JP2824165B2 - Double coating for various substrates - Google Patents

Abstract

A duplex coating for substrates comprises an undercoat of tungsten carbide-cobalt having a strain-to-fracture of greater than 4.3x10<-><3> inch per inch and a top coat of a ceramic material such as alumina and in which the ratio of the thickness of the top coat to the undercoat is from 6:1 to 1:3.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

FIELD OF THE INVENTION The present invention relates to a dual coating exhibiting good fatigue properties which includes a tungsten carbide-cobalt undercoat with a ceramic material such as alumina.

[0002]

2. Description of the Related Art Various rotary seals are commonly used in gas turbine engines of the type in which a rotating member cooperates with another member that is relatively stationary across a narrow gap. Such seals are sometimes used between a rotating member and a rotating shaft or drum to maintain a differential pressure within the chamber on each side of the seal. For example, in one type of gas turbine engine, a plurality of rows of rotating blades extend radially outward from a rotating shaft into a passageway for working medium gas. Secondly, a plurality of vanes extend radially inward from the stationary case or shroud. There are several types in which the stationary blade protrudes inward from the stationary case in a cantilevered manner. The wings are positioned to direct the working gas toward or away from the adjacent rotating blade. The vane has a sealing surface circumscribing the tip of the blade in each row of blades, and in the cantilever vane type, the rotor is provided with a sealing surface circumscribing the tip of the vane in each vane row. I do. As the clearance between the blade or wing tip in each row and the corresponding seal surface increases, a significant amount of working medium gas leaks over the blade and / or stator tips to the periphery and aerodynamics. Decrease efficiency. Moreover, increasing the clearance causes a greater amount of working medium gas to leak axially from the downstream end of the blade or rotor to the upstream end. Therefore, it is desirable to keep the clearance to a minimum. However, it is also necessary to accommodate the various dimensional changes that occur during initial startup, thermal processes, high G turns, and the like. Under these conditions, especially during engine startup, there is usually some wear-in of the part.

[0003] It is known that a more desirable condition is for the tip or knife edge to cut the groove into a corresponding sealing surface, rather than for the tip or knife edge to undergo wear. U.S. Pat. Nos. 4,238,170 and 4,239,452 provide an internal peripheral groove circumscribing the blade tip in the stator or shroud seal surface, but this arrangement presents alignment difficulties. At the same time, the thermally induced axial displacement of the blade cannot be adapted to the stator or shroud. A variety of rotating seal arrangements have been disclosed in the literature, in which rotating members are provided with softer, e.g., polished cooperating members, such as filled honeycombs, porous metals, brittle ceramics, and the like. And cutting or abrasion is reduced. It has been found that in some of these arrangements, improper sealing or seizure of the cooperating members can occur. In such other arrangements, local "hot spots" and burns of the non-abrasive member may occur. Examples of seals using abrasive members are disclosed in the following U.S. Patents: 3,068,016
No. 3,481,715; 3,519,282;
3,817,719; 3,843,278; 3,9
18,925; 3,964,877; 3,975,
No. 165; 4,377,371; 4,540,336
issue. The abrasive seal is adapted to flake or wear out in the event of a thermal flash or impact load that causes the blade tip to impact the seal. US Patent 4,377,
No. 371 points out that some materials used as abrasive seals are susceptible to massive delamination, which is propagated by the presence of cracks in the seal surface, and polishes the seal surface by using a laser beam. To produce a network of fine microcracks on the seal surface. I. E. FIG. Sumner and D.M. Ruckle is AIAA
/ SAE / ASME 16 times Joint Propul
The paper “Development of Improved-
Durability Plasma Sprayed
Ceramic Coatings for Gas
Turbine Engines ", AIAA-80
At -1193, segmented laser scanning coatings were reported to perform poorly.

