Note: Descriptions are shown in the official language in which they were submitted.
~LZS5~7~
T~CMNICAL FIEL~
This inven-tion relates primarily to the flexible
belts used in continuous casting machines for the castinq of
ferrous and non-ferrous metals. More particularly, this
invention is directed to protective and thermallY i'nsulatinq
~matrix coatings, the methods of formi.ng such coatings, the
composition of the coatings, and the coated belts so produced.
The casting helts are usually made of mild steel. Secondarily,
the invention applies to the coating of other molten-metal-
contacting surfaces in continuous casting machines, such as the
coating of edge-dam blocks.
BACK~,ROUND
Numerous combinations o~ oils, graPhite, soot,
diatomaceous earth, silica, organic bindexs, etc., have been
used to protect metallic casting helts or to insulate them
and/or to act as parting agents to prevent adherence to the belts
in continuous casting machines for castinq molten metal. Such
prior coatinqs are temporary or transitory in nature and may be
continually applied and replenished during castinq. The con-
tinual applicatlon of such coatings ~hile castinq requires
precise maintenance and control in view of the need for con-
sistent thermal conductivity. This continual application and
replenishing oE temporary insulative coatings is a difficult
and imprecise art. For example, excess liquid or solvent
or binder in the insulatinq coatinq material is likely to
emanate gas in such quantity as to disturb the soundness of
the cast product, resultinq in porosity.~ Some of the qas thus
liberated is at times hydrogen, which can detrimentallY alter
the metallurgical quallties of the cast metal. Also excess
--2--
~q
553!3'7~:i
amounts of the temporary insulative coating material itself
may accumu]ate near the edges of the cast product and usurp
part of the continuously~moving mold space, causing defects
in the cast product.
A two-layer be]t coating, including thermosetting
resin and solvent, for use in continuouslv casting relatively
low melting-point metals, such as aluminum, zinc and lead is
described in U. S. Patent No 3,871,905. Coatin~s containing
resins are generally unsuitable for use for continuously
casting metals having melting-point temperatures significantly
higher than aluminum.
A casting belt made of mild ~illed steel containing
0.2% to 0.8% by weight of titanium has been multiple-laver
coated, as described in U. S. Patent No. 4,298,053.
The surface of the belt lS first coated bY a "primer"
layer of a nickel-aluminum alloy (80% by wt. of Ni and 20%
by wt. of Al) statéd to be 0.005 mm thick in the
.
specification but claimed to be 0.0~5 mm thick in the only
claim. This primer layer is coated by another layer between
0.01 and 0.5 mm thick made of chromium, or of an alloy of
chromium, or of nickel, or of an alloy of nickel or of a
stainless steel. Then,a third layer of colloidal graphite
anti-adhesion agent is applied overthe second layer. However~,
in-our experience more thermal insulation and additional
non-wettability are required than can be obtained by following
the teaching of that patent.
Canadlan Patent No. 1,062,877 of Thym and Gyongyos
describes the coating o~ endless casting helts by several
thin layers (80-100 micrometers,preferably 50-70 micrometers)
~L255~
on the endless casting belts until the desired thick~ess of
ceramic layers is achieved to give the re~uisite thermal
resistance. Such a'build-up of multiple ceramic layers is
laborious, time-consuming and expensive. The resulting huilt-
up coating is machined mechanlcally, e.g. by grinding, in
order to achieve the desired uniform surface finish and
wetting behavior between this multiple-layer ceramic coating
- and the aluminum being cast. This built-up ceramic coating
consists of A1203' CaZrO3~ A1203 g 4
Tio2- It is built up in thickness until it provides thermal
resistance in the range of 10 4 to 10 3 m2.h. C/kcal.
'Such built-up ceramic coatings are usually relatively
thick and relatively fragile and brittle~ They have insufficient
durability to withstand thermal shock, or to withstand the
mechanical stretching and relaxing, the flexing and abrading
which are inherent in continuous casting employing one or
-more movlng belts as molten-metal-contacting-cooling surfaces.
Durability to withstand such mechanical and thermal
stresses are important, as otherwise bits of the ceramic coating
become loose and spall during the demanding service imposed
upon them in continuous casting of molten metals. The loosened
bits inevitably become inclusions in the cast metal product.
Such lnclusions can become a serious 'problem, as for instance
in the case of copper destined for drawing into fine wire. Such
inclusions cause the wire to break in the dies, resulting in
significant productivity losses as the wire is restrung.
Ceramic coatings are generally not flexible and tend to be
fragile.
--4--
~2S5~7~
Problems associa-ted with brittleness, cera~]c
flake~off and contamination of the cast product by ceramic
particles are highlighted in German patent 24 11 44~ oE
Theobald, in which patent an attempt was made to solve
this problem when casting alurninum by applying over the
relatively thick ceramic a second and protective abrasion
resistive metal layer which has a higher temperature point
of fusion than the metal to be cast.
SUMMARY OF T~IE DISCLOSURE
A unitary-layer partially metallic, suitably adherent,
mechanically and thermally durable, non-wetting, fusion-bonded
matrix coating on endless, flexible metallic casting belts for
continuous casting machines is described. This fusion-bonded
matrix coating is also advantageous for coatinq other molten
metal-contacting surfaces in continuous casting machines,
such as edge~dam blocks that define moving side walls of a
mold cavity. The fusion-bonded matrix (or reticulum) coating
provides advantageous accessible porosity throughout the
coating and comprises a nonmetallic refractory material
interspersed substantially uniformly throughout a matrix of
heat-resistant metal or metal alloy, for example, nickel or
nickel alloy, such metal o~r metal alloy~being usion-bonded
to a grlt-blasted surface of the belt and serving to anchor
and hold the nonmetallic material. The coating is applied
by thermally spraying a powdered mixture directly on the
roughened surface. The result~is to insulate and protect
the underlying belt from~intimate molten metal contact, from
heat sress~and consequent distortion and from chemical or
stress-corrosive action by the molten metal or its oxides or
-5-
~'
~25~i~7~
slags. The nonmetaLllc material may be present, at least
partly, in the form of isolated partlcles encased wlthin the
metallic reticulum and/or in the form of a second reticulum
intertwined with the metallic reticulum. The life of the coated
belts is dramatically increased, and the surface quality and
properties of the cast product are significantlg improved.
The coating controls and renders more uniform the rate of
freezing of the metal being cast, resulting in improved
metallurgical properties.
Formulations and a method of Eorming such coatings
by thermal spraying are descrlbed.
SUMMARY OF THE INVENTION
The invention consists of a machine for applying
an insulative and protective coating on an endless flexible
casting belt characterized by: a plurality of spaced
circular cylindrical pulley rolls having substantially
parallel axes, drive means for rotating at least one of
the pulley rolls for revolving a casting belt around the
pulley rolls, means Eor tensioning the belt revolving
around the rolls, means for steering the belt revolving
around the rolls, a thermal spray gun, and means for
supporting the gun and for traversing the gun back-and-forth
relative to the revolving belt with the gun aimed at
the belt from a predetermined stand-off distance.
BRIEF DESCRIPTION OF THE DRAWINGS
~ . . _ .
FIG. I illustrates a side view of the casting zone,
the casting belts and pulleys~ and one of the casting side
dams of a twin-belt continuous casting machine;
FIG. 2 is an enlarged cross-sectional view of the
casting space and its surrounding parts, taken along the
line 2-2 of F:tG. 1,
FIG. 3 is a perspective view of a belt coating
machine as seen from the idling end;
FIG. 4 is an enlarged perspective view of the
~S5~37~j
steerlng mechan:Lsm portion of the belt coating machine of
FIG. 3 as seen from the location 4-4 i.n FIG. 3;
FLG. 5 is a perspective view of a modification of
the machine of FIG. 3;
FIG. 6 is a perspective view, shown enlarged, of
a preferred, laterally "floating" thermal spray gun assembly
as seen looking from the working end of the belt-coating
machine. FIG. 6 illustrates an improvement with respect to
the belt-coatin,g machines shown in FIGS. 3 and 5.
