Note: Descriptions are shown in the official language in which they were submitted.
lZ';~53~
17GE 29 8 2
_ 1 _
CHROMIUM CARBIDE COATING FOR PROTECTING STEAM TllRBINE
COMPONENTS SUBJECT TO SOLID PARTICLE EROSION
Background of the Invention
This invention relates to a solid particle
erosion resistant coating for components of a steam
turbine, and in particular relates to a chromium
carbide and a matrix material coating. The coating
protects against solid particle erosion of various
components of a steam turbine, such as the turbine
blades, the nozzle partitions and other components
directly affected by solid particle erosion due to
their location in the steam flow path of the
turbine.
During the normal operation and start-up of a
steam turbine, solid particles of relatively small
size may be present in the steam which is supplied to
the steam turbine. It is believed that the solid
particles are predominantly magnetite, Fe3O4,
which is sometimes called ~boiler scale" and which is
thou~ht to originate in the steam boiler tubing and
adjoining pipes. These solid particles are carried
through the entire steam turbine system by the steam.
The ability of the solid particles to erode various
components in the steam turbine system is dependent
,s;
,j
~Z'7353l3
17GE 2982
--2--
upon the velocity of the particles relative to that
particular component and other parameters such as
steam temperature and pressure. The turbine blades
which rotate within the steam turbine at high
velocities are particularly subject to solid particle
erosion. The nozzle partitions, which direct the
steam flow onto the turbine blades, and the web
diaphragms proximate the steam admission ports, are
also subject to solid particle erosion. Other
components such as certain valve bodies, ~tems, and
skirts, which are directly in the steam flow path,
are affected by solid particle erosion due to the
pressure differentials and force of the steam as it
flows through the component. Typical temperatures
in an operating steam turbine approach 1000~.
The turbine blades and the nozzle partitions are
commonly made of 12-chrome martensitic stainless
steel. O~her components may be constructed of a
similar material or of chrome molybdenum vanadium
ferritic steel. Components made of these
compositions and exposed to solid particle erosion
often suffer accelerated metal loss and do not
achieve their expected equipment life spans. In the
past, the component substrates have at times been
coated with a tungsten carbide coating to prevent
solid particle erosion or a cobalt based alloy such
as Stellite which is a tradename for an alloy made by
Cabot Corporation. As is well known in the art,
tungsten carbide has a hardness which exceeds the
hardness of many other materials, including the
family of chromium carbides. The family of chromium
carbides includes Cr3C2, Cr7C3 and
Cr23C6. This quality of tungsten carbide has
- :
~z7353~3 17GE 2982
been thought to be of primary importance in the
prevention of solid particle erosion.
Chromium carbide, in combination with a nichrome
matrix, has been utilized in gas turbine systems to
prevent hammer wear or fretting wear between adjacent
components. Hammer wear occurs when the two
components impinge upon each other in a state of high
frequency vibration. This wear is observed on the
shrouds linking together the outer radial portions of
the gas turbine blades. Fretting wear occurs when
two contacting surfaces experience oscillatory
tangential displacements of small amplitude. In gas
turbine technology, fretting wear is observed at the
roots of the turbine blades which are fit into the
disks. The disks are affixed to the shaft of the gas
turbine rotor. Sometimes, to prevent this wear, the
roots of the gas turbine blades and the shrouds are
coated with the chromium carbide and nichrome
coating. However, these particular components are
not subjected to solid particle erosion degradation
due to their position outside the gas flow path of
the turbine.
It is recognized in the art that fretting wear
is distinctly different than solid particle erosion.
Fretting wear is caused by metal-to-metal contact.
Solid particle erosion is caused by hard particles
impinging upon an object at high velocity and
traveling along its surface. Abrasive wear is also
produced by solid particles, however those particles
impinge the surface at negligible velocities or move
over a surface because they are trapped in a
metal-to-metal interface. A solid particle eroded
surface possesses a much greater surface roughness
1~7~S31~
- ~ - 17GE 2982
than an abraded surface because the high velocity
impinging particle has a higher kinetic energy
and may more readily remove material from the
surface. Attendant metal losses produced by
each particle in the soiid particle erosion
process are greater than the corresponding losses
produced by the abrasion mechanism. It should
be noted that the industrial usage of the terms
"wear," "abrasive wear," and "erosive wear" are
lQ not always consistent with the textbook classifications
of those terms.
