinkage relationship focused on getting consistent
graphite pore former to minimize variability in firing shrinkage
(McCauley, et al., see copending U.S. patent application Ser. No.
12/395,005, now U.S. Patent Application Publication No.
2010/0052200). The prior approach dynamically changed the amount of
graphite in the green body to change the firing shrinkage
properties, and compensated for natural shrinkage variability seen
in the body and which variability is also seen in FIG. 2. The
x-axis represents the extrusion interval time (in
hours.times.1,000), that is for example, 0 to 14 hours
(.times.1,000) in 2 hour (.times.1,000) increments, or about 13,000
hours of data accumulated over about 1.5 years.
[0086] FIG. 3 illustrates an exemplary decision tree for filter
shrinkage control via secondary fine graphite addition or
subtraction. In embodiments, the steps of the disclosed process can
follow the illustrated decision tree (300), for example: First,
measure at least one characteristics of the incoming raw materials
(305), such as the particle size distribution of a graphite pore
former. Second, calculate the expected green-to-fired shrinkage
based on the measured raw material characteristic (310). Third,
query "Is the expected shrinkage within .+-.0.15 rel % of the prior
expected shrinkage?" (315). If the answer to the expected shrinkage
question (315) is "Yes" (317), then use the existing batch sheet
for the next production batch (318) (i.e., null adjustment). If the
answer to the expected shrinkage question (315) is "No" (319), then
query "Is the expected shrinkage higher or lower than the prior
expected shrinkage?" (320). If the answer to the penultimate (e.g.,
a previous production day) shrinkage prediction question (320) is
"Lower" (321), then "Add 0.5 wt % fine graphite to batch." (325).
However, if the answer to the penultimate (e.g., a previous
production day) shrinkage prediction question (320) is "Higher"
(326), then "Subtract 0.5 wt % fine graphite to batch" (330). Next,
the fine particle size graphite addition (325) or fine graphite
subtraction (330) directive is accomplished in the batch
formulation, followed by honeycomb green body extrusion, and firing
the green body steps. The resulting fired articles or wares can be
analyzed for porosity, or other properties, and the measured
porosity property information can be used in the subsequent
calculations to iteratively (340) further refine the green-to-fired
shrinkage behavior and shrinkage properties.
[0087] FIG. 4 illustrates the relationship between graphite surface
area (SA in meters squared per gram), which is analogous to
particle size in graphite, to the co-efficient of thermal expansion
(CTE) of the fired ceramic. This figure also shows that CTE is
unchanged over a wide range of graphite surface areas or particle
sizes.
[0088] FIG. 5 illustrates the relationship between surface area
(SA), which is analogous to particle size in graphite, to %
porosity tested at two different field locations indicated by the
respective squares and diamonds. This figure also shows that %
porosity is essentially unchanged over a wide range of graphite
surface area particles sizes.
[0089] FIGS. 6A and 6B, graphically illustrate the relation of
expected shrinkage in the diameter of the fired ceramic article
compared to actual measured diameter shrinkage. Data reduction
using principal component analysis (PCA) was significant for using
curves as predictors. A classical regression method was initially
helpful in defining and refining which predictors should be
followed. A classical regression used, for example, 22 variables
having all XiXj interactions and squares, to provide the following:
RMSE=16.5%; Press=50.4%; R.sup.2=55.5%; and R.sup.2.sub.Adj=52.9%.
A neural networks method included non-linear relationships and
further improved initial shrinkage correlations. A neural network
used, for example, 22 variables having 6 neurons to provide the
following: RMSE=8.7%; Press=13.2%; R.sup.2=87.1%; and
R.sup.2.sub.Adj=81.9%.
[0090] A typical batch sheet for the aluminum titanate ceramic
source composition is listed in Table 1, and can be a starting
point for batch compositions having a fine particle size graphite
formulation. The level of fine graphite in the batch can be changed
(i.e., added or subtracted) to compensate for expected shrinkage
shifts. Additional fine particle size graphite can be added to the
batch to increase shrinkage and to counteract a drop in expected
shrinkage. Conversely, fine particle size graphite can be removed
or subtracted from the batch, i.e., a reduced amount or an amount
left out the batch recipe, or other ingredients can be, for
example, increased, to effectively decrease shrinkage and to
counteract a rise in expected shrinkage. For this method to be
implemented in large scale manufacturing, the standard "home"
composition can contain some amount of fine particle size graphite
to allow for flexibility in both shrinkage directions. FIG. 3 shows
a flow chart illustrating the formulation method.