6a-
~S5~
BEST MO~E 1~ CA~RYING oUrr T~lr, INV~TION
~ i-th reference to FI~S. 1 and 2, there is illustrated
the casting zone and nearby componen-ts of a twin-helt casting
machine which includes a lower casting helt 10 revolved around
pulleys 12 and 1~, which are parts associated with a lower
carriage L. Pulley 12 is located at the input or u~strearn
end of the machine, and pulley 14 is at the output or do~7nstream
end of the machine. A continuous movinq castinq mold C is
defined by and between the lower casting bel-t 10 cooperating
with a pair of spaced casting side darns 16 and l8(FIr~ 2)
and with an upper casting belt 20, as they move -together along
the casting zone C. The side dams are guided by rollers 22
They each comprise a multiplicity of slotted dam blocks 24
strung on straps 25. Seals 26 keep water from entering
between the belts so as to isolate the casting region C
from water. Stationary guides 27 serve to quide the moving
side dams. Upper casting belt 20 revolves around pulleys 28
and 30, which are parts of an upper carriage U. Finned backup
rollers 32 define the position of the belts in casting ~one
C and permit fast-moving liquid coolant to travel along the
reverse surface of each belt. Molten metal is introduced
into the machine at its upstream end as indicated by the
arrow 31 in FIG~ 1. The cast product P issues from the
downstream end.
In accordance with the present invention each of
the belts 10 and 20 is coated before being installed on the
respective belt carriages L and U. It will be understood
from FIGS. 1 and 2 that the molten-metal-contacting surface
of each belt is its outer surface, sometimes called its front
surface, w~ile its inner surface is sometimes called the
reverse surface. Such flexible casting belts 10 and 20 are
usually made from low carbon steel rolled ~to be moderately
hard and usually have a thickness in the range from 0.035 of
-7
~25~i~7~
an inch up to 0.065 of an inch, but thinner or -thic~er belts
may be used. Occasionally, for more demanding service, the
belts are made from a titanillm-containing steel, as described
in Dompas U. S. Patent No. 4,092,155, which i5 work
hardened by rolling sufficiently to become full hard.
To coat a castinq bel-t, such as belt 10 or 20,
in accordance with the invent:ion, any oily resldue on the
outer surface of the belt must first he thorouqhly removed,
as by alkali-detergent cleaning followed by wiping with
a clean solvent.
Next, the outer surface of the belt is roughened hy
grit-blasting. For example, this grit-blasting is carried
out with 20-grit aluminum oxide, applied at an air pressure
between about 40 and 100 psi (between about 300 and 700 kilo-
pascals). The size 20-grit means particles of aluminum oxide
which have passed through a screen having 20 wires per inch.
Air pressure within the lower portion of this range is used
when grit-blasting thinner belts in the lower portion of the
belt thickness range described above, since the impacts of the
grit may otherwise cause roughness on the reverse belt surface.
Air pressure within the lower portion of the range may also be
advisable when the belt is not intended to be subsequently
roller-stretcher leveled. Usually, the belt will be roller-
stretcher leveled after grit-blasting in order to control
distortlon within acceptable limits, as described below.
Roughness of the blasted surface is normally in a preferred
., .
range from 0.002 of an inch up to 0.003 of an inch (2000 to
3000 micro-inches or 52 to 76 micrometers), which range is
readily obtained, though the useful ran~e of roughness may
occasionally extend from akout 0.001 of an inch up to about
0.005 of an inch.
~2SS~7~i
Surface roughness figures as s-tated above are deter-
mined as measured by our preferred method, that of the method
of surface grinding. In this preferred method, the thickness
of a blasted belt sample is first measured by means of an
ordinary machinist's micrometer caliper. The s~mple is then
placed on the magnetic chuck oE a surface grinder, and the
roughness is carefully ground off to ~ust that level at which
the resulting ground surface aE~pears smooth. ~he belt sample
is then again measured with the micrometer caliper, and the
difference in readings is taken as the roughness. By com-
parison, the extremes of roughness of a given grit-blasted
surface as measured by a vertically measuring microscope at
400X are, in our experience, on the order of 150% of the
measured values obtained by the surface grinding and micrometer
method.
The grit-blasting process ordinarily distorts the
belt, and roller-stretcher belt leveling will usually be
required. Levellng lS done by passing the belt with reversals
in bending and ironing action through multiple closely spaced
rollers, for example, as shown and described in U. S. Patent
2,904,860 of C. W. Hazelett.
Thermal spraying i~ then utilized to a~ply the one-
coat fusion-bonded matrix protective insulative coating directly
to the grit-blasted roughened belt surface. A successful method
is to thermally spray the coating materials b~ means of a
combustion flame--an oxyacetylene flame--at a standoff dlstance
~ ~ .
of at least 5 inches (127 mm), and at a traverse speed in the
range of 30 to 50 feet (9 to~15 meters)~ per mlnute.
_ g _
~i'e25~&~;
Ox~acetylene-sprayed coatings are successful if the
material being sprayed does not burn u~ excessively in the
flame.
Oversi~e nonmetallic particles may not entirely
melt. ~oreover, oxyacetylene f].ame may not be sufficient to
retain nonmetallic particles molten for the kime required to
fuse them to other particles of the same species as fina]ly de-
posited on the belt surface. If there is a Preponderance oP
metallic particles intermixed with nonmetallics, the environ-
ment is not conducive for interfusion of the nonmetallic
constituents. Thus, in such cases, the nonmetallic material
may be present, at least partly, in the form of isolated
partic].es encased within or surrounded by the metallic reticulum.
Plasma spraying is an alternative method of thermal
spraying that uses electricity. Combustlon (oxyacetylene)
spraying is often called flame spraying. Such usaqe is apt to
be confusing in that the plasma spray is often said to utilize
a plasma flame. Both kinds of spraying may he said to utilize
a flame. It is our terminology to use the Phrase "thermal
spraying" as being inclusive of both oxyacetylene flame spraying
and eleotrically energized pIasma spraying. ~lasma spraying
as ordinarily used runs hotter than oxyacetylene spraying and
so results in less porosity.
; It is our present.belief that the higher temperatures
provided by electrically energized (plasma) sprayinq may enable
the rapid fusing of metallic and nonmetallic materials supplied
in coarse forms, such as sticks, rods or wires (as distinct
from powdered form) and therefore may enable such coarse forms
of metallic and nonmetallic materials to be employed. But
regardless of whether this belief proves true in practice, the
use of mixtures of appropriate metallic and nonmagnetic con-
stltuents as described further below is dramatically successful
in providing fusion-bonded matrix coatings with suitable
--10--
~ 2~S~
percenta~es of "accessible" porosity as descrihed hereina~ter.
In most prior applications of thermal s~raying, porosity
is avoided so far as possible. In the present invention we
have found the opposite to be true. Con-trolled porosity
characteristics in the fusion-bonded matrix coat are desirahle
and important. An appropriate :Level of controlled porosity
contributes substantially to the insulative value of the matrix
coating, while at the same time an appropriate level of porosity
enhances the desired characteristic of rlon-wettability by molten
metal. We believe that this non-wetting enhancement is due
in large part to the air retained in the pores of the porous
coating. When molten metal is introduced ad~acent to the
coated belt the air in the pores is heated and expands out
of the pores and so supplies a gaseous film between the molten
metal and the belt coating, thereby preventing the molten
metal from wetting the coated belt, during the critical initial
time when a skin of solidified metal is being formed on the
product being cast in the continuous casting process.
Equally important is the fact that controlled
porosity within the matrix coating has the virtue of acting
as a blotter or disperser for moisture picked up on the surface
of a casting belt, caused by condensation or hy stray droplets
of coolant. This blotting or dispersing of moisture prevents
blowholes, rosettes, or needles that would otherwise appear
in the surface of the cast product P adjacent to the location
of a liquid contaminant. This feature of blotting dispersion
.. .
of moisture is important, for example, in the casting of
aluminum sheet product P with a high quality surface suitable
for anodization, as opposed to lower surface quality which is
acceptable for painting.
'
--11--
~25~37~
In addition, there are two more reasons why controlled
porosity is ~esirable. One is its improvement of thermal shock
resistance. The other is its increasing oE resistance to
spalling under mechanical rough handling. sOth of these
characteristics are important in a coatinq consis-ting, on a
volume basis, largelv of ceramic material or brittle material
generally. Under thermal shock, the porositY appears to allow
internal ad~ustments to occur without relatively massive dis-
locations appearing, there being already countless tiny
dislocations present as pores, each of which we now believe
contributes minutely to a myriad of needed internal mechanical
adjustments for accommodating thermal shocks and mechanical
flexings and stretchings. Thus, controlled porosity, far
from detracting from effective strength of the matrix coating,
actually increases it.
The desired porosity appears to extend throughout
the unitary-layer, fusion bonded matrix coating. That this
porosity extends omnipresently throughout the matrix coating
is evidenced by the fact that a steel belt so coated will rust
if left moist.
- In sum, substantial but controlled porosity within the
unitary-layer, fusion-bonded matrix coatings on belts of con-
tinuous casting machines in accordance with this invention has
four advàntages that are important to the ~resent invention.
