Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02260197 1998-12-29
W O 98/0140S PCT/GB97/01706
CF~AMIC CQMPOSITIONS
This invention relates to ceramic compositions which are of
particular value in the handling and casting of high melting temperature
metals such as iron or steel.
It is common practice to make articles, which are used in the
handling and casting of molten metals such as steel, from carbon bonded
ceramics (also known as black refractories). Examples of such articles are
pouring nozles for molten metal-containing vessels such as ladles or
tundishes, and shrouds which surround the metal stream flowing from one
vessel to another. These carbon bonded ceramics are formed from a
mixture of graphite, one or more oxides such as alumina, magnesia and
zirconia, and a binder such as a phenolic resin or pitch which will
decompose to produce a carbon bond.
The above carbon bonded ceramic materials suffer from a number
of disadvantages. They have poor thermal shock resistance and tend to
crack, so that it is necess~ry to treat articles such as nozzles and shrouds
in some way so as to minimise the thermal shock produced when the
articles are heated rapidly to elevated temperatures. The materials also
have low oxidation resistance as they contain a relatively high proportion
of carbon, mainl!~ in the form of graphite. The materials also suffer from
additional disadvantages in specific arplications. For example, the outer
surface of a nozzle is susceptible to attack by slag present on the surface
of the molten metal in which the nozle is immersed (known as slag line
attack), and the bore of a nozzle tends to become clogged in use due to
the build up of alumina, when casting aluminium killed steel.
It has now been found that a carbon bonded ceramic material
consisting of a mixture of boron nitride, zirconium diboride and at least one
other refractory material, is particularly useful as an alternative to
conventional graphite-containing carbon bonded ceramics for the
CA 02260197 1998-12-29
W O 98/01405 PCT/GB97101706
production of articles used for the handling and casting of molten metals,
such as steel.
According to a first feature of the invention there is provided a
ceramic composition comprising a mixture of particles of boron nitride,
zirconium diboride and at least one other refractory material bonded
together by carbon produced by the decomposition of an organic binder.
The other refriactory material may be for example a refractory metal,
an oxide, a carbide, a boride or a nitride.
The refractory metal may be for example boron.
Examples of suitable refractory oxides include aluminium oxide,
zirconium oxide, magnesium oxide, yttrium oxide, calcium oxide,
chromium oxide and silicon oxide. More than one oxide may be used, and
the oxide may be a rnixed refractory oxide such as mullite.
Examples of suitable carbides include silicon carbide, boron
carbide, aluminium c:arbide and zirconium carbide. More than one carbide
may be used.
Examples of suitable borides include titanium diboride and calcium
hexaboride, and examples of suitable nitrides include silicon nitride,
aluminium nitride, titanium nitride, zirconium nitride and sialon. More than
one boride and more! than one nitride may be used.
According to one preferred embodiment of the invention the
ceramic composition comprises a mixture of boron nitride, zirconium
diboride and zirconium oxide, and the ceramic composition preferably
contains 5 - 70 % by weight of boron nitride, more preferably 15 - 50 % by
weight, 5 - 60 % by weight of zirconium diboride, more preferably 15 - 50
CA 02260197 1998-12-29
WO 98/01405 PCI'IGB97101706
% by weight, and !; - 80 % by weight of zirconium oxide, more preferably
10 - 60 % by weight
According to another preferred embodiment of the invention the
ceramic composition comprises a mixture of boron nitride, zirconium
diboride and aluminium oxide, and the ceramic composition preferably
contains 5 - 70 % by weight of boron nitride, more preferably 15 - 50 % by
weight, 5 - 60 % by weight of zirconium diboride, more preferably 15 - 50
% by weight, and 10 - 70 % by weight of aluminium oxide, more
preferably 15 - 60 ~'0 by weight.
In the above preferred embodiments the proportion of each of the
components of the ceramic composition is expressed as percentage by
weight based on the total weight of the cerdl";c composition, excluding the
carbon bond.
The organic binder which decomposes to produce a carbon bond
may be for example a phenol-formaldehyde resin such as a novolac or a
resol phenol-formaldehyde resin, a urea-formaldehyde resin, a
melamine-formaldehyde resin, an epoxy resin, a furane resin or pitch.