[0004] British Patent Nos. 853,314 and 1,00
No. 8,526 discloses a turbine or compressor blade having ribs formed thereon to provide a seal with respect to the rotor or stationary shroud, wherein the ribs and associated sealing surfaces are removable when worn. ing. U.S. Pat. No. 4,148,494 discloses a gas having an abrasive tip consisting of an electrodeposited matrix of nickel or an alloy containing nickel, wherein confined abrasive particles, such as borazon particles, protrude from the tip. A turbine blade or blade is disclosed. Grindable tips of the type disclosed in this patent are difficult and extremely expensive to manufacture. U.S. Pat. No. 3,339,933 discloses a blade blade coated with bonded alumina that extends to a cooperating honeycomb member to form a seal. U.S. Pat. No. 3,537,713 discloses a blade protruding inward,
A rotating sleeve is disclosed which is coated with a hard protective material such as molybdenum or nickel aluminide which forms an alternative ridge and groove in place of the rub resistant material on the stationary cooperating member. U.S. Pat. No. 4,884,820 discloses that a blade of a blade coated with a ceramic or metal carbide coating is laser treated to produce a wear resistant cutting surface that can be cut into an abrasive material in a rotary gas seal. It is disclosed to form a plurality of laser formed depressions.

[0005] While the above coatings provide a good wear resistant cutting surface to the blade, sometimes the coating adheres well to the blade when used in a high temperature environment and does not remain adhered to the blade. For example, ceramic coatings such as alumina-based coatings provide good wear resistant cutting surfaces for many substrates, but have poor adhesion to substrates such as titanium alloys. However, when a substrate such as a titanium alloy is roughened by grit blasting with an abrasive such as alumina grit, good adhesion for the ceramic coating can be obtained. Unfortunately, grit blasting usually causes large fatigue debits on the substrate. Tired to the material,
A progressive fracture phenomenon that occurs in a material when a cyclic load is applied with a stress having a maximum value smaller than the tensile strength of the material.
Fatigue can usually eventually break after a sufficient number of cyclic loads. Since fatigue causes the material to fail faster and / or at lower loads than expected, its net effect either reduces the useful life of the material at the same load or decreases the allowable load during the same life. Met. Thus, the use of grit blasting is undesirable because it can cause fatigue debits in the substrate. U.S. Pat. No. 4,826,734 discloses a substrate coated with a tungsten carbide-cobalt coating having a strain to break greater than 4.3.times.10.sup.- ³ inches / inch. However, while such coatings provide a suitable wear resistant cutting surface, ceramic coatings typically provide a more desirable wear resistant cutting surface for use in compressor blade tips.

It is an object of the present invention to have good fatigue properties,
At the same time it is also to provide a double coating for the substrate, which gives a good wear-resistant cutting surface. It is another object of the present invention to provide a dual coating for a substrate such as titanium including a tungsten carbide-cobalt undercoat and a ceramic coating topcoat. Another object of the present invention is to provide a strain to break (str)
To provide a dual coating for a titanium-based substrate, including a tungsten carbide-cobalt undercoat having an ain-to-fracture greater than 4.3 × 10 ⁻³ inches / inch and a topcoat of an alumina-based coating. is there. The foregoing and further objects will become apparent from the description and disclosure hereinafter.

[0007]

SUMMARY OF THE INVENTION The invention includes a substrate such as a titanium alloy, wherein a tungsten carbide-cobalt undercoat layer is adhered to the substrate and a topcoat layer of a ceramic material such as an alumina-based material is undercoated. The present invention relates to a coated article fixed. Preferably, the tungsten carbide-cobalt undercoat is 4.3 ×
Greater than ^10-3 inches / inch, preferably 5.0x
^4. Greater than ^10.sup.-3 inches / inch, more preferably 5.
It should have a strain to break greater than 5 × 10 ⁻³ inches / inch. The compressive residual stress of the tungsten carbide-cobalt undercoat is preferably from about 30 to about 50 kilopounds per square inch (KSI) (2,100 to
3,500 kg / cm ² ). The undercoat layer preferably has a surface roughness of at least 100 microinches (2.5 microns) Ra so that the topcoat can be sufficiently secured to the undercoat layer. Preferably, the ratio of topcoat thickness to undercoat thickness should be between 6: 1 and 1: 3, more preferably between 3: 1 and 1: 2. The novel dual coating of the present invention provides an undercoat layer that has good adhesion to the substrate and good fatigue properties while the topcoat provides a surface with excellent wear-resistant cutting properties. Was found. The good fatigue properties of the undercoat effectively prevent cracks in the topcoat from spreading down to the substrate when the coated substrate is operated, especially in an operating state under cyclic loading. Thus, the dual coated substrates of the present invention survive longer under cyclic loading conditions than coated substrates having only a coating layer of ceramic material.