There are upper limits to the desired range o~ such omnipresent
porosity. The upper limit in a given formulation is reached
when the integrity of the coating becomes impaired. In those
matrix coatings where ~he metallic constituents are predominant
(as determined by weight), this upper limit is at least about
35 percent "accessible" porosity by volume. In those matrix
-12-
~255~37~
coatings where the nonmetallic constituents are predominant
(as determined by weight) this upper limit is ahout 12 to 20
"accessible" porosity by volume.
There is a lower limit to the desired ranqe of
"accessible" porosity by volume in the ma-trix coating, hecause
insufficient porosity will not yield the four advantages
described above. This lower limit is about 4 to 8%.
As described below, tests and measurements were
made of "accessible" void space, i.e. effective porosity, as a
percentage of the volume of the matrix coating. These tests
and measurements were conducted to give a better understanding
of the parameters contributing to the desired porosity. Sam~les,
usually of about 14 square inches of mlld steel belt stock,
were thermally spraye~d to a thickness usually of about 0.050
inch (1.3 mm). They were thermally dried and then weighed.
Then they were soaked briefly in water with detergent (Kodak
Photo-Flo) added; then they were withdrawn and all unabsorbed
water was wiped off. The specimen was weighed again, the
increase in grams noted and divided bv the coating volume in
cubic centimeters to obtain the percentage of void space
that was accessible to water, which had become blotted or ab-
sorbed within the coating~ In a given samPle there may be
other voids that are closed ~nd so not measurable by this water-
absorption method, but we believe that those 'laccessible'l
voids which emit gas on heating and which absorb stray water
are the more imPortant voids with respect to overall advantageous
performance of the matrix coating during casting. ~ence, a
method of measuring effective porosity which takes into account
only fluid-accessible or, specifically, water-accessible
porosity is especially suitable to our purposes.
-13-
.
~zss~
Ta~le A helow lists the water-accessible porosities
as a percentage of the total volume of the matrix coating
which were observed by measuring various test sam~les thermal
spray coated with powdered mixtures of the listed ~ormulations
under the conditions stated.
TABLE A
WATER-ACCESSIBLE POROSITIES OF VARIOUS
THER~LY SPRAYED COATINGS
Oxyacetylene flame sprayed, except as noted.
Standoff distance of 5 inches, except as noted.
Traverse speed approximately 40 feet per
minute.
Composition ~ by weight: Accessihle Porosity:
a) 56 Ni - 19 graphite - 25 ZrO2~ 32 percent
b) Same, sprayed at 10 inches 34 percent
c) 75 Ni-25 C (graphite) pellets 30 percent
d) 85 Ni-15 C (graphite) pellets 14 percent .
e) 87 Ni - 8 Al - 5 Mo ~8 percent
f) 60 Ni - 6 Al-4 Mc - 30 ZrO2 4 percent
g) 72 Ni - 13 graphite - 15 ZrO2 14 percent
h) Zirconia 12 percent
,
i) Zirconia, plasma sprayed less than 2 percent
COATING~COMPOSITION
-
The pre~erred unitary-layer, fusion-honded, pro-
tective matrix coating is of the same composition throughout
its~thickness. ~This matrix coatlnq comprises a nonmetallic
refractory material interspersed substantiallv~uniformly
throughout a matrix of heat-resistant metallic component or
constituent. This metal].ic constituent is a meta~ or a metal
alloy, and it must exhibit five critical properties, as
~14-
.
~:~S5~7~i;
follows:
1) The metallic constituen-t must have heat resistance
and resistance to thermal cycling. In other words, the
metallic constituent must have a sufficiently hiqh meltinq point
relative to the tempera-ture of -the rnolten metal beinq cast tha-t
the metallic constituent resists unclue degredation during the
lifetime of the belt in continuous casting and also must resist
undue deterioration due to the extreme and repeated thermal
cycling which occurs during continuous casting. The melting
point of the metallic constituent must be at least close to,
but not necessarily above, the temperature at whIch the molten
metal enters the continuous casting machine.
2) The metallic constituent must have thermal
fusion bonding compatibility with the flexible steel casting
belts normally used to which the matrix coa~ing is fusion-
bonded.
3) The metallic constituent must have at least a
modicum of ductility in order to withstand the mechanical
rough handling to which the matrix~coated belt is subjected
during continuous casting. The moving belt is repeatedly
flexed around pulley rolls and straightened out, and in addition
the moving belt is subjected to a relatively high tension
stress during use.
4) The metallic constituent must have thermal
expansion rates that are not too far~different from the thermal
expansion rates of the nonmetallic constituents included in the
matrix coating to withstand repeated extreme thermal cycling
occurring during continuous casting without fla~-es spallinq off.
5) The metallic constituent must have sufficient
resistance to oxidation under the conditions of thermal spraying
'
-15-
~; 2S5~37~;
and also under the conditions of continuous casting so as to
avoid undue deleterious oxidation.
We have found that nickel and nickel alloys are
especially suitable for form:ing the metallic constituents of
the matrix coatings of this invention. Cobalt, iron and
titanium would also appear to have the hereinbefore described
fi~e critical properties so as to be useful as the metal or
metal alloy for forming the metallic constituents of the matrix
coatings. Those skilled in the art may find that other metals
or metal alloys are also suitably operable.
The matrix coating of this invention is formed by
thermal spraying of the metallic and nonmetallic constituents
mixed in formulations within the following ranges:
Metallic constituents: about 38 to about 90 percent
by weight,
Nonmetallic constituents: about 62 to about 10 percent
by weight.
Our observations have led us to conclude that there
must be a sufficient volume of the metallic constituents present
in the matrix coating relative to the nonmetallic constituents
so as to form an integral, fused network, reticulum or matrix
of the metallic constituents for suitably holding or anchoring
the nonmetallic constituents to the belt. Nonmetallic consti-
tuents, when present in the l~pper portion of the above range,
may also form a network or reticulum entwined (intertwined)
throughout the metallic reticulum. Nonmetallic constituents,
when present in the lower portion of the above range, may be
present, at least partly, in the form of isolated particles
encased within or surrounded by the metallic reticulum. Thus,
the metallic component forms the anchoring and holding matrix
or reticulum, while the nonmetallic component is distributed
-16-
~25S~37S
uni~ormly throughout this reticulum either as a second reticulur~
or as discrete particles. The metallic constituent qenerally
has a specific gravity averaging about one and one-half to about
four times that oE the nonmetallic. Thus, when both constituents
are present in the coating at 50% by weiqht, the volume ratio
of nonmetallic Particles to metallic particles is about 2.5 to
1 in our usual formulations. On the other hand, when the
metallic constituent comprises 85~ by weight of the coating
composition, then the volume ratio of nonmetallics to metallics
is about 1 to 2.5.
Presently preferred compositions utili~e at least
as part of the metallic and nonmetallic constituents a comnosite
nickel and graphite powder in which grains of nickel en-
capsulate graphite powder, the graphite comprising either about
15 or about 25 percent of the combined weight. Such com~osite
nickel and graphite owders are available commerciall~, for
example from Bay State Abrasives of Westborough,.~assachusetts.
Preliminary tests in the Pouring of mild (1010)
steel melting at about 1530 C (2786 F) onto steel ca.sting
belt samples having fusion-bonded matrix coatin~s in accord
with the present invention have shown that commercially avail-
able, predominantly nickel alloy containing a~out 8 ~ercent
of aluminum and about 5 percênt of molybdenum is a suitable
metallic alloy for use with powdered zirconia or graphite as
suitable nonmetalllc constituents for forming a durable matrix
coating
for example, silica or alumina
.ther metals, alloys, or nonmetallic refractoriesi
could be suitable as constituents in the fusion-bonded,
accessible-porosity, matrix coating provided by.the present
-17-
3~25S~3'7S
invention. The critical proPerties to he looked for in metals
and alloys are set forth explicitly in greater detail above.
They have suitable heat resistance and resistance to thermal
cycling, bonding compatibility wi-th low-carbon steel belts, a
modicum of ductility, thermal ex~ansion rates that are not
too far different from those o~ the nonmetallic consti-tuents
included, and oxidation resistance if oxyacetylene flame s~raying
is to be the method of application.
A presently prefexred insulati.ve materlal for use at
least as part of the nonmetallic constituent is zirconium
oxide, ZrO2, also called zirconia, which is used in Powdered
form, preferably o~ particle size running from 0.0005 to 0.0014
of an inch (12 to 36 micrometers). This zirconia nonmetallic
constituent has the advantage that its coefficient of expansion
more closely approximates that of steel and nickel than some
other available metallic oxides which have a lower coefficient
of expansion.
