Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
1- 2v~ 8
POWDER FEED COMPOSI~ION FOR FORMING
REFRACTORY OXIDE THERMAL SHOCK
~ESISTANT COATI~G, PROCESS AND ARTICLE
Field of the Invention
The present invention relates to a thermal
spray powder feed composition for forming a
refractory o~ide coating having high thermal shock
resistance, increased wear resistance and resistance
to spalling in thermal cycling environments and to a
process for forming a refractory o~ide coating and
to an article having a refractory o~ide coating.
~ackaround of the Invention
This invention is related to the problem of
providing a high wear and thermal shock resistant
coating for hearth rolls for annealing steel,
stainless steel and silicon steel sheet in a hearth
(furnace). The hearth rolls carry the steel sheet
through the hearth. The temperature in the hearth
may vary from about 1500~ to over 20~0~F depending
upon the type of steel, the travel speed of the
sheet steel as it passes through the furnace and the
duration of time in the furnace.
A major problem encountered in the
annealing operation is the transfer or pick-up of
material from the steel sheet to the hearth rolls.
If pick-up occurs, it will accumulate on the hearth
rolls and damage the steel sheet being processed.
To avoid this problem frequent roll changes are
required with concomitsnt costs fos replacement and
lost production. This problem has become more
severe in recent years since thinner sheets are
D-16458
',
..
- 2 - 2~ 8
being used, along with higher ~peeds and temperature
to increase productivity.
To suppress the transfer of material to the
hearth roll and to increase wear resistance it is
desirable to coat the hearth roll with a coating
composition which is substantially chemically inert
at elevated temperatures. An undercoating of metal
or a ceramic-metal alloy is used to prevent
spalling. Spalling may also be prevented using a
graded coating in which the composition of the
undercoating is gradually varied from 1~0~~ alloy to
100~~ ceramic. Unfortunately, the ceramic coatings
presently available usually crack in thermal cycling
due to a large difference in thermal expansion
between the substrate, a heat resistant alloy, and
the coating. The alloy undercoat at the interface
is o~idized at or above l000~C in the presence of
o~ygen leading to spalling of the ceramic layer.
When a graded coating is used, the alloy component
of the coating is also o~idized which, in turn,
increases the volume of the coating. Upon cooling,
the coating spalls due to excessive compressive
stress created by the shrinkage of the substrate.
SUMMARY OF T~ INVF~TION
It was discovered in accordance with the
present in~lention that a thermally sprayed coating
formed from a powder feed composition of zirconium
silicate and partially stabilized zirconia possesses
high resistance to thermal shock in thermal
cycling. The powder feed composition of the present
invention produces a coating particularly useful to
D-16458
-- 3 --
2 ~ 8
protect a hearth roll in a continuous annealing line
for annealing steel, stainless steel or silicon
steel sheets. The powder feed composition comprises
particles of zirconium silicate in a mi~ture with
particles of zirconia stabilized or partially
stabilized with an o~ide selected from the group
consisting of Y2O3, CaO, MgO, CeO2 and HfO2. The
powder feed composition is applied by a thermal
spray techni~ue to produce an as-deposited coating
having a composition which by ~-ray phase analysis
comprises zirconia, silica and zirconium silicate.
The major component of the powder composition is
partially stabilized zirconia with the zirconium
silicate preferably limited to a ma2imum of 60 wt%.
For plasma spray application the powder feed
composition should comprise at least 65 wt%
stabilized zirconia with the remainder substantially
zirconium silicate whereas for detonation ~un
application the powder feed composition should
comprise at least 40 wt% stabilized zirconia and up
to 60 wt% zirconium silicate. The zirconia may be
either fully or partially stabilized although
partially stabilized zirconia is preferred.
It was further discovered in accordance
with the present invention that spalling of the
coating may be prevented at elevated temperatures
e~ceeding 1150~C by thermally spraying a metallic
undercoat. The preferred undercoat is a cobalt
based metal matris comprising Co-Cr-Al-Ta-Y with a
dispersion of A12O3.
~-16458
_ ~ _ 2 ~J~ g
BRIEF DESCRIPTION OF THE DRAWING
The single figure is a schematic
representation of the equipment used to test the
propensity of the as-deposited coating on a hearth
roll for pick-up o~ me~al or metal o~ides under
dynamic and static local conditions.