The undercoat of the present invention has good fatigue properties and cohesive strength and can be applied to a substrate by means such as a detonation gun process. The thickness of the undercoat layer must be sufficient to stop the propagation of the topcoat cracks and thus not significantly reduce the fatigue properties of the base. The thickness of the topcoat will be at least 2 to about 20 mils (0.051 to 0.51 mm) for most applications, preferably at least 3 to about 10 mils (0.076 to 0.21 mm)
Can be The undercoat surface should have a roughness of at least 100 microinches (2.5 microns) R to secure the topcoat as it is deposited thereon.
a, preferably at least 150 microinches (3.8
Micron) should have Ra. The topcoat of the present invention should have a good abrasion resistant cutting surface and should be able to adhere to the undercoat layer so that it does not peel when used in the intended operation. This topcoat can be applied to the undercoat using a conventional detonation gun process. The thickness of the topcoat should be sufficient for most applications to provide a good abrasion resistant cutting surface for the intended application. A thickness of at least 1 to about 20 mils (0.025 to 0.51 mm) is generally suitable, and a thickness of at least 2 to about 10 to 15 mils (0.051 to 0.051 mm).
To 0.25 to 0.38 mm). The tungsten carbide-cobalt undercoat layer should include about 7 to about 25% cobalt, about 0.5 to about 5% carbon, and about 70 to about 92.5% tungsten by weight. Cobalt is about 8% to about 18% by weight and carbon is about 2% to about 4%.
Weight percent and the tungsten is preferably from about 78 to about 90 weight percent. The most preferred coating should include about 9 to about 15% by weight cobalt, about 2.5 to about 4.0% by weight carbon, and about 81 to about 88.5% by weight tungsten. Inventive tungsten carbide-cobalt coating materials can include small amounts of chromium, preferably about 3 to about 6% by weight, and most preferably about 4% by weight. The addition of chromium will improve the corrosion properties of the coating.

The ceramic topcoat layer comprises: alumina; a composition of alumina and titania, chromia and / or zirconia (including alloys and mixtures); a composition of chromia and alumina; zirconia on silica, yttria; A composition in which calcia and / or magnesia are mixed; chromium carbide; The topcoat is preferably alumina and a composition of alumina and titania, chromia and / or zirconia, most preferably alumina.
The substrate can include titanium, steel, aluminum, cobalt, nickel, alloys thereof, and the like. The substrate is preferably a titanium alloy. If the substrate is a titanium alloy, the preferred top coating is an alumina-based material, such as alumina, and the undercoat layer is 7-25 wt% cobalt, 0.5-5 wt% carbon and 70-92.5 wt% tungsten. % Of tungsten carbide-cobalt. Undercoat thickness is 2-20
Mil (0.051-0.51 mm) and the topcoat thickness is 1-20 mil (0.025-0.51 m)
m). The ratio of topcoat layer to undercoat layer is usually 6: 1 to 3: 1. When using a titanium alloy substrate, the tungsten carbide-cobalt undercoat provides excellent adhesion to the titanium alloy without effectively altering the fatigue properties of the titanium alloy, and at the same time allows the top coat to be fixed. Has been found to result in a surface having Topcoats of alumina-based materials provide excellent wear resistant cutting surfaces that can be used in various applications without breaking under cyclic loading. To further enhance the surface of the topcoat, laser treatment can be performed to provide a plurality of depressions defined by high land areas.
Such depressions can act as a set of cutting edges. In addition, the depressions defined by the land areas are believed to enhance cutting performance by providing space for receiving fine cutting debris when the double coated substrate is used as a blade in a compressor. Upon cooling, the blade tip retracts and debris is released from the surface.