The organic binder is preferably a phenol-formaldehyde resin, and it
is preferred that thle resin is used in the form of a liquid. A powdered
phenolic resin can be used but it is necessary to dissolve the resin in a
suitable solvent, such as furfural, in ofder to mix the resin with the other
components and produce the ceramic composition. The amount of liquid
phenolic resin usecl will usually be of the order 5 - 25%, preferably 10 -
15% by weight, balsed on the total of the other components, and after
production of the ceramic composition, the composition will usuatly contain
2 - 12% by weight, preferably of the order of 5% by weight, of carbon
produced by decomposition of the resin, based on the total weight of the
ceramic composition.
CA 02260197 1998-12-29
W O 98/0140S PCT/GB97/01706
The ceramic compositions of the invention may be produced by first
mixing together particles of the boron nitride, the zirconiurrl diboride and
the other refractory rnaterial, and then adding the liquid resin and mixing
until the mixture of the particles and the resin is homogeneous. It may be
necessary to heat the mixture to reduce the liquid content of the resin to
render the mixture suitable for forming. The mixture is then formed to a
desired shape, preferably by cold isostatic pressing of the mixture in a
suitable mould. After forming the shape is heated to cure and cross-link
the resin, for example at about 150~ - 300~ C for about 1 hour, and then
heated at about 70()~ - 1200~ C to pyrolyse the resin and produce a
carbon bond.
Although the ceramic compositions of the invention may be used for
other applications, for example in the melting and handling of glass or in
the melting, handling and casting of relatively low melting temperature
metals such as aluminium and its alloys, the compositions are particularly
useful for use in the! handling and casting of high melting temperature
metals such as iron or steel.
When used in the handling and casting of a metal such as steel
each of the three cornponents of the ceramic compositions of the inventio
confers particular properties on the compositions. The boron nitride
makes the compositions non-wetting in the presence of molten steel or
molten slag, and hence when used for example in a composition which is
used for a casting nozzle will preve lt clogging of the nozzle due to
alumina build up. In addition the boron nitride makes the compositions
resistant to thermal shock, and helps to protect the compositions from
oxidation. The zirconium diboride confers erosion resistance, gives
protection against oxidation at higher temperatures ( up to about 1250~ C)
than does the boron nitride, and improves the resistance of the
compositions to attack by molten slag. In the preferred embodiments both
the aluminium oxide and the zirconium oxide improve the resistance of the
composition to attack by molten steel.
.
CA 02260197 1998-12-29
- W O 98101405 PCT/GB97/01706
In order to increase the oxidation resistance of the compositions at
higher temperatures, for example up to about 1400~ C, it is desirable to
include in the compositions a proportion, for example ~ - 20% by weight
based on the weight of the composition, of silicon carbide and/or titanium
diboride, as at least part of the third refractory material.
Examples ol applications for the ceramic compositions of the
invention in the handling and casting of steel are lining materials, and
nozzles and shrouds, such as those used in continuous casting. The
zirconium oxide-containing composition described above is particularly
suitable for forming that part of a nozzle which in use is at the boundary
between the surface of molten steel and molten slag which lies on top of
the steel. The aluminium oxide-containing composition described above is
particularly suitable for forming the inside of a nozzle, since it can readily
be co-pressed with an alumina-graphite material which forms the rest of
the nozzle, and it prevents build up of alumina and clogging of the nozle.
While these compositions may be used to form the whole nozzle if
desired, it is preferred to use them only to form portions of the nozzles as
described. The rernainder of the nozzles can then be formed from a
conventional carbon bonded ceramic material such as a carbon bonded
alumina and graphite mixture.
The following examples will serve to illustrate the invention:
Exan~le 1
~ A series of compositions was prepared as in Table 1 below. The
amount of each of the refractory components is expressed as percentage
by weight based on the total, and the amount of liquid resin is expressed
as percentage by w~!ight of the total of the refractory cornponents.
. ... ..
CA 02260197 1998-12-29
W 098/01405 PCT/GBg7/01706
Table 1
CGIII - -~ BN ZrEI7 Al20, ZrOi SiC Resin
NO.
- ~ 10
2 25 40 35 - ~ 1 3
3 30 35 35 ~ ~ 1 5
4 20 35 45 ~ ~ 1 0
- 5 1 5
6 1 5 35 - 50 - 7
7 15 25 60 - 7
8 50 35 - 1 5 - 20
Ceramic compositions according to the invention were produced by
first mixing together particulate boron nitride, particulate zirconium
diboride and, if present particulate aluminium oxide, zirconium oxide and
silicon carbide in an intensive mixer and then adding a liquid
phenol-formaldehyde resin, and mixing until the mixture of the particles
and the resin was hornogeneous.