At least one manufacturer of coating
materials has suggested a chromium carbide and
nichrome coating for jet engine, i.e., gas turbine,
rotors which are not in the gas flow path of
the turbine. The typical composition of the rotor
coating is, in weight percent, 75% chromium
carbide particles and 25% nichrome (the nichrome
being, in weight percent, about 80% nickel and
about 20% chromium). That coating is obtained
by ~lasma spraying a blended, chromium carbide
and nichrome coarse powder capable of passing
through a 140 mesh screen. That same
manufacturer recommends generally, for particle
erosion resistant coatings, a powder having 75%
chromium carbide and 25% nichrome capable of passing
through a 325 mesh screen. Again, that same
manufacturer recommends a 85~ chromium carbide and
1~;3S313
17GE 2982
15~ nichrome coarse powder, capable of passing
through a 140 mesh screen, where tungsten carbide
coatings cannot be used because of poor oxidation
resistance. The manufacturer states that the 85
chromium coarse powder, (hereinafter a "coarse"
powder is a powder which cannot pass through a 270
mesh screen" and powders which can pass through such
a screen are classified as "fine") is not as
effective a solid particle erosion resistant coating
as the 75~ chromium carbide, coarse powder coating.
Therefore, a review of the man~facturer's literature
illustrates that tungsten carbide is recommended
unless severe oxidation is a problem and then
recommends, a 75~ chromium carbide 2S~ nichrome
coarse powder capable of passing through a 140 mesh
screen to prevent solid particle erosion.
Experimental hardness tests of different
percentages of chromium carbide and nichrome coatings
and different powder sizes has generally confirmed
- 20 the manufacturer's conclusions. The coarser powders
have higher microscopic hardnesses than the finer
powders. As the chart below shows, the 85% Cr3C2
alloy and 75~ Cr3C2 alloy of -325 mesh powder
particle size ~a " - " before the '!mesh" size
indicates that the powder passes through a 325 mesh
screen) has much lower microhardness values than the
coarser powders of the same composition. The
microhardness values are supplied below in Knoops
indentation hardnesses which is well recognized in
1~ the,
~1~ 30 the metallurgical art. The higher/value of the
microhardness, the harder the substance is on a
microscopic level.
,~ . . -
1~7353~
- 6 - 17GE 2982
TA~LE 1
Powder Particle Microhardness
Alloy SizeAverage Knoops
B5Cr3C2-15NiCr -325 mesh 610
585cr3C2-15Nicr -140 mesh1243
75Cr3C2-25NiCr -325 mesh 732
75Cr3C2-25NiCr -140 mesh 843
65Cr3C2-35MiCr -140 mesh1192
Although the hardness test data show some
10 statistical variations, the hardness on a micro-
scopic level suggests that the coarser powders
may be better suited as a solid particle erosion
resistant coating than the finer powders.
Additionally, the apparent hardness of tungsten
15 carbide, which exceeds chromium carbide through
the applicable temperature range, points away
from the use of chromium carbide as presented
herein.
Accordingly, it is an object of this
20 present invent:ion to provide for a chromium
carbide-based coating resistant to solid particle
erosion for components in a steam turbine which
are directly affected by solid particle erodents
in the steam flow path.
Another object of this present invention
provides for a mixture of fine chromium carbide and
matrix material having about 85% chromium carbide and
15% nichrome matrix material, app
~ . . . _
,~
~,
1273~i3~ 17GE 29 8 2
transfer assisted process onto components subject to
solid particle erosion.
Summarv of the Invention
In one embodiment, the components in a steam
turbine directly affected by solid particle erosion
are coated by plasma spraying with a blended chromium
carbide and matrix powder having, in weight percent,
about 85% chromium carbide and about 15% nichrome as
a matrix material. Essentially all of the blended
powder is capable of passing through a 325 mesh
screen. The coating is an aggregate of chromium
carbide particles and nichrome matrix material which
has a metallographic porosity of 3~ or less. The
coating has a density of substantially 6.1 grams per
cubic centimeter.