A preferred method of applying the undercoat is to flame the plate using a detonation gun and includes the following steps: introducing the desired fuel and oxidant gas into the detonation gun to form an explosive mixture. The powdered coating material is introduced into the explosive mixture in the gun and the fuel-oxidant mixture is exploded to strike the substrate coated with the coating material. The fuel-oxidant mixture preferably comprises a fuel mixture of at least two combustible gases selected from the group consisting of oxidants and saturated and unsaturated hydrocarbons. Preferably, the oxidant used is oxygen and the combustible gas mixture is acetylene gas and propylene gas. The powder of the coating material used to obtain the undercoat is preferably a powder made by a cast and crash process. In this method, the components of the powder are melted and cast into shell-shaped ingots. The ingot is then ground to obtain the desired particle size distribution. The resulting powder particles contain square carbides of various sizes. Various amounts of metal phase are associated with each particle. This morphology gives the individual particles non-uniform melting characteristics. In fact, under some coating conditions, some of the particles containing some of the larger horned carbides may not melt at all. Preferred powders are
Usually made to a size of 1 to 25 microns and W ₂ C, Co
A coating having a polished metallographic appearance consisting of approximately 2-20% of angular tungsten carbide particles distributed in a matrix consisting of a mixed carbide such as ₃ W ₃ C and a Co phase. The powder of the coating material used to obtain the top coat is preferably a powder made by firing a metal salt. The resulting powder particles are typically sized in the range of 1 to 45 microns. In testing the coated samples, various data were observed,
Some of the data was obtained using the test procedure described below.

The strain leading to the breakage of the coating in the strain test example leading to the break was determined using a four-point bending test. In more detail, Hardness Rockwell is made by quenching 4140 steel.
One side of a cross-section rectangular beam made on a C-scale (HRC) 40-45 is coated with the material to be tested.
Typical substrate dimensions are 0.50 inch (1.3c wide)
m), thickness 0.25 inch (0.63 cm) and length 1
0 inches (25 cm). The coating area is 0.
50 inches x about 7 inches (1.3 x 18 cm)
Center along the 10 inch length of the substrate. The coating thickness is typically 0.015 inch (0.38 mm), but the applicability of the test is that the coating thickness is 0.0
Unaffected by the range of 10 to 0.020 inches (0.25 to 0.51 mm). An acoustic transducer is attached to the sample using a masking tape and a coupler such as Dow Corning high vacuum grease. Acoustic Chick transducer has a narrow response band coordinated to it and approximately 90～120KH _z piezoelectric. Transducer, gain 3
Couple to a preamplifier with a fixed gain of 40 dB that passes the signal through an amplifier set to 0 dB. Thus, the overall system gain is 70 dB. The amplifier is coupled to a counter that counts the number of times the signal exceeds the 1 millivolt limit and outputs a voltage proportional to the total count. In addition, a signal proportional to the peak amplitude of the event is also recorded.