The boron nitride was a refractory grade containing up to 7% by
weight of oxygen and had a particle size of less than 10 microns, and the
zirconium diboride had a particle size of less than 45 microns. The
aluminium oxide and zirconium oxide ~ere both 50/50 w/w of particles of
less than 500 microns and particles of less than 53 microns. The silicon
carbide had a particle size of less than 150 microns.
The resin was a liquid novolac phenol-formaldehyde resin having a
solids content of 60~/t, by weight.
The mixture ol particles and liquid resin was heated to reduce the
liquid content of the resin to render the mixture suitable for forming. The
,
CA 02260197 1998-12-29
W O 98/01405 PCTIGB97101706
mixture was then formed into test specimens by cold isostatic pressing of
the mixture in a mould. After forming the specimens were stripped from
the mould, and heated for 1 hour at 200~ C heated to cure and cross-link
the resin. Finally the test specimens were heated at 900~ C to pyrolyse the
resin and produce 'a carbon bond.
Example 2
Compositions 1, 2, 3, and 4 from Example 1 were tested to assess
their resistance to molten steel in comparison with a conventional carbon
bonded alumina-graphite material, by measuring their corrosion rate when
immersed in molten steel at 1650~ C.
Rods 50 mrn in diameter and 300 mm in length were made by
isostatic pressing L~sing the method described in Example 1, and their
diameter was accurately measured. The rods were then held in jigs, and
immersed for one hour in molten steel in an induction furnace. At the end
of the test the diameter of the rods was remeasured.
The results obtained are tabulated in Table 2 below.
Table 2
Composition No. Corrosion Rate (mm/hour)
0.3
2 0.2
O.I
4 0.6
Alumina/Graphitc 2
CA 02260197 1998-12-29
wO 98/01405 PcTtGss7/0l706
Exampie 3
Compositions l6, 7, and 8 from Example 1 were tested to assess
their resistance to rnolten slag in comparison with a carbon bonded
zirconia graphite material, by measuring their corrosion rate when
immersed in molten slag at 1580~ C.
Rods of the same dimensions as those in Example 1 were made
using the method clescribed in Example 1, and their diameter was
accurately measured. A borosilicate glass was sprinkled on to the surface
of molten steel in an induction furnace, and allowed to melt to form a slag.
The rods were then held in jigs and immersed in the molten steei for one
hour. At the end of the test the diameter of the rods was remeasured in
the area which had b~en in contact with the molten slag.
The results obtained are shown in Table 3 below.
Table 3
C'omposition No. Corrosion Rate (mm/hour)
6 2
7 2.5
0.5
Zirconia/Graphite 4
~xample 4
All eight compositions from Example 1 were tested to assess their
resistance to oxidation, by measuring their oxidation rate at 1200~ C at
various time intervals.
CA 02260197 1998-12-29
W O 98101405 PCT/GBg7/01706
Disc shaped specimens 30 mm in diameter and 10 mm high were
made by the method described in Example 1. The specimens were
weighed and plac~ d in an electric oven for various times, and then
removed, cooled and reweighed.
The results, which are expressed as weight change of the
specimens in mg/crn2/hour, are shown in Table 4 below.
Table 4
C'omposilioll No.2 Hours 26 Hours 130 Hours
0.97 - 0. 1 4 ().()00 1
2 2.77 0 37 0 00005
3 1.7~ - 0.6() - 0.00002
4 5.54 2.97 - 0.001 5
0.63 0.20 0.00001
6 15.10 1.89 0.00025
7 1 0.73 1 .69 0.0000
8 0.54 0.~9 0.0000~
As the results in Table 4 show. the rate of oxidation decreases
substantially with time. reaching virtually zero after 130 hours. This can be
explained by the phenomenon of passiYe oxidation which is inherent in the
compositions.
Fxample 5
Compositions 1 and 3 were tested in comparison with a
conventional carbon bonded alumina-graphite material to assess their
ability to suppress clogging due to alumina build up when used to form the
inside surface of a nozzle though which molten steel is cast.
CA 02260l97 l998-l2-29
W O 98/01405 PCT/GB97~17o6
Tubular nozzles having an outside diameter of 50 mm, an inside
diameter of 15 mm and a length of 300m were made using the method
described in Example 1. The nozzles were immersed in aluminium killed
steel having an aluminium content of 0.2% by weight. After immersion of
the nozzles, oxygen was bubbled into the steel and the no~les were
agitated continuously to distribute the oxygen. After 30 minutes the tests
were concluded and l:he nozzles were removed. The nozzles were then
sectioned and inspected to assess the build up of alumina.