~ Yttria (yttrium oxide, Y2O3) added in any of various
amounts up to about 20 percent may be helpful in stabilizing
the structure of the zirconia crystals exposed to high
temperatures, thus preventing premature loosening of the
crystalline particles due to subtle changes in mechanical~
propo~rtions during thermal cycling. Other metallic oxides
may also be used for this heat-stabilizing purpose, notably
magnesia (MgO) and lime (CaO). The latter is economical and
has afforded acceptable results in our experience. Thus,
economical lime (calcium oxide) is presently preferred as
a heat-stabilizing compound. It is normally an ingredient in
purchased zirconia, comprising about 4 to 5 percent by weight
of the zirconia.
~ -18-
~ZSSI~g7S
The particles or powder o~ the nonmetallic comnonent
are thorouqhly mixed and blended with the powdered metallic
component, and the resulting mixture is thermal-s~rayed directlY
onto the grit-blasted surEace of the belt. ~egregation of the
mixed powders durinq application must be avoided.
As discussed earlier, coatinqs of zirconia alone
or of other nonmetallic substances alone may under certain
adverse conditions lose adhesion and release ~its o the
nonmetallic substance into the freezing metal product. This
flake-off problem has been minimized or avoided in the matrix
coatings of this invention by attention to the following factors.
The zirconia powder is preferably of fine particle size,
sufficiently fine to pass through a screen havinq 300 or more
wires per inch. There should be enough metallic constituents
in the powder mix to form on the belt an integral, fused-
together network or reticulum that will securely anchor and
hold the zirconia particles in a relatively discrete and dis-
continuous array and/or in a second reticulum ~hich is intert~ined
with the metallic reticulum as described above.
Additionally, the finished unitary-layer, Fusion-
bonded, matrix coating should be brushed and dusted or vacuum
cleaned before use.
- Graphite is a highly heat-resistant separating agent
which sublimes at about 3700 C. without melting. It is a
useful nonmetallic constituent for the reason that it is
non-wetting with respect to nearlv all molten metals. ~ore-
over, should particles of graphite get into the metallic pro-
duct, its softness, friability, lubricity, and inertness
forestall most of the problems associated with the incidental
inclusioll of foreign substances. Under the pressure of
rolllng or drawing, graphite particies break or divide into
progressively finer particles.
-19-
~zss~s
Our experience has shown that when suitable ~owdered
metallics and powdered nonmetallics are thorouqhlY mixed and
blended together, some of the resultinq mixtures (particularly
those containinq very fine particles) are apt not to flow
freely and uniformly through the passages of a thermal-spray
gun. The result is uneven coat:ing. For producing a free-
flowing powder blend in many cases, an addition to the Powder
blend of at least about 0.25 percent by weight o~ spherical
fumed silica (SiO2) particles as a lubricant has substantiallY
enhanced flowing of the powder mixture and uniformity of
thermal spray coating. The amount of this fumed silica
lubricant is not critical, and good results have been obtained
with most powder mixtures. A grade of 0.014 micro meter
(14 millimicrons) fumed silica particles has been successful
for producing a free-flowing powder blend. This size of
0.014 micro meter is less than a millionth of an inch and lS
a nominal size.
~20-
~:~S5~37~ii
Examples of suitable formulatlons for forrning the
matrix coatings of this invention are set forth below.
EXAMPLE I
Constituent: Welght percent:
Metallic component:
Aluminum 4 to 5
Molybdenum 2 to 3
Nickel, plus trace impurities 55 to 57.5
Nonmetallic component:
Zirconia 35
Calcium oxide, in the zirconia 1.5 to 2
100%
XAMPLE II
Constituent: Weight percent:
Metallic component:
` Aluminum 6
Molybde~um 4
Nickel, plus trace impurities 52
Nonmetallic component:
Zirconia 22 to 23
Calcium oxide, in tpe zirconia 1 to 1.5
Graphite 13 to 14.8
; Spherical fumed silica 0.2 to 0.5
100%
: ~ :
::
-21-
~;~S5~7~
EXAMPLE I I I
Constituent: Weight percent:
Metallic component:
Nickel, plus trace impurities~ 57 to 60
Nonmetallic component:
Zirconia 25 to 28.8
Calcium oxide, in the zi:rconia 1 to 1.5
Graphite 13 to 15
Spherical fumed silica 0.2 to 0.5
100%
EXAMPLE IV
Constituent: Weight ~ercent:
Metallic component:
Chromium 14
Nickel, plus trace impurities 54
Nonmetallic component:
Zirconia 29.8 to 30.4
Calcium oxide, in the zirconia 1.4 to 1.6
Spherical fumed silica 0.2 to 0.6
100%
EXAM~3LE V-VIII
; Similar formulations for forming matrix coatings
; of this invention may be obtained by substituting cobalt
partially or fully for a corresponding weight percent of nickel
in the forego~ng ~our Examples.
~: :
--22--
~s~
EXAMPLE IX
Component: Weight percent:
Metallic component: 38 to 90
Nonmetallic component 62 to 10
100%
Constituents of metallic component: Weiyht percent:
Aluminum 0 to 35
Nickel, plus trace impurities balance
Constituents of nonmetallic component: Weight percent:
Graphite
0 to 40
Spherical fumed silica 0.3 to 0.8
Lime 4 to 20 of the sum
of Zirconia plus Lime
Zirconia balance
.---
100%
In this ExampIe IX, the weiqht percent of
aluminum is shown in the range 0 to 35, but the upper end of
this range is subject to the limitation that the ratio of
aluminum to nickel does not signi-ficantly exceed a one:-to-one
atomic ratio. Slnce the ratio of the atomic weight of alumi-
num to that of nickel is about 41%, the weight percent of
aluminum in the above Example does not significantly exceed
about 41% of the weight percent of nickèl in this formulation.
.
: : :
~: -23-
:
~551371~
XAMPLES X-XVII
Magnesium zirconate can be substituted partially
or fully for both a corresponding weight percent of zirconia
and its proportionate weight percent of the heat-stabilizing
agent Calcium Oxide in each of the foregoing Examples I-IX.
EXAMPLES XIX-XXII
Formulations a, d, e and f of Table A above, each
modified to include at least about 0.25~ by weight of spherical
funed silica as a lubricant, are further Examples suitable for
forming fusion-bonded matrix coatings on flexible casting
belts.
The preferred minimum deposited thickness of the
fusion-bonded matrix protective insulating coating for use on
flexible metal continuous casting belts 10, 20 (FI5. 2) is'about
0.002 inch (0.05 mm), said minimum measurement being the thick-
ness o~er the generality of the peaks of the underlying grit-
blasted belt surface, which is the way a magnetic thickness
gauge normally measures. However, advantages may be obtained
by using matrix coatings as~thin as about 0.0015 inch (0.038 mm).
: Thermally sprayed coatings even thinner than 0.002
of an inch (0.05 mm~ appear to be useful .in some applications
where nonwetting is more important than thermal insulation.
Thus, a lower practical limit to thickness is not readily
apparent. For extra insulation, thicknesses of several times
this amoant of 0.002 of an inch will on o.ccasion be useful,
since the coatlng which is the subject of the present invention
is rugged and can withstand much flexing around the,pulleys
-24-
~Z~i5~37~
(rolls) of a continuous casting machine. But, depending on
the casting application, more thickness is not necessarily
better, not on flexible belts and especially not in uses where
coating-loss impurities could seriously interfere with the
~quality of the cast product, as in the continuous casting of
copper wire bar intended for fi.ne wire drawing. Thicknesses
as great as 0.015 of an inch (0.4 mm) are readily produced
and are rugged. However, the expense of such thick coatings
is also a limiting factor.
The accuracy with which insulation can be applied
and controlled with these thermally sprayed fusion-bonded
matrix belt coatings is not only a desirable feature in
itself but, further, it enables planned proportioning of in-
sulation between belts 10, 20 and edge dams 16, 18. That is,
it enables the attainment of optimum comparative heat flux
density through the belts 10, 20 as compared to heat flux
into the edge dams 16, 18. The accurate proportioning of the
density of heat flux between the broad belt surfaces on the
one hand, and the relatively narrower moving edge dams on the
otherl is of importance in producing cast slab of first-class
metallurgical quality where the thickness is greater than 1/4
of an inch (6 mm); see U.S. Patent Application, Serial No.
493,359, flled May 10, 1983, now U. S. Patent No.
4,545,423. The theory therein may
explain the importance of proportioning the density of heat
flux between the wide belt mold surfaces and the narrow edge
dam mold surfaces.