The coated beam is placed in a folding tool. The folding tool is designed to load the four-point bending beam so that the coating is in tension. The outer load point is 8 inches (20 cm) apart on one side of the beam and the midpoint of the load is 23/4 inches (7.0 cm) on the opposite side of the substrate. This test geometry places the central 23/4 inch of the coated beam into uniform stress. The two sets of load points are replaced with each other using a universal testing machine to produce a bend in the center of the test sample. The sample is bent such that the coating is convex, ie, the coating is placed in tension. The deformation of the sample during bending is monitored either by a load cell attached to a universal tester or by a strain gauge attached to the sample. When measuring loads, the strain in the coating is calculated using engineering beam theory. During bending, the acoustic count and peak amplitude are also recorded.
Data is collected simultaneously using a three pen chart recorder and calculator. If the coating cracks,
Accompanied by acoustic emissions. Thickness penetration (thr
Acoustic emission characteristics associated with an out-thickness crack include about 10 ⁴ counts per event and 100 dB peak amplitude for 1 millivolt at the transducer. The strain present at the onset of cracking is recorded as the strain leading to the breaking of the coating.

[0013] The residual stress of the coating in the residual stress test example, the blind hole (blin
d hole) test. The specific procedure is A
It is a revised version of STM Standard E-387. Specifically, a strain gauge rosette is adhered to the sample to be tested. The rosette used is Texas Measurements, College Station, Texas.
, Sold down by the gauge FRS-2. The device has three gauges oriented at 0, 90 and 225 degrees to each other and mounted on a wheel backing. The gauge has a centerline diameter of 5.12 mm (0.202 inch), a gauge length of 1.5 mm (0.0509 inch), and a gauge width of 1.4 mm (0.055 inch). The procedure for attaching a rosette to a sample is described in North Carolina,
Measurements Group In in Lori
c. , Published in Bulletin B-127-9. A metal mask is adhered to the strain gauge to help locate the hole during drilling and to protect the strain gauge during drilling. The mask has an annular shape with an outer diameter equal to 0.382 inches (9.70 mm), an inner diameter equal to 0.160 inches (4.06 mm), and a thickness of 0.0485 inches (1.23 mm). be equivalent to. This mask is positioned at the same center as the strain gauge using a 6 × microscope. When the mask is centered, apply a drop of glue to the edges, dry the mask in place,
Fix it. The three gauges are connected into three equal signal conditioners, producing a reading in the distortion unit. Before starting the test, all three units are adjusted to show a zero reading.

The test apparatus includes a plate capable of moving vertically and horizontally in one direction and having a rotating grit blast nozzle mounted thereon. Grit blast nozzles are available from Piscataway, New Jersey. S. It is made of White and has an inside diameter of 0.026 inch (0.66 mm) and an outside diameter of 0.076 inch (1.93 mm). The nozzle is offset from the center of rotation, so that the result is 0.096-. 1002 inches (2.44 to 2.545
mm). Place the sample to be pierced into the cabinet and place the strain gauge in the center below the rotating nozzle. Position adjustment of the part is achieved by rotating the nozzle without flowing either the polishing media or air, and manually adjusting the position of the sample so that the nozzle rotation is concentric with the mask. Set the standoff between the nozzle and the part to 0.020 inches (0.51 mm). The position of the plate is indicated by a stop. The abrasive used to make the holes is 27 micron alumina,
It is carried with 60 psi (4.2 kg / cm ² ) of air. Erosive or abrasive media is used at a rate of 25 grams / minute (gpm). The abrasive is dispensed with a conventional powder dispenser. Drill holes for 30 seconds, during which time the flow of abrasive and air is stopped. Move the nozzle away from it and release it. The positions at the top of the strain gauge and at the bottom of the hole are measured with a portable focusing microscope and the differences are recorded. The hole depth is the difference-the thickness of the strain gauge. The distortion released around the hole is indicated by the signal conditioner, and these values are also recorded. While recording the data, the sample does not move, so the nozzle can be returned to the original starting point and the test can continue. The test is repeated until the hole depth is greater than the thickness of the coating, at which point the test is stopped. The strain released at a given hole depth in the incremental layer is related to the stress in that layer using data from a calibration sample of mild steel loaded to an empirically known stress state. The residual stress is determined from this data.