DESCRIPTION OF THE PREFERRED E~RODIM~T
The present invention is based upon the
discovery that a starting powder feed composition
consisting essentially of a mi~ture of zirconium
silicate and zirconia with the zirconia being
stabilized with a stabilizing o~ide such as yttria,
calcia, or magnesia may be thermally sprayed to form
a coating possessing the characteristic of being
resistant to thermal shock and resistant to steel or
steel o~ide pick-up from a continuous annealing
line. Any conventional thermal spray technigue may
be used to form the coating including detonation gun
deposition and plasma spray deposition. The chemical
composition of the thermally sprayed coating should
consist of a mi~ture of at least about 40 wt%
zirconia (ZrO2), including a stabilizer for the
zirconia selected from the group consisting of CaO,
Y203~ MgO, CeO2 and HfO2 with the balance zirconium
silicate (ZrSiO4~ and/or its decomposition products
SiO2 and ZrO2. The preferred weight percent of the
component o~ides in the coating is 55 to 85%
stablized ZrO2 and 15 to 45% ZrSiO4 and/or its
decomposition products SiO2 and ZrO2. The optimum
weight percent of the component o~ides in the coating
is 70 to 85% stablized ZrO2 and 15 to 30% ZrSiO4
and/or its decomposition products. The stabilizer
D-16458
.
- ::
- 5 - 2~3~3~
should be between 2 and 20 wt% ~f the zirconia
component.
The coatings are preferably applied by
detonation gun deposition or plasma spray
deposition. A typical detonation gun consists
essentially of a water-cooled barrel which is several
feet long with an inside diameter of about 1 inch.
In operation, a mi~ture of o~ygen and a fuel gas,
e.g., acetylene, in a specified ratio (usually about
1:1) is fed into the barrel along with a charge of
coating material in powder form. Gas i5 then ignited
and the detonation wave accelerates the powder to
about 2400 ft./sec. (730 m/sec.) while heating the
powder close to or above its melting point. After
the powder exits the barrel, a pulse of nitrogen
purges the barrel and readies the system for the next
detonation. The cycle is then repeated many times a
second.
The detonation gun deposits a circle of
coating on the substrate with each detonation. The
circles of coating are about 1 inch (25 mm) in
diameter and a few ten thousandths of an inch
(several microns) thick. Each circle of coating is
composed of many overlapping microscopic thin
lenticular particles or splats corresponding to the
individual powder particles. The overlapping splats
interlock and bond to each other and the substrate
without automatically alloying at the interface
thereof. The placement of the circles in the coating
deposition are closely controlled to buila-up a
smooth coating of uniform thickness and to minimize
substrate heating.
.
D-16458
.
, - 6 - 2~~ 3~
In the plasma arc spray process, an electric
arc is established between a non-consumable electrode
and a second non-consumable electrode spaced
therefrom. Gas is passed in contact with the
non-consumable electrode such that it contains the
arc. The arc-containing gas is constricted by a
nozzle and results in a high thermal content
effluent. The powders used to produce the coatings
are injected into the effluent nozzle and are
deposited onto the surfaces to be coated. This
process, which is described in U.S. Pat. No.
2,858,411, produces a deposited coating which is
sound, dense and adherent to the substrate. The
applied coating also consists o~ irregularly shaped
microscopic splats or leaves which are interlocked
and bonded to one another and also the substrate.
In general the coating composition for the
plasma arc spray process will be substantially
equi~alent to its corresponding starting material
composition. When using the detonation gun to apply
the starting material evaporation of the components
may result in a significantly different ratio of
constituents in the as deposited coating. Thus some
change in chemistry may occur during deposition,
using any thermally sprayed process. Such changes
can be compensated for by adjusting the powder
composition,or deposition parameters.
8ecause of the complex phase diagram for
Zr-Si-O, the solidifying ZrSiO4 powder particles may
contain ZrSiO~ as a crystallographic phase and/or
ZrO2~SiO2 as the decomposition products of the molten
ZrSiO4 in separate crystallographic phases within
D-16~58
~ 7 ~ 2~ 8
individual splats. Thus the ZrO2 and SiO2 are
intimately associated within each splat which had
previously been ZrSiO4 in the powder form. By
~associated" is meant the e~tremely fine and
intermi~ed crystalline structure of SiO2, ZrO2 and/or
ZrSiO4 crystallites within the splat.
Although the coatings of the present
invention are preferably applied by detonation or
plasma spray deposition, it is possible to employ
other thermal spray techniques such as, for e~ample,
high velocity combustion spray (including hypersonic
jet spray), flame spray and so called high velocity
plasma spray methods (including low pressure or
vacuum spray methods). Other technigues can be
employed for depositing the coatings of the present
invention as will readily occur to those skilled in
the art.