General Descriptlon of the Drawings
Figure 1 illustrates graphically the results of
the solid particle erosion experiment testing
different percentages and coarse and fine powders of
2G chromium carbide composition;
Pigure 2a is a photomicrograph using a scanning
electron microscope, at 300X power, of a coated
surface resulting from a -325 mesh powder composition
of 85Cr3C2-15NiCr;
Figure 2b is a photomicrograph from a scanning
electron microscope, at 1000X power, of the same
surface as shown in Figure 2a;
Figure 3a is a photomicrograph from a scanning
electron microscope, at 300X power, a prior art
coated surÇace resulting from -140 mesh powder
composition of 75Cr3C2-25NiCr;
~Z7353~ 17GE 2982
Figure 3b is a photomicrograph from a scanning
electron microscope, at 1000X power, of the prior art
surface of Figure 3a;
Figure 4a is a photomicrograph from an optical
microscope, at 1000X power, of a cross section of a
composition coating of 85Cr3C2-15NiCr, -325 mesh;
and,
Figure 4b is a photomicrograph from an optical
microscope at 1000X of a cross section of coating of
85Cr3C2-15NiCr, -140 mesh.
Experimental Testing
To determine the actual solid particle erosion
resistance of chromium carbide, tests were conducted
on the chromium carbide and nichrome alloys listed in
Table 1 above. In this instance, the chromium
carbide and nichrome powders were blended and not
presintered or otherwise prealloyed together. Both
the chromium carbide powder and the nichrome powder
have other elements therein which are recognized in
the art as incidental or trace elements. The
chromium carbide particles are crystalline faceted
particles. The nichrome acts as a matrix material
and has, in weight percent, about 80% nickel and
about 20~ chromium along with other trace elements.
The various alloy powders were plasma sprayed
onto substrates of 12-chrome martensitic stainless
steel which were previously machined into wedge
specimens. One wedge specimen was not coated for
control purposes. Prior to bein~ coated, the steel
was prepared and cleaned by washing with isopropanol,
dried and grit blasted. The plasma spray, coating
procedure was conducted in accordance with the
coating alloy manufacturer's recommendations
~Z7~3~ 17GE 2982
utilizing argon as a primary effluent gas and
hydrogen as a secondary gas.
The coated wedge specimens and the control
specimen were placed in an erosion simulator. An
inert erodent, chromite spinel FeO-Cr2O3~ of a
predetermined powder size was introduced into the
flow path of the simulator. The chromite is similar
to the magnetite, which is thought to be boiler scale
carried by the steam in a steam turbine, in strength,
particle size and metallurgical structure. The size
of the chromite averaged 4 microns, and the
temperature was maintained in the simulator at
approximately 970~ ~ahrenheit. The rate of gas fiow
through the simulator as well as the amount of
particles released into the gas flow was controlled
to substantially represent the long term conditions
present in the steam flow path of a steam turbine.
The velocity of the mixture of gas and eroding
particles was 500-1000 feet per second in the region
of the test specimens. However, the experiment, by
necessity, greatly accelerates the solid particle
erosion conditions in a steam turbine environment.
Various measuremen~s were made of the specimens
prior to the test in the erosion simulator. The test
lasted approximately 100 test hours. The results of
the experiment are shown in Table 2 below.
,\ ....
. .~
1273~S38
- 10 - 17GE 2982
Table 2
Alloy Coating Weight Change
On Specimen Erosion Rate
85cr3C2-15Nicr 0.27 mg/hr
-325 mesh
75Cr3C2-25~iCr 0.66 mg/hr
-325 mesh
75Cr3C2-25NiCr 1.16 mg/hr
-140 mesh
65Cr3C2-35NiCr 1.32 mg/hr
-140 mesh
85cr3C2-15Nicr 2.02 mg/hr
-1~0 mesh
Uncoated 12Cr ~.53 mg/hr
stainless steel
Figure 1 graphically illustrates the
results of this test.