To achieve such relative proportionlng of heat
extraction(heat flux), one may adjust the thickness of
coatings on the belts as compared to that same coating composi-
-25-
tion on the hlocks of the edge dams. Meta]s are usually better
thermal conductors than non-metals; hence the ratio of metal to
non-metal in the ~usion-bonded matrix coating may be adjusted to
control conduc-tivity. For example, the thermal conductivity of
nichrome (80~ Ni, 20% Cr by wt.) is on the order of about ten
times that of zirconia. Again, the metallic constituents them-
selves in the matrix coatings can be selected according to thermal
conduc-tivity or`insulative value, and adjusting the content of
metals of relatively low thermal conductivity to the content of
metals of higher thermal conductivity. The conductivity of ni-
chrome is on the order of about one-fourth that of nickel or of
some low alloys of nickel.
The present invention may be applied to edge-dam
blocks in themselves, in order to achieve advantages generally
similar to those attained with belts. However, in accoxdance
with the ab~ve-noted patent application relating to the
insulation of edge-dam blocks, more insulation will generally be
required on the edge-darn blocks than on the adjacent casting
belts. This difference will ordinarily be achieved through apply-
ing a greater thickness of thermally-sprayed, fusion-bonded
matrix coating insulating material, though composition ratios for
adjusting and proportioning heat flux may be used.
MACI~E FOR FORMING FUSION-BONDED
~_MATRIX COATINGS ON BELTS
A machine for e~ploying the method for applying
the coatlngs is illustrated in FIGS. 3 and 4. Two circular,
cylindrical pulleys or rolls 34 and 36 have parallel horizontal
axes. These parallel axes lie in the same horizontal plane,
as a matter of convenience. The idler pulley 34 is mounted on
its supporting pedestal 38 which is movable on wheels 40, 41
rolling on rails 42 and 44, to adjust for belts of differing
lengths. The rail 42 is a steel arigle with the legs downward,
forming an inverted Y, and the wheels 40 have peripheral grooves
engaging the ridge of this rail. Rail 44 is a flat bar. The rails are m~unted
-26-
~255~
on bed structures 46. The rails are long eno~lgh -to accomodate
the longest belt which is to be coated.
The belt 10 to be coated is placed around these
pulleys and tension is applied. The tension is exerted by a
double-acting hydraulic cylinder 48, in line with a suitable
~rigid tubular spacer 50. This spacer 50 is removed and re-
placed with a longer or shorter spacer depending on each range
of belt length to be coated. The cylinder 48 ls mounted on the
horizontal longitudinal centerline between the pulleys 34,
36 in order to avoid substantial turning torque on the idler
pedestal and its supporting rails. This cylinder 48 is mounted
on a rigid arm 58 projecting from a pedestal 52. The tubular
spacer 50 is mounted on a similar rigid arm (not seen) pro-
jecting from the pedestal 38.
Since one side of the machine must be open for
belt mountlng and removal, the pulleys 34, 36 are canti-
levered from pedestals 38 and 52, by means of two bearings
54 and 56 on each pulley shaft 57 to absorb the overhung load.
We use a tension of roughly 2200 pounds tlOOO kilograms) per
reach of belt (upper and lower reaches), making a total force
of 4400 pounds (2000 kilograms), though this tension force
is not a critical factor, since the purposes of the tension
are simply (l) to enable the driving and steering of the belt
and (2) to force the belt to come close to the pulley 36 at
the working end of the machine in order that the belt may be
cooled where the thermal-spraying flame is to impinge on it.
The side of the machine where the belts are inserted and
removed is called the l'outboardll side, and the side near the
pedestals 38 and 52 is called the l'inboard" side.
.
-~7-
~255~37S
A four-way hydraulic valve 60 controls the tension.
Hydraulic-oil under pressure comes from a pump 62. A limit
switch 64 with an upstanding probe 65 senses the edge of the
belt and causes a buzzer to sound a warninq in the event that
the belt creeps too far inboard, i.e., too close to the
~pedestal.
The belt 10 is ordinarily revolved relatively
fast, while the traverse of a thermal-spray gun 66 is slow,
resulting in a pattern of deposit path not unlike that of a
screw thread or helix, with overlapping borders of the path.
This helical application path is the preferred method, since
the starting and stopping of the application can thus ad-
vantageously take~place in the margins of the casting belt,
vutside of the casting area, where the location and effects
of starting and stopping are not critical. The pulley 36
that supports the belt is revolved by means of a variable-
speed drive (not shown) inside of the p~destal 52 at the workingend, at a predetermined peripheral speed ordinarily between
30 to 50 feet (9 to 15 meters) per mlnute for oxyacetylene
thermal spraying, and at lO0 feet (30 meters), approximately,
per minute for plasma thermal spraying. However r speeds well
outside these suggested ranges may be suitable under some
circumstances. For example~ the thermal-spray gun 66 can
conceivably be run back and forth rapidly across the belt like
a shuttle, while the belt is rotated slowly or, preferably,
the belt is stepped ahead with each pass of the "shuttle."
But the attainment of uni~orm coating around the belt at the
places of starting and stopping is not readily achievable by
this shuttle method.
The presently preferred method involving relative-
ly fast revolution of the belt as described above is apt to
-28-
~ZSS137S~
result in the belt creeping inboard or outboard on the pulleys,
Unless suitable adjustments or guides are available. The pre-
sently preferred adjustment for counteracting belt creep is
that of skewing the cantilevered idler pulley 34 in a vertical
plane, causing its axis 68 ko be inclined a trifle upward or
downward, within a plane perpendicular to the straight reaches
of the belt. The mechanics of roll-skewing steering
have been described in U. S. Patent
No. 3,123,874. For roll-skewing steering, the hand
adjusting screw 70 is arranged to shift one bearing 54 up-
wardly or downwardly slightl~ for tilting the axis as needed
to keep the belt from unduly creeping either wav.
The details of the mounting o the idler pulley
34 are as follows. Inside the idler-pedestal housing 38, the
moment from the tension of the belt on the cantilevered
pulley 34 is absorbed from the pulley shaft 57 by the two
self-aligning pulley shaft bearings 54 and 56. Bearing 54,
the one nearest to the viewer of FIG. 4,is housed in a rec-
tangular block 74 which is able to slide up and down between
gibs 76. The weight of the cantilevered-pulley 34 pivoting
relative to bearing 56 fixed in a block 77 tends to raise the
movable block 74, but the aforesaid hand adjusting screw
70, threaded into yoke 78, limits the rise of the block 74.
Hardened wear plate 80 on top of the block 74 prevents galling
at the end of the adjusting screw 70. The gibs 76 are mounted
on pedestal frame plates 85 and 87 which are welded to the
block 77. These plates 85 and 87 are also welded to the yoke
78 and to a palr of angle members 89. Alignment of the pulley
axis 68 in a horizontal plane, i.e., in a plane parallel to
the straight reaches of the belt is achieved through four
--29--
~ ;25iS~7'j
adjustinc~ screws (only three are seen) 81, 82, 83 thre~ded
into solidly fixed angles 84, whlch in turn are anchored to
a base plate 86, which is part oE the pedestal 38. The angle
members 89 on the plates 85 and 87 are adjustably secured to
the base plate 86 by stud assemblies 91 including studs welded
to the base plate and extending up through elongated slots in
the flange of the angle member 89, with a washer and nut on each
`stud.
The thermal-spraying gun 66 is mounted to a;,m at the
belt 10 where the be].t is passing around and is in contact with
the pulley 36 at the working end of the machine. This pulley is
cooled, which cooling is arranged by running water through it by
means of axially mounted connections 88 (only one is seen) and a
hose line 93. It may be expedient to cool both pulleys, but so
far the idler pulley 34 has not been cooled. Cooling the work-
ing pulley 36 in this way keeps the pulley and also the belt from
overheating.
Water which is cold and which is supplied in too
large a flow rate will keep the working pulley 36 too cold,
resulting in condensation of atmospheric moisture as water on
the belt. Such condensation interferes with the adherence
of the sprayed material and must be avoided at all times. It
is helpful to allow the cooling water to flow through the hose
line 93 only when the pulley power is on. This control of water
flow to o,ccur only when the belt lS revolving is arranged by
placing a solenoid-controlled valve (no-t shown) in the line
93 which supplies the pulley-cooling water. This solenoid-
.
~ control~led valve is energized ~rom the same switch which
.
energizes the pulley drive.