The correlation between the strain leading to the breaking of the coating and the residual stress is as follows. If the material is under a combined load set, the stress and strain from each of the load conditions can be calculated, and the total stress and strain map can be determined by superimposing the stress resulting from each load. This must be applied to the coating to determine the residual stress in the coating in addition to the stress applied during the four-point bending test, as well as the actual stress of the coating when its fracture occurs. The four-point bending test is performed to place the coating in tension. That is, using the fact that stress and strain are correlated by a constant, the total stress in the coating at break is actually given by:

Σ _t = Eε _f + σ _r (Equation 1) σ _t = Stress applied E = Coating elastic modulus ε _f = Strain leading to fracture from 4-point bending test σ _r = Coating residue measured from blind hole test Stress (conventional compressive stress is negative).

The coating will usually crack at a constant stress value, regardless of whether the stress is the result of residual or applied stress or a combination of the two. A coating having a predetermined compressive residual stress,
Before placing the coating in tension, an equal amount of applied tensile stress must be applied. Rearranging equation (1) and expressing strain to failure as a function of residual stress (Equation 1), it is clear that an increase in compressive stress in the coating results in an increase in strain to failure of the coating.

(Equation 2) Thus, the stress or strain that can be applied before the coating breaks is affected by the amount of residual stress or strain present in the coating. More information on blind hole testing to measure residual stress can be found in Residual Stress InDesign, published by ASM International, Metal Sparks, Ohio.
n, Process and Materials S
It can be found in the publication entitled selection. This publication contains the same title of ASM C, April 27-29, 1987 in Cincinnati, Ohio.
L. on reference. C. Contains a paper published by Cox. The disclosure of this article is incorporated herein by reference.

Adhesion Strength Test The adhesion strength of the coating to the substrate was determined according to ASTM 633.
The coating layer was thinner, as determined using the method disclosed in US Pat. Specifically, a series of six cylindrical adhesive specimens, each having a diameter of one inch (2.5 cm), were applied. The surface of the coating layer was ground with a diamond wheel to provide a smooth surface perpendicular to the axis of the cylinder. Next, each of the cylindrical test pieces was subjected to a trade name “SCOTCH-WELD EPOXY” by 3M.
ADHESIVES 2214 NON-METAL
It was fixed to a butt-unmatched (unmatched) uncoated cylindrical test piece using an epoxy adhesive sold by LIC. The epoxy adhesive is composed of: component percent epoxy resin blend 70.0-80.0 aliphatic glycidyl ether 1.0-10.0 nitrile latex 1.0-10.0 dicyandiamide 1.0-10. 0 3-p-chlorophenyl) -1,1-dimethylurea 1.0-10.0 Amorphous silicon dioxide 1.0-10.0 Water 1.0-10.0 After curing epoxy, cylindrical test piece Were separated to break, and the load at break was observed. After this test, each coated cylindrical specimen was inspected to determine if a break had occurred in the coated or epoxy bonded area.

Fatigue Test The fatigue test specimen was machined to a length of 3.5 inches (8.9).
cm) (the segment at each end is 0.8 inches (2.0 cm) in length) and was a cylindrical specimen having a diameter of 0.6 inches (1.5 cm). The interior portion of each end segment is tapered inwardly to a length of 0.75 inches (1.9 cm) and a diameter of 0.2
A center segment having 5 inches (0.64 cm) was produced. The central segment is called the gauge section of the test bar. The coating is applied to the gauge section and outwardly tapered segments using a detonation gun. The coatings are tested "as-coated" which retains the natural surface roughness of the coating. The fatigue test is performed as an axial tension-tension test with a minimum stress ratio R0.1 to a maximum stress. The test is a simple load-adjustable axial fatigue tester in air at ambient temperature.
Perform at 30 Hz. For all specimens, the maximum stress was derived as the maximum load across the gauge cross section of the substrate material. In each test, the test piece was placed in the gauge section and fractured (gauge section failure = GS
F) or up to 10 million cycles (runout = R
O). Some examples are provided below to illustrate the invention. In these examples, the coating was as shown in Table 1.
Using the powder composition shown in Table 1.