Detailed Description
According to the e~perimental test
results as illustrated in Figure 1, a fine powder,
85% chromium carhide and 15% nichrome as a
matrix material performs much better as a solid
particle erosion resistant coating than the coarse
powder of the same alloy. Also, Figure 1 graphically
illustrates that the 85% chromium carbide
composition, capable of passing through a 325 mesh
screen in its powdered form, provides more than
twice the erosion resistant protection than the
75% chromium carbide, fine powder. It should
be recognized that the testing of these various
alloy in the erosion simulator greatly
accelerates the solid particle erosion process
evidenced in the steam flow path of a
steam turbine and the actual field performance of
:
~Z73~3~3
17GE 2982
these coating alloys may differ accordingly. The
test results may present some statistical variations,
nonetheless the results clearly show that a fine
powder, high chromium carbide alloy coating provides
much better solid particle erosion resistance than
that disclosed and tauyht by the prior art. A
similar test on tungs~en carbide coatings revealed an
erosion rate in the same range as the uncoated steel,
i.e., approximately 4~5B mg/hr. A Stellite weldment
tested in a similar manner eroded at a rate of 5.56
mg./hr.
Generally, it is believed the chromium carbide
crystalline particles are softened and some are
partially melted during the plasma spraying
operation. It is believed the matrix material,
herein the nichrome alloy material, melts during the
plasma spraying process. When both the softened or
wetted chromium ~arbi~e parti~les and matrix ~lloy
material land on the surface of the component being
coated, the nichrome matrix material bonds to the
substrate and the chromium carbide adheres to the
matrix. Some of the wetted chromium carbide
particles are embedded into the matrix material to
form an aggregate "splattered droplet" surface
morphology. Figure 2 illustrates this surface
morphology. Some of the chromium carbide particles
may dissociate and others actually alloy or bond with
the matrix material or base substrate metal. The
coating process continues in a bonding and layering
fashion grossly similar to the diffusion and
attachment process noted in commercially available
concrete until a coating covers substantially the
entire surface and forms an aggregate. The aggregate
17GE 2982
1~73~3~
-12-
is, in this system chromium carbide particles bonded
together with a matrix material, which in this test
is nichrome.
Figures 2a and 3a show the surface of a coated
substrate. Figure 2a is a photomicrograph taken with
a scanning electron microscope at 300X power of ~he
85~ chromium carbide alloy which in its powder form
is capable of passing through a 325 mesh screen. In
contrast, Figure 3a is a photomicrograph of the 75
chromium carbide alloy, obtained by plasma spraying
the coarse, or -140 mesh, powder as described in the
prior art.
A comparison of these two photomicrographs
illustrates that the fine powder provides a much
smoother coating surface than does the coarse powder.
The 85% chromium carbide fine powder provides a
coating with a surface roughness of approximately 200
microinches of arithematic average amplitude
deviation from the centerline. This arithmetic
2~ average amplitude deviation is a measure of surface
roughness well known in the art. The centerline is a
line parallel to the general direction of the profile
of the coating such that the sums of the areas of the
profile which lie on either side of it are equal. In
other words, the centerline is the mathematical
medium line for the coating surface. In contrast,
the coating resulting from the 75~ chromium carbide
coarse powder, shown in Figure 3a, has an arithmetic
average amplitude deviation of approximately 660
microinches from the centerline.
Figures 2b and 3b are provided by a scanning
electron microscope amplifying the surface of the
respective coatings 1000X. The 85% chromium carbide
1~7~3~ 17GE 2982
-13-
fine powder has much shallower valleys, fissures and
crevices than the 75% chromium carbide coarse powder.
On a microscopic scale, the chromium carbide
particles seem to remain essentially intact during
the plasma spraying operation with the matrixing
material bonding to some of the chromium carbide
particles and forming an aggregate composition as the
coating. This feature is illustrated in Figures 4a
and 4b. The microstructures consist of elongated
carbide rich islands separated by thin,
semi-continuous laminar oxide "stringers" and small,
globular oxides as pointed out in Figure 4b. Since
the-coating is an aggregate, some pores or voids are
noted in the microstructure. The photomicrographs in
Figures 4a and 4b show coatings which have been
electrolytically etched with chromic acid to better
reveal the various features, i.e., the carbide
islands, matrix bonding-t~e -aggregate together, and
the laminar oxide "stringers". Figure 4a is a
photomicrograph of a fine powder 85~ chromium carbide
alloy composition, whereas Figure 4b is a coarse
powder, 85% chromium carbide composition. Although
the various microstructural elements are labeled in
Figure 4b, similar materials can be visually
confirmed in Figure 4a but arrows have not been
included in that latter picture due to the fine
detail presented therein.