The thermal spray coating may be applied
directly to the metal substrate. However, an
undercoat compatible with the substrate and resistant
to o~idation is preferred. An ùndercoat of a
ceramic-metal alloy mi~ture having a cobalt based
metal matrix containing alumina is preferred.
Optimum coatings are a cobalt based alloy with
alumina dispersions de~cribed in U.S. Patent No.
4,124,737, the disclosure of which is herein
incorporated by reference. The refractory o~ide
coating reacts with the preferred undercoat to
produce an impervious thin leyer of aluminum and/or
zirconium o~ide phases at the interface which
prevents o~idation of the undercoat as well as to
provide good bonding between the undercoat and the
.
D-164~8
..
- 8 ~ 3 ~ ~ 8
ceramic layer. The impervious layer may be less than
5 microns (0.005mm) and is produced as a result of
interdiffusion or o~idation in an environment at an
elevated temperature in the presence of osygen. A
similar layer may form in an inert atmosphere as a
result of reaction between the o~ide overcoat and the
metallic undercoat.
E~am~le l
To substantiate the superiority of the
coating composition of the present invention, various
powder mixtures containing a yttria stabilized
zirconia (ZrO298% Y203) and zirconium silicate
(ZrSiO~) were mechanically blended into the blend
ratios identified in Table I and fed to a plasma
torch in a conventional manner to produce a coating
on a 304 stainless steel bar. The ceramic coating
was applied on one face of the 304 stainless steel
bar (2 3/4~ ~ 3/9 ~ 1/2~h) which was first coated
with a detonation gun with an undercoat of 50 to 75
microns of a cobalt-based coating undercoating of 90
(Co-25Cr-7.5Al-0.8Y-lOTa) + 10% A1203. Then the
ceramic coating was ground to a 100 micrometer
thickness before heat cycling.
The coating specimens were heated in air to
1150~ ~o 1200~C and held for a minimum of si~ (6
hours followed by air cooling. After five (5)
cycles, the specimens were water quenched on the
si~th cycle. If no spalling or partial spalling
occurred, the coating was described as acceptable.
The results are shown below in Table I.
458
9 2 ~
TARLE I
Coating Powder MixSure % Test Result Hardness
No. Al~ B" Spall;ng After ~ tiV 0.3
1 100 0 No 402
2 85 ~5 11~ 587
3 75 25 No 539
4 65 35 Parti~l Spalling~ 532
Spa11 ed 482
6 25 75 Sp~lled 535
7 0 100 Spalled Before tlQ 481
P~wder A-yttria st~biliz~d zircon1a (ZrO2 . 8% Y203)
' Powder B-zirconium silicate (ZrSiO4)
A Parti~ll spall;r)g indicates A lift1ng of th~ cer~m;c layer at edges.
It is apparent from Table I that for plasma
torch applications spalling did not occur using a
starting powder composition of greater than 65 wt.
zirconia with above about 75 wt. % zirconia being
optimum.
le 2
Similar tests were conducted using a powder
feed composition of calcia stabilized zirconia
(ZrO2-5CaO) and zirconium silicate in a mechanical
blend at the various powder ratios labeled 8 to 12
as reflected in the following Table II. The powder
feed composition was applied through a detonation
gun over a preferred cobalt-based metal matri~
undercoat of (Co-25Cr-7.5Al-lOTa-.08Y) +
(30% A12O3). The wear performance, thermal shock
resistance and anti-pickup performance was optimum
for coatings made with mi~ture number 8.
D-16458
: :
- TA8LE II
~Q~ g ZrSiO4/Zr~ ~ater Ouench fro~ 900~C ~Coatinq Chenistr~ ~t%
ZrSiO4 ZrOz
8 50/SO ZO c~cles OK 43 5 56 5
9 15~85 20 c~cles OK 21 79
30/70 5 cycles 50% spall 34 66
11 70Z30 1 c~cles 50Z spall 46 5 53 5
lZ 65/lS ZO c~cles OK 45 55
'' ' '
- - ~Coatins chemistr~ - based on metallic elerent concen-ration ~easured b~ Electron ~icroprobe Anal~zer
.: -- 10 --
~: .
:,
c
C~
Q~
- 11 - 2 ~t~8
Additional tests as set forth in Examples ~,
4 and 5 were conducted to substantiate the
superiority of a detonation gun coating formed from a
thermal spray powder feed composition comprising
zirconium silicate and partially sta~ilized zirconia
(2rO2-5CaO) on a hearth roll based on its anti-pickup
characteristics compared to detonation gun coatings
of a conventional composition of
(Co-2SCr-7.5Al-~8Y-lOTa)+10% A1203, a coating from a
feed composition of ZrSiO~, and bare steel. It
should be noted that while a detonation gun coating
of (Co-25Cr-lOTa-7.5Al-0.8Y) ~ 10% A1203 powder is a
preferred undercoat for the ceramic coatings of this
invention it has been used in the past as a hearth
coating by itself and for comparison purposes is
designated as coating number 13 in Table III.