The thermal-spray gun 66 is made to traverse some
or most of the width of a belt by means of a nut 90 engaging a
lead screw 92 which is turned at a predetermined speed by an
adjustable speed drive 94. This assembly 92, 94 is suspended
from an upright rack 96, and the travelling nut 90 is guided
by a carriage 98 travelling along a guideway 99, such as a
-30-
~;~SS~37~
guide bar. I'he preferred speed of traverse depends on the
width of the spray which can be laid down in one pass, together
with the speed of travel of the belt and the length of the belt
as measured once around the loop. Naturally, a longer belt
will take longer to pass once around the pulleys and so will
~require a slower traverse of the thermal spray yun 66 than a
shorter belt. A typical range of traverse speed per belt
revolution is 3/4 to 1 1/4 inches (38 to 63 mm) per belt re-
volution, but a wide range of available traverse speeds should
be provided for the gun 66. For instance, if the above-
mentioned plan of making a kind of "shuttle" of the Elame-spray
gun were to he adopted, traverse speeds of many feet per minute
would be required. For such reasons, no hard and fast limits
to the speeds of either belt revolution or gun traverse can
be laid down.
The dust from overspray is efficiently collected.
Incorporated within the machine are exhausting and washing
equipment to catch this dust. Through a hood 100, which
extends along near the work pulley 36 above the full length
; ~ of the traverse of the gun 66, the air containing the ove~prayed
dust is sucked away by a suction blower driven by a ~,otor 102. This
air is blown through perfora~ed, continuously wetted metal
baffles located in a housing 104. The holes in these wetted
baffles are as small as 1/16 of an inch (1.5 mm). The
filtered and washed air is finally exhausted through a vent
duot 106. The hood 100 has a lip 107 projecting down
beyond the crown of the work roll 36. This downwardly pro-
jecting hood lip 107 is at a level just a few inches above
the top of the housing 105 of the traversing thermal spray gun 66.
At the point of thermal-spray impingement, the belt
may expand due to heating and bulge enough to lift away Erom
-31-
;S~7~
the pullcy 36, thus resulting in localized loss of contact
between the belt and the cooled pulley 3~. Such loss of
cooling contact can result in localized over-heating of the
belt.
An alternate method of cooling the belt is to sheath
the periphery of the pulley 36 with a continually moistened jacket
of moderately heat-xesistant material, preferably somewhat
resilient, such as a mat of silicone rubber or fiberglass mat-
ting, or a combination of such mat and matting. The objective
is to present to the reverse side of the casting be:Lt a textured
or porous surface that will retain a controlled amoun-t of cool-
ing water or aqueous cooling liquid. A film of moisture 50
deposited on the reverse belt surface will cool and protect the
belt from overheating. We believe that this cooling effect
results largely from the water acting as a heat-transfer medium
between the belt and the water-cooled pulley 36.
A presently preferred method of supplementing the
belt-co~o11ng by the cooled work pulley 36 is to use a wetted, mop-
like cloth or fibrous mass 109 in contact with the belt and
having a width about equal to the belt width. This wetted porous
fibrous wiping mass 109 is placed in contact with the lower reach
of the belt where thé belt is approaching the work pulley 36,
which is unsheathed steel. ~he direction of belt travel and
pulley rotation are shown by the arrows 111 and 115 in FIGS. 3
and 4. Such fibrous belt-cooling devices 109 are continually
moistened as needed in order to prevent the belt from reaching a
temperature in excess of 450F.
An alternate belt coating machine, a four-pulley
machine, is shown in FIG. 5. This modification employs two idler
pulleys 108, 110 extending from a pedestal 38' and a pair of work
pulleys 112, 114 extending from a pedestal 52'. At least one of
-32-
~;~55~S
the pulleys 112, 11~ is a drive pulley. This four-
pulley machine may allow more reliahle cooling of the helt
at the point of thermal-spray impingement, since the coating
is applied, not where the belt is in contact with a nulley,
but on a flat portion of the casting belt accessible to other
means of cooling from the reverse side, to a coolant, such
as an aqueous liquid. Excess cooling, as by application of
copious quantities of cold water, is not desirable, as there
results condensation of atmospheric moisture on the side o
the belt being flame-sprayed. Such condensed moisture inter-
feres with adherence of the coating. Also, the disposition
of excess water will be a problem, if sizeable quantities are
used. Water cannot be allowed to contact the side of the belt
being thermally sprayed.
For these reasons the water or aqueous liquid is
preferably applied by a nozzle 116 making fine spray, or by a
porous wiping device such as a wad or "muff" of fibrous,
moderately heat-resistant material. Such a fine spray nozzle
116 or porous wiping mass or muff is acting upon the reverse
surface of the belt preferably over a limited area being moved
by a carriage 117 that is made to travel parallel to and in
opposed aligned relationship with the thermal-spray gun 66 by
means of a second screw 118, so as always to be opposite to
the gun 66. The spray from the nozzle 116~should not be so
fine as to create mist, unless a second suction hood is pro-
vided to prevent the mist from wandering to the front side
of the belt. The carriage 117 for the nozzle 116 is mounted
on a nut 120 which rides on screw 118, which in turn is driven
by a chain sprocket 122, driven from the other screw shaft 92
-33-
~255t~7~
for the gun 66 and synchronized with it so tha-t the cooling
means 116 always stays opposite to the traversing gun 66. A
forked guide 124 sliding along a frame member 126 keeps the
nozzle carriage 117 from rotating.
In this machine, four pulleys, not three, are
~generally necessary in order to provide a uniform belt steering
effect at both ends of the machine. The steering method is
similar in principle to that described in U. S. Patent
3,310,849~
Inviting attention back to FIGURE 3, we have found
that the use of the wet, porous, wiping cooling mass ].09 is
very successful in avoiding any overheating of the helt. The
work pulley 36 has a bare steel surface and is moderately cooled
by a flow of water through the line 93 and connection 88.
In additon, a film~of water is applied to the inner belt
suxface by the wet, porous, fibrous mass 109. In order to
achieve this thin, nicely spread water film on the reverse
belt surface, liquid detergent is added as needed to the porous
mass 109.
This system and method of using the wet, porous,
belt-wiping mass 109 plus the moderately cooled bare work
pulley 36 has recently been found to operate so successfully,
that at present we believe this is the optimum arrangement.
The adjustable speed drive 94 includes an electric
motor 128 (as seen most clearly in FIG. 6) driving an adjust-
able speed and reversible mechanical transmission 130, for
example, such as a cone drive. A handle 132 is used to
adjust the speed of the output and also to reverse the direction
of the output drive. A dial 134 shows the adjusted speed and
direction. This mechanical transmission 130 includes right-angle
gearing, and the output from this transmission is a`sprocket
-34-
~255~37~
and chain drive 136 located within a protective housiny and
serving to drlve a sprocket secured to the end of the leadscrew
92. Thus, the speed and direction of the leadscrew 92 can
be adjusted by means of the handle 132. After an adjustment
has been made, the leadscrew 92 turns constantly at the adjusted
speed in the adjusted direction, until another adjustment is
made. Such adjustable,reversible drives 94 are commecially
available, for example, from Graham Company, oE Milwaukee,
Wisconsin. In FIGS. 3 and 5, this drive 94 is shown mounted on
a shelf 138 secured to an upright leg 140 of the stationary
rack 96 which has a base frame 142.
The metallic and non-metallic powder constituents
are thoroughly mixed by agitating impeller elements in a closed
container with a removable cover, for example, the cover may be
a screw-onj or latchable, top. The objective is to obtain
thorough and uniform mixing and to prevent subsequent segregation
before the mixture is fed into the powder feed passage leading
to the nozzle of the thermal spraying gun 66. In FIGS. 3 and 5,
this thermal spraying gun 66 is shown as an oxyacetylene flame
gun to which the oxygen and acetylene are supplied through a
pair of hose lines 144 and 146, respectively. The oxygen and
acetylene are mixed within the gun 66 and are fed to an annular
nozzle 148 having multiple orl~ices arranged around a forwardly
aimed central axial outlet, with the powder mixture to ~e
sprayed issuing from this axial outlet.
One way in which the powder may be fed to the gun
66 is to mount a hopper (not shown) onto the top of the gun
housing 105. ~This hopper includes a closable cover and contains
electrical motor-driven mixing agitator impeller elements for
maintaining the powder -thoroughly and uniformly mixed. The
-35-
~l25~ l7~i
hopper walls may also be vibrated by an electrically energized
vibrator for preventing "bridging" or compacting of the powder
mixture within the hopper. A metering escapement mechanism
serves to meter the flow of the powder mixture down from the
bottom outlet of the hopper into the powdcr feed passage lead-
ing -to the nozzle 148 of the gun 66, for example, this metering
escapement may comprise a feed screw having an adjustable speed
drive.