The expected density of the coating resulting
from the coarse, -140 mesh, powder is in the range of
5.9 grams per cubic centimeter. The expected density
for the coating resulting from the fine, -325 mesh,
powder is substantially 6.1 grams per cubic
centimeter. To differentiate the fine powder
127353~ 17GE 2982
-14-
coating from the coarse powder coating, the
percentage of porosity or voids in the coating is
illustrative. The fine powder coating has a
metallographic porosity range of 4~ or lower. In
contrast, the coarse powder coating has a
metallographic porosity of 8% or greater. It is
recommended that the porosity acceptable for adequate
solid particle erosion resistance with the chromium
carbide coating be 44 or lower and the density be 6.1
grams per cubic centimeter or greater. A porosity of
3~ and a density of 6.2 g./cc with the chromiu~
carbide being at least 80% by weight percent of the
total powder provides a good solid particle erosion
resistant coating.
The matrix material used in the test herein was
nichrome. Nichrome is commonly recognized as a
composition of nickel, at about 80% by weigh~
content, and chromium, at about 20% weight content,
along with incidental trace elements such as carbon,
manganese, silicon and iron. It is recognized by
those skilled in the art that a matrix material is
necessary to bond the chromium carbide particles
together as an aggregate and to bond the aggregate to
the substrate. ~ence, the percentase of chromium
carbide particles has an effective upper limit of
about 95% since the matrix should be at least 5% of
the powder being sprayed. An important part of the
matrix material is the chromium content. It is
believed that 10% of the matrix material must be
chromium to confer adequate oxidation resistance to
the coating, to bond the chromium carbide crystalline
particles together in the aggregate, and to attach
the aggregate to the s~bstrate metal. Other matrix
17GE 2982
538
-15-
material alloys which may be utilized are based on
the transition metals Fe, Co and Ni. Typical
compositions are Fe10Cr, Fe20Cr and Co20Cr. Small
amounts of silicon or boron or other elements in the
matrix material alloy in combination with an
appropriate amount of chromium may also be utilized.
The method ~f coating the components subject to
the solid particle erosion may be altered by those
skilled in the art. Generally, a heat transfer
assisted process is necessary to properly coat the
substrate. In addition to plasma spraying the
blended powder onto the cleaned and prepared
substrate surface as discussed herein, detonation
spraying, flame spraying, diffusion coating and
plasma arc welding the powder can be utilized as the
heat transfer process. The process must adhere a
preferred 85% chromium carbide alloy composition to
the substrate and obtain a metallographic porosity of
4% or less and a density of at least 6.1g/cc and
preferably 6.2g/cc, thereby achieving good solid
particle erosion resistance.
The plasma spraying operation utilized, in the
test herein, argon gas as a primary effluent gas and
hydrogen as a secondary gas. As is well recognized
in the art, nitrogen gas may be used as the primary
effluent. However, argon gas generally produces a
chromium carbide coating which is more ductile than
that resulting from the use of nitrogen gas as a
primary gas in the plasma spray process.
The powder may be presintered or pre-alloyed
prior to coating the substrate by a heat transfer
process.
lZ735~ 17GE 2982
-1-6-
The particular components of a steam turbine
which are significantly affected by solid particle
erosion due to their presence in the direct steam
flow path include valve surfaces as well as the
blades of the turbine rotor and the nozzle partitions
between the axially spaced apart sets of turbine
blades. Another component in the steam turbine which
is affected by solid particle erosion is the
diaphragm which changes the flow direction of the
steam from a radially inward direction to an axial
direction. The use of an 80% chromium carbide, 20%
nichrome coating with no mnre than 4% porosity and a
density of at least 6.lg/cc would reduce ~he solid
particle erosion on the diaphragm surface exposed to
the steam flow.
The claims appended hereto are meant to cover
all modifications and equivalents readily apparent to
those of ordinary skill in the art, hence, the
specific method of coating is not meant to limit the
scope of the claims, nor is the specific discussion
of the turbine blades, nozzles, etc., so limiting
since all components exposed to extensive solid
particle erosion in the steam flow path of a turbine
are meant to be covered herein.