The alumina-graphite material became badly clogged. Composition
3 showed no clogginl, and while composition 1 did show some clogging
the material was considerably better than the alumina-graphite material.
I~xample 6
Four compositions were prepared as in Table 5 below using the
method described in Example 1. The boron nitride, zirconium diboride,
aluminium oxide and ;zirconium oxide which were used were the same as
those which were used in Example 1. The titanium diboride, boron and
calcium hexaboride were powders of particle size less than 50 microns.
The magnesium oxide had a particle size of 53 to 500 microns. The
amount of each component is expressed in the same manner as in
Example 1.
CA 02260197 1998-12-29
WO 98/01405 PCTIGB97/01706
Table S
Component Composition C~omposition Composition Composition
9 10 1 1 12
B~ 40 20 10 4()
.~rB. 35 30 35 30
TiB. 15 15 10 15
B 10 lo
~I~Ol - 20 1 0
~Z~-O. - 1 ~
" - 5
- - 3 5
Resin 18 15 15 20
The compositions were tested to assess their resistance to molten
slag using the method described in F~rnple 3. and they were tested to
assess their resistance to oxidation using the method described in
Example 4.
The results obtained are shown in Table 6 below. The results of the
oxidation resistanc:e tests are expressed as weight change of the
specimens in mg/crn2lh~L
table 6
Coi"position No. corrosion Rate 2 Hours 26 Hours 130 Hours
(mm/hour)
9 0.4 0.3 0.2 o
1 o 0.8 0.4 0.3 o
11 0.7 20 4 0.01
12 1 04 0.2 o
. _ .... .
CA 02260197 1998-12-29
WO 98/01405 PCTIGB97101706
FY~mple 7
A mixture was prepared having the following composition by
weight:-
Boron nitride 20 %
ZirconilJm diboride 20 %
Zirconium dioxide 55 %
Silicon carbide 5 %
Each of the four components was as described in Example 1.
The mixture of the ceramic components was mixed with 6.5 % by
weight, based on the total weight of the four ceramic components, of a
liquid novolac phenol-formaldehyde resin having a solids content of 60 %
by weight as described in Example 1.
Ceramic test specimens in the form of rods 4 cm in diameter and 30
cm in length were then produced using the procedure described in
Example 1, and the diameter of the rods was accurately measured.
A slag containing 7 % by weight of fluoride was melted on top of
molten steel held at 1650 ~C in a 250 kg capacity high frequency induction
heating furnace.
The rods wer~e then held in jigs, and tested by immersing them in
the molten steel for ltwo hours to assess their resistance to thermal shock,
the degree of penetl alion of molten steel and slag, and the rate of
corrosion at the slag/metal interface. Similar rods made from a carbon
bonded zirconia-graphite material were tested in a similar manner. Both
2~ types of rod had adequate thermal shock resistance and resistance to
penetration, but the! rods made from the composition according to the
CA 02260197 1998-12-29
W O 98/01405 PCT/GB97~1706
13
invention was superior in terms of its rate of corrosion at the slaglmetal
interface. The carbon bonded zirconia-graphite rods had a corrosion rate
of 3.05 mm per hour at the slag line whereas the rods made form the
composition according to the invention had a corrosion rate of only 0.95
mm per hour.
Fxample 8
A mixture was prepared having the following composition by
weight:-
Boron nitride 25 %
Zirconium diboride 20 %
Alumin ium oxide 55 %
Each of the three components was as described in Example 1.
The mixture of the ceramic components was mixed with 7.5 % byweight, based on the total weight of the three ceramic components, of a
liquid novalac phenol-formaldehyde resin having a solids content of 60 %
by weight as described in Example 1.
Ceramic test specimens in the form of rods 4 cm in diameter and 30
cm in length were then produced using the procedure described in
Example 1.
The rods were then held in jigs and immersed in aluminium killed
steel containing 0.05 to 0.1 % by weight aluminium in a 250 kg capacity
high frequency induction heating furnace. The surface of the molten steel
was covered with a layer of rice husks, and in order to prevent excessive
oxidation of the steel during the test argon gas was also used to protect
the surface of the sl:eel. The temperature of the molten steel was 1570 to
1580 ~C and the immersion time was 2 hours. Similar rods made from a
carbon bonded alumina-graphite material were tested in a similar manner.
CA 02260197 1998-12-29
W O 98101405 PCT/GB97101706
14
At the end of the test the rods made from the composition according to the
invention had apprec:iably less build up of alumina on their surface than
did the rods made from the carbon-bonded alumina-graphite material.