The presently preferred way in which the powder
mixture is fed to the gun 66 is to use a remotely located
powder mixing and feed apparatus, as shown at 150 in FIGS. 3 and
5. This apparatus includes a control console 152 and a container
154 which is loaded with the powder mixture by removing a screw-
on cover 156. The powder composition is thoroughly mixed
before loading into the container 154, and it is agitated and
vibrated therein to prevent stratification, segregation, compact-
ing or "bridging" within this container. The interior of this
container 154 is adjustably pressurized by an inert gas, for
example,~ such as nitrogen, with the container pressure being
shown by a dial 158 on the console 152. Such pressure serves
to propel the powder mixture toward an outlet from the contain-
er. This outlet communicates with a powder feed hose line 160
connected with the gun 66 and communicating with the axial
passage leading to the central outlet of the nozzle 1~8. In-
creasing the container pressurization as shown by the dial 158
increases the powder feed rate, i.e. the quantity of powder
mixture per minute being fed to~ the container outlet leading
into the hose line 160. Conversely, decreasing the container
pressurization decreases the powder feed rate.
-36-
~:25S137~i
The powder mlxture velocity -through the powder
feed hose line 160 is controllable separately from the feed
rate, and is indicated by a gas flow meter lSg.
This gas flow meter 159 indicates the
velocity of the inert gas flowing through the powder feed line
160. This inert gas flow fluidizes the powder mixture
adjacent to the container outlet and conveys the fluidized
powder mixture through the line 160 to the central axial outlet
in the nozzle 148 of the gun 66.
The oxygen and acetylene supply tanks ~not shown)
each has a conventional shut off valve. There is a manually
adjustable flow meter downstream from each shut off valve for
independently adjusting the rate of feed of oxygen and acetylene
through the respective lines 144 and 146. A manually operated
valve at the gun 66 simultaneously turns "on" or "off" the
flows through both of these lines 144 and 146. The gun 66 is
manually ignited by a spark striker.
An electric switch at the gun 66 is connected'
through an electrical cable 162 wlth the control console~for
turning the mixing, and feed apparatus 150 "on" or "off", as
desired by the operator, who may be standing somewhat to the
rear of the gun housing 105. Thus, the operator may turn
"on" and lgnite the-gun. Then, when desired, the operator
actuates the electric switch for causing the mixing and feed
apparatus 150 to feed the powder mixture to the gun. An
example of a suitable oxyacetylene flame spraying gun 66 and
mixing and feed apparatus 150 is equipment which can be obtained
commercially from Eutectic-Castolin Company, of Flushing,
New York, under their trade name designation TERODYN System
3000. Another example of a suitable oxyacetylene flame spray-
-37-
~;~5i5~
ing gun 66 and mixing and feed apparatus is an oxyacetylene
flame spraying gun, as described previously, with hopper rnixing
and feeding apparatus mounted directly upon the gun housing
105. Such a hopper needs to be reloaded with the powder mix-
ture at about ten-minute intervals during operation; whereas,
the container 154 only needs to be reloaded with the powder
mixture at about one-hour intervals during operation, and thus
we presently prefer to use the remote apparatus 150.
As discussed previously, the belt 10 or 20 being
coated tends to creep or "drift" sideways (edgewise) one way or
the other during its revolving travel around the pulley rolls
34 and 36 (FIG. 3), or around the pulley rolls 108, 110, 112
and 114 (FIG. 5). Therefore, it is necessary to counteract the
drift by steering the belt by turning the steering screw 70, as
already explained. This sideways creeping or drifting of the
belt and the counteracting steering action causes a problem
with respect to the desired uniformity of the matrix coating
being applied. If the thermal spray gun traverses uniformiy
constantly with respect to the frame of the coating machine,
as the leadscrew 92 ordinarily constrains it to do, then some
non-uniformity of the transverse motion of the gun with respect
to the belt may occur. The result of such non-uniform relative
: : motion between the gun and t~e belt is that more coating is
deposited in some areas of the belt and less in others.
~ In order to cause the thermal spray gun *o
: .
:~ traverse constantly and consistently uniformly with respect to
.. : : the belt 10 or 20 being coated, regardless of any sideways
~ .
(edgewlse) belt movement, the presently preferred apparatus
and system, as shown in FIG. 6, is advantageously employed.
,
-38-
~2S~l~37~
The upright leg 140 of the rack 96, as shown in FIG. 3 or 5,
is cut off below the level of the shelf 138, thereby creating
a rack assembly 94' which is laterally "floating"; that is,
which is free to move back and forth in a direction parallel
with the axis of the leadscrew 92, in order to allow the gun
~to "track"any lateral ~edgewise) motion of the belt, as will be
explained~ The entire thermal spraying apparatus is "floating"
for accommodating lateral motion, including the leadscrew 92
and its drive 94 and their support frame 96', together with the
thermal spray gun and its carriage 98. In other words, this
"floating" allows the leadscrew and gun to be moved freely
laterally with respect to the face of the casting belt being
coated; that is, to be moved in a horizontal direction parallel
to the axis of the leadscrew 92.
There is a stationary horizontal track frame 164
which extends parallel with the leadscrew 92 and also parallel
with the axis of the work roll 36. This track frame 164 is
supported and secured by brackets 166 and 168 to the stationary~
hood 100, for example, by welding attachment of these brackets.
This track frame 164 has a generally hollow rectangular con-
figuration as seen looking at its left end in FIG. 6. There is
an elongated cut-out clearance opening or slot 170 in the upper
surface of the track frame~l64, and this ~longated slot opening
170 extends to the left (outboard) end of the track frame.
A removable plate 172 bridges the gap at the left end of the
slot opening 170, being fastened by four machine screws,
washers and nuts~174.
~ his track frame 164 has a pair of parallel in~
turned flange tracks 176 and 178 which are spaced apart and
are located in the same horizontal plane for serving as track-
-39-
3~25~ 37~
trackways parallel with the axis of the leadscrew 92. Riding
along these parallel trackways 176 and 17~ are a pair of whe~l~d
carriages 180 and 182 comprising plates welded to the top of
the floating frame 96' and projecting out on both sides. Each
carriage 180 and 182 has four supporting wheels 184 with hori-
zontal axes in planes perpendicular to the axis of the leadscrew
92. There are two wheels 184 on each side of each carriage, so
that each carriage has two wheels rolling along each trackway
176 and 178 for supporting the floating frame 84'.
In addition to these four supporting wheels 184,
each carriage 180 and 182 has a guide wheel 186 with
vertical axis. These guide wheels 186 are located below each
carriage 180 and 182 for rolling along the edges of the respect-
ive trackways 176 and 178 for guiding the movable frame 9~'for
causing it to move parallel with respect to the face o-f the belt
in the region being thermally sprayed by the gun.
It is noted that the inboard (right) end of the
leadscrew 92 as seen in FIG. 6 is mounted in a bearing assembly
190 bolted to the lower surface of the movable frame 96'.
The movable frame 96' ends just beyond the location of the bear-
ing 190. Thus, seen as a whole, the movable frame 96' has an
L-shape, with the longer shank of the L extending horizontally
and with the shorter leg~of the L extending down vertically,
with the platform or shelf 138 secured to the lower ends of this
vertical leg.
; In order to track the edge of the belt 10 or 20
being coated, there is a tracklng roller 196 having a vertical
axis mounted on an arm member 194 carried by an inverted V-
shaped s-pport member 192 with a foot pad secured at 193 to the
.
-40-
S5~7~i
top of the movable frame 96' between the carriayes 180 and 182.
This inverted-U support member 192 has sufficient height and
width to reach completely over an~ to clear the hood 100, in
all positions, and its upstanding leg 195 extends up through
the clearance slot opening 170~ A tension spring 188 extends
between the s-tationary bracket 168 and the support member 192,
thereby urging the movable frame 96' toward the left (toward
the outboard direction) for causing the tracking roller 196 to
maintain contact with (and thus to follow) the inboard (right)
edge of the belt.
Conse~uently, the thermal spraying assembly is
caused by the spring 188 and the sensing roller 196 to track the
belt regardless of any edgewise creeping or drifting of the belt
as the belt revolves. If there were no turning of the lead-
screw 92l then the path of thermal spraying on the belt surface
would be aligned at a fixed distance from the sensed belt edge
regardless of any lateral (edgewise) movements of the belt.
Now, when the uniform leadscrew induced motion
of the spray gun is superimposed on the aforesaid automatic
tracking of the belt edge, the resul-t is to produce a desired
uniform coating action, regardless of any lateral (edgewise)
movements of the belt in ei~her direction. In other words,
as the belt revolves and as the leadscrew 92 causes the gun to
move relative to the floating frame 96', the resultant adjacent
passes of thermal spraying are always at a uniform predetermined
distance from each other blending into each other in predict-
able fashion on the belt surface, resulting in applying a coat-
ing of uniform thickness onto the belt, regardless of any
lateral (edgewise) movements, i~e., regarless of any lateral
-41-
~255~
wobbling, of the revolving bel-t.