E~ample 3
The following anti-pickup test was conducted.
Evaluation Method: A semicircular roll
simulator as depicted in Figure 1 was used. The
rocker 6 is rotated a~out pivot point 7 by
reciprocating arms B and 9. Two coated specimens 10,
12 were mounted under the roll for simulating
reciprocating sliding motion over the specimens and
static reaction under load. A powdered Fe203 and/or
Fe powder 14 is placed on specimen 10 for dynamic
pickup evaluation and powdered Fe203 or Fe powder 16
is placed on specimen 12 for static pickup
evaluation. The test measures the propensity of Fe
or Fe203 to stick to surface faces A, B and C
respectively under a load of 8.Skg.
. . .
D-1645B
- 12 - 2 ~
Test Conditions
Temperature: : gOOu~
~uration Time : 4 hours
Atmosphere : Nitrogen 98% + Hydrogen 2%
Pickup source : ~e & ~e203
Anti-Pickup Index: is determined by the sum
of the anti-pickup points (AP) for faces A, ~ and C
based on the following Table III.
.
~able III
Graded
ValueDe~cription
Good 3 No adhesion
2 Adhe~on can be removed by 6kg/cmZ air blow
1 Adhesion can be removed by hand-rubbing
Bad O Adhesion cannot be removed by the above method6
lts
AP Point with AP Point with Grand
Fe Powder Fe304 Powder Total
Coating A B C Total A B C Sotal
9 0 3 1 4 0 2 3 5 9
13* 0 1 1 2 0 0 1 1 3
14 0 0 0 0 0 3 2 5 5
uncoate~ O O O O O O O O O
steel
* Coatin~ 14 was made with a ~owder consisting of
only zlrconium ~ilicate (Zr iO~,).
** Bulk Material for reference: R~frgctory Case
Steel
D-16458
:' :
- 13 - 2 ~
~ample No. 9
~ Wear Resistance under elevated
temperature
Equi~ment: High Temperature Ball on Disk
Wear Tester
Test Conditions:
Temperature: 900~C
Atmosphere: N2 98% + H2 2%
Ball size: 9.252mm in diameter
Weight: 2 kg
Sliding Speed: 0.2m/sec
Track: 35 mm in diameter
Total Sliding length: 720m
~all: High carbon chromium bearing steel,
SVJ2
Disk: Coated sample
Wear Rate Calculation:
The normalized wear rate is calculated by
the following procedure.
a. The wear volume is calculated from
area of cross section of scar measured from a
profilometer chart multiplied by the circumference
of the scar.
b. Wear volume is then divided by applied
load and total sliding distance.
Results
Coati~q Wear rate mm~k~
8 1.2 x 10-8
13 7.2 ~ 1~-8
1~ 2B0 ~ 10-8
uncoated steel 850 ~ 10-8
D-16458
- ~4 - 2~3~
E~am~le No. 5
~eat Shock Test:
1000~C - 15 minute heating from ambient to
1000~C, water quench for 15 minutes - 1 cycle
The sample substrate material had a thick
plate shape with a dimension of 50250~10 mm, one
50~50 side coated. The sample substrate was 304
stainless steel.
Thermal shock resistance is evaluated by
counting how many cycles the coating can survive
without spalling off.
S~m~le No. of CYcles
8 no spalling after 20 cycles
1~ no spalling after 20 cycles
14 spalling after 6 cycles
S~ rY of Testina
As is apparent from the tables and all of
the above e~amples the optimum starting powder blend
will necessarily vary depending upon the stabilizer
and thermal spray technique used. The as-deposited
coating should preferably have at least about 40 wt~
stabilized zirconia and up to 60 wt% zirconium
silicate. The preferred concentration of the
component o~ides in the coating independent of
crystalline structure should be between about 55 to
about 85 wt% stabilized ZrO2 and 15 to 45% ZrSiO4
and/or its decomposition products SiO2 and ZrO2 with
70 to 85 wt~ stabilized ZrO2 being optimum. The
concentration of stabilizer in the as-deposited
coating composition should be betweén 2 to 20 wtS of
the zirconia component.
D-16458
. .
'
.: .~ ,