TABLE-US-00001 TABLE 1 Batch sheet for an aluminum titanate
composition having a blend of two graphite sources having different
particle size distributions. Material Wt % Inorganics (100%) Silica
quartz 10.31% SrCO.sub.3 8.10% CaCO.sub.3 1.39% Aluminum Oxide
49.67% Titanium oxide 30.33% La.sub.2O.sub.3 0.20% Pore Former
Coarse Graphite 3.50% (Super-addition to Inorganics) Native Pea
Starch 8.00% Fine Graphite 1.00% Methocel/Soap(Super- hydroxypropyl
methyl 3.00% addition to Inorganics + cellulose (a blend of K and F
Pore Formers) Methocels) Formers) hydroxypropyl methyl 1.50%
cellulose (Methocel F240) Liquid Organics Tall Oil L5 1.00%
(Super-addition to Inorganics + Pore Formers) Batch Water
15.60%
[0091] In embodiments, one exemplary aluminum titanate composition
of the disclosure can have, for example, about 50% porosity and a
13 mil web thickness when prepared using a Methocel F240. In
embodiments, another exemplary aluminum titanate composition of the
disclosure can have, for example, about 45% porosity and a 10 mil
web thickness when prepared using a blend of K and F Methocels. In
embodiments, one exemplary aluminum titanate composition of the
disclosure can be, for example, prepared using a mixture of
Methocel F240 and a blend of K and F Methocels.
[0092] In embodiments, the coarse graphite pore former particle
size distribution properties can be, for example: D10 9 to 13
microns; D50 35 to 45 microns; and D90 80 to 120 microns. In
embodiments, the fine graphite pore former particle size
distribution properties can be, for example: D10 1 to 3 microns;
D50 3 to 5 microns; and D90 6 to 8 microns.
[0093] Extrusion experiments using graphite as a shrinkage control
additive were accomplished in a CelPro laboratory extruder. The
entire particle size distribution of the graphite was purposely
changed to affect shrinkage. Shrinkage did change with large scale
graphite moves (i.e., particle size distribution changes) but other
properties, such as % porosity and CTE, remained unchanged. For
this experiment, surface area (m.sup.2 per gram) and porosity was
used as a surrogate for particle size as shown in FIGS. 4 and 5,
respectively, since there is a close correlation between the
graphite particle size properties and the resulting fired ceramic's
internal pore structure.
[0094] Other shrinkage control strategies have been previously
investigated and developed by Corning, Inc. The strategies
generally fall into two categories: passive control, and active
control. Passive control strategies, i.e., strategies that seek to
lessen the magnitude or retard the rate of change of shrinkage, can
include, for example, ensuring incoming raw materials are very
consistent, particularly Al.sub.2O.sub.3, along with blending raw
material lots together to dampen any changes of a new starting
batch material. Active control strategies can include, for example,
sodium additions to the batch to counter act "natural" shrinkage
variation and as described in the present disclosure.
Example 1
Predicting Shrinkage
[0095] To predict shrinkage, statistical techniques were used based
on raw material sample characteristics, specifically particle size.
First, a sample of every powder raw material ingredient is obtained
from the batch house and cataloged. Ideally, the resulting product
is within a suitable fired product specification. After dispersing
and de-agglomerating each powder raw material, the particle size
property was measured by, for example, laser light scattering
(e.g., Microtrac or CILAS).
[0096] An example procedure for sample preparation for Microtrac
particle size measurements can include, for example: 1) obtaining a
particle sample; 2) loosely mixing the sample by repeatedly turning
over or rolling the sample container; 3) adding a small amount of
sample (e.g., about 0.05 grams) directly to the sample bath on a
Microtrac instrument; 4) adding about 3 mL of a 5% aqueous Triton-X
solution to the vessel containing the graphite particle sample; 5)
sonicating the sample at 50 watts for 35 seconds; 6) verifying that
obscuration is within the allowable machine limit; and 7) begin the
particle size measurement process.
[0097] The full curve incremental data was then analyzed by the
principal components analysis (PCA) technique which transforms the
data from about 40 to 100 numbers (set 1 or original variables) to
about 4 to 5 numbers (set 2 or principal components). Both number
sets accurately describe the whole curve. These principle component
numbers along with water call (i.e., added water) are then inserted
into the calculation, for example, an empirical or neural net, as
predictors. The calculation produces a response of the expected
shrinkage.
[0098] Principal component analysis (PCA) is a mathematical
procedure that uses an orthogonal transformation to convert a set
of observations of possibly correlated variables into a set of
values of linearly uncorrelated variables called principal
components. The number of principal components is less than or
equal to the number of original variables. This transformation is
defined in such a way that the first principal component has the
largest possible variance (that is, accounts for as much of the
variability in the data as possible), and each succeeding component
in turn has the highest variance possible under the constraint that
it be orthogonal to (i.e., uncorrelated with) the preceding
components. Principal components are guaranteed to be independent
only if the data set is jointly normally distributed. PCA is
sensitive to the relative scaling of the original variables.
Alternative names for PCA are the discrete Karhunen-Loeve transform
(KLT), the Hotelling transform, or proper orthogonal decomposition
(POD).
[0099] The disclosure has been described with reference to various
specific embodiments and techniques. However, it should be
understood that many variations and modifications are possible
while remaining within the scope of the disclosure.
* * * * *