This desired uniformi-ty of application of the
coating is advantageously achieved regardless of sideways drift
of the belt or the steering of the belt in correcting such
drift. This uniformity is also advantageously achie~ed regard-
less of any camber which may happen to exist in the belt e~ge,
since all that is required of the adjacent passes of the spray
is that they be of uniform predetermined distance from each
other for appropriately blending, not necessarily that they
be perfectly straight, i.e., that they lie in a perfect helical
path on the belt surface.
In order to enable the fork-shaped gun carriage 98
to move accurately relative to the floating frame 96', this
carriage 98 includes a chassis 199 on which are mounted a pair
of wheels 19~ having vertical axes. These wheels 198 roll along
an accurately~machined guide way or track 99 on the side of the
horizontal leg of the movable frame 96'. A similar pair of
wheels (not seen) on the other side of this chassis 199 roll
along a similarly accurately machined guide way on the opposite
side of the horizontal leg of this movable frame 96'. Thus,
the wheels 198 of the carxiage 98 are in straddling relationship
with the frame 96' for holdi~ng the carriage 98 accurately align-
ed for holding the gun housing 105 accurately spaced from the
~elt surface as the leadscrew 92 rotates. Another pair of
wheels 200 (only one is seen) mounted on opposite ends of the
chassis 199 on horizontal axes roll along an accurately
machined guideway on the under sur~ace of the horizontal leg
of the movable frame 96' for steadying the gun carriage 98 to
prevent it from swaying. A strut 202 extends down from the
carr~age 98 and is ad~ustably secured at 204- to the side of the
gun housing 105. In FIG. 6, the viewer sees the rear of the
gun housing 105, for its nozzle is aimed at the belt.
-42-
~255~
Although FIG. 6 shows the therrnal spray gun aimed
at the belt as the belt passes around the roller 36, it is to
be undexstood that this laterally-floating belt-tracking
thermal spraying apparatus of FIG. 6 can also be employed
advantageously with a four-pulley coating machine as is shown
in FIG. 5.
It is to be understood that the roller 196 serves
as a sensor of the belt edge location, and the spring 138
serves as motive means for moving the movable frame 96' in
response to the sensing action of the roller 196. Other belt-
edge sensor means, for example, such as sliders, electrical
contacts, light beams and photoelectric cells, pneumatic or air
jet position sensors, magnetic sensors and so forth, can be
used in connection with other motive means for moving the frame
96', for example, such as electrical, pneumatic or hydraulic
motive means in a servo loop control system responding to such
sensor means, such servo loop control systems being well known
to those in the field of machine motion control.
- Moreover, instead of tracking the belt edge,
it is possible to paint or apply a narrow strip of contrasting
color along the margin of the belt near its edge and then to
~ ~ track such a strip.
- However, the edge of a steel belt is very
definitive by nature, and we have found this completely
mechanical sensing and mo~tive means for producing automatic
belt tracking movement of the whole laterally-floating thermal
spray assembly to be~eminently practical and very reliable
and durable.
-43-
~Z5~i~37~
RESULTS OF THE INVENTION
__.
The present invention of thermally spraying a
unitary-coat fusion-bonded matrix protective coat'ing of powder
mixtures of heat-resistant metallic and refractory non-metal,lic
components is capable of meeting all of the following essential
or desirable conditions. The fusion-bonded matrix coating (1)
is adherent to the flexible base metal of the belt or to edge-
dam blocks; (2) provides adequate thermal insulation;
(3) is resistant to mechanical damage, ~- i.e., spalling flake-
off or abrasion; (4~ is resistant to thermal shock; (5) affords
an acceptable often attractive surface finish on the cast
product; (6) is acceptably non-wetting with respect to molten
metal cast; (8) affords accurate proportioning of insulation
between the belts and the edge dams; (9) has desirable access-
ible porosity throughout the matrix coating; (10) is compatible,
because of surface characteristics, with additional min~imally
applled temporary top-coatings, such as oil or graphite or a
combination; and (11) can be applied practically by means of a
readily constructed and readily operated machine as described.
- In accordance with customary practice in using
belt casting machines, the user may find it desirable or may
wish to apply a temporary top coating over the fusion-bonded
matrix coa'ted belts. For example, a temporary coating of
colloidal graphite applied and dried from an aqueous or solvent
solution has been found sui~able for use on such matrix coated
belts for casting copper product P.
, Judging from previous experience, we believe
that amorphous carbon or soot, applied for instance as a col-
loidal suspension, may be substituted for the graphite top-
coat.
-44-
~ Z55~37S
In the case o~ casting aluminum slab as the
product P, diatomaceous silica may be included in this temporary
top-coating. In the casting of copper, a trace of oil appears
to be desirable and may be sprayed onto the fusion-bonded ma-trix
coating oE a new belt in minute quantities, however not enough
to appear wet or to result in any decomposition of the oil.
In the casting of copper bar to be used for draw-
ing into wire, belt life top and bottom was increased by a
margin of nearly 2 to 1, when the belts had been fusion~bonded
matrix coated in accordance with this invention. Surface quality
was remarkably improved, owing in part to the ability to use
much less oil or top-coating than conventional practice, thus
reducing its attendant hydrogen-related porosity in the cast
product. Improved metallurgy of the copper rod indicated
that improved drawabllity was present also.
In an early test of casting of copper bar, the
matrix coating of ~xample I was used on a top belt 20 only. The
thickness was around 0.002 of an lnch on a hard-rolled, low-
carbon titanium steel belt 0.044 of an inch thick. This cast was
stopped after three hours, Eor reasons not related to the belt
coating, which was still in excellent condition. No precoat
of graphite was used at first, and a l~ittle pickup of copper was
experlenced. The next cast on this top belt ran 24 hours with
two interruptions not related to the belt coating. The quantity
of oil appli~ed onto the belts was~ reduced as compared with
conventional practice in casting copper bar in a twin-belt
machine, with good resultsO The test was terminated after 24
hours due to reasons not related to the belt coating.
The above copper bar casting test was repeated
with an Example I matrix coated low-carbon-steel upper belt 20
of No. 2 temper, no titanium content. The results were just as
-45-
~25~37~
good as with the titanium steel belt, and such good results
were not expected, because such good results were contrary to
previous experience in attempting to cast copper bar on such
a non-titanium-containlng steel belt. Priox experience had
been that hairline cracks might be expected to occur in such
a non-titanium--containing belt after 8 to 10 hours of repeated
cyclic contact with mo].ten copper and cyclic flexing. Such
cracks did not appear in the matrix coated non-titanium-
containing belt that was tested for eight to ten hours.
A further copper bar casting test was conducted
with a fusion-bonded matrix coating according to Example III.
This coating was applied onto low-carbon, hard-rolled titanium
steel belts of 0.044 inch thickness. This time, such fusion-
bonded matrix coated belts were used both as the top and bottom
belts 20 and 10. Oil was lightly sprayed onto the bottom belt.
After an initial light applica-ti.on of oil on thetop belt, it
was only necessary to wipe the top belt perhaps three times an
hour, in order to dislodge slight pickup. Results were thé
best ever, including the longest belt life which we have seen
for casting copper. Belt life, top and bottom, was increased
by a margin of nearly 2 to 1.
An example of the benefits of the subject
invention has been the experimental casting of aluminum alloys.
Surface improvement of the metal being cast was remarkable.
Rosettes and streaks formerly observable during the castlng
process wereeliminated~ on both the top and the bottom of the
cast slab. Re]ectable material was greatly reduced. The
fusion-bonded matrix coated belts were still in good condition
well beyond the useful life of conventional belts. The edges
of the cast slabs were excellent, owing to the proportioned
heat transfer between edges and belts by use of the insulative
coatings.
-46-
~2S5~37~
~ n our expericnce, in order to operate advantayeously
in use, an endless flexible casting belt having a fusion-bonded
matrix coating thereon in accordance with this invention will be
capable of repeatedly flexing around a pulley roll having a
diameter of 20 inches (508 mm) without occurrence of flaking or
spa]ling of said coating.
Although the examples and observations sta-ted herein
have been the results of experi.mental field trials of belts
matrix-coated, as described,on which were cast molten copper or
molten aluminium and aluminum alloys, and tests with molten
steel poured onto stationary sections of coated belt, allowing
a vertical fall of fourteen inches before the molten steel
impacted against the coated belt, this invention appears applica
ble to the continuous casting of~any metal or alloy having a
meltinq temperature equal to or less than steel.
Although speclfic presently preferred embodiments
of the invention have been disclosed herein in detail, it is
to be understood that these examples~of the invention have been
described for purposes of illustration. This disclosure is
not to be construed as limiting the scope of the invention.
.
,
~:
,
:::
:
-47-