[0019]

[Table 1]

[0020]

EXAMPLE 1 A gaseous fuel-oxidant mixture of the composition shown in Table 2 was
Each was introduced into a detonation gun to form an explosive mixture having the oxygen to carbon atom ratios shown in Table 2.

[0021]

[Table 2]

The sample coating powders shown in Table 2 were also supplied to the detonation gun. Each gaseous fuel
The oxidant mixture flow rate, the flow rate of each coating powder, the volume percent of the gaseous fuel-oxidant mixture and the oxygen to carbon atomic ratio for each coating sample are also shown in Table 2. The coating sample powder was fed to the detonation gun simultaneously with the gaseous fuel-oxidant mixture. The detonation gun was ignited at a rate of about eight times per second and the detonation gun coating powder was bombarded against various steel and titanium substrates to form a dense adhesive coating of shaped microscopic leaves that entangled and overlapped each other. % By weight of cobalt and carbon in the coating layers of sample coatings 1 and 2
Was determined along with the hardness of the coating. The hardness of the sample coatings 1 to 3 in Table 2 was measured using a Rockwell surface hardness tester, and the Rockwell hardness value was converted to a Vickers hardness value. The Rockwell surface hardness method used is according to ASTM standard method E-18. Hardness is measured on a smooth, flat surface of the coating applied to the quenched steel substrate. Rockwell hardness values were converted to Vickers hardness values according to the following formula: HV _0.3 = -1774 + 34.433HR45N where HV _0.3 represents the Vickers hardness value obtained at _0.3 kgf load, HR45N represents the diamond indenter and 45 kgf load. Represents the Rockwell surface hardness value obtained on the N scale. The value of strain and the value of residual stress leading to fracture were obtained as described above, and the obtained data are shown in Table 2.

Example 2 The end of a 1 inch (2.5 cm) diameter Ti-6Al-4V test rod was coated using the same method and powder composition used for sample coating 3 of Example 1. Then, each cylindrical rod was SCOTCH-WEL
The mating rod was secured along the axis using D epoxy adhesive. After the epoxy was cured, the two cylindrical rods were pulled apart to break and the load at break was recorded for each sample. The test bars were then inspected to determine if the break occurred at the coated bond area (adhesion of the coating to the substrate-referred to as CSD) or at the epoxy bond area (epoxy). Table 3 shows the thickness of the coating, the adhesion strength to failure and the type of failure observed.

[0024]

[Table 3]

Example 3 The end face of a 1 inch (2.5 cm) diameter Ti-6Al-4V test bar was coated using the same method and powder composition used for sample coating 2 of Example 1. In addition, a topcoat was applied using the same method and powder composition used for sample coating 3 of Example 1. Each double-coated cylindrical rod was then replaced with S
It was secured to the mating bar along the axis using a COTCH-WELD epoxy adhesive. After the epoxy was cured, the two cylindrical rods were pulled apart to break and the load at break was recorded for each sample. The test bars were inspected to determine if the break occurred at the coated bond area (CSD) or at the epoxy bond area (epoxy). Table 4 shows the thickness of each coating layer, the adhesion strength to breakage, and the type of breakage observed.

[0026]

[Table 4]

There was some minimal edge chipping of the coating. The average adhesion strength of the dual coating of the present invention was much better than the average adhesion strength shown in Example 2.

Example 4 The end face of a 1 inch (2.5 cm) diameter Ti-6Al-4V test bar was coated using the same method and powder composition used for sample coating 1 of Example 1. In addition, a topcoat was provided using the same method and powder composition used for sample coating 3 of Example 1. Each double-coated cylindrical rod was then replaced with S
It was secured to the mating bar along the axis using a COTCH-WELD epoxy adhesive. After the epoxy was cured, the two cylindrical rods were pulled apart to break and the load at break was recorded for each sample. The test bars were inspected to determine if the break occurred at the coated bond area (CSD) or at the epoxy bond area (epoxy). Table 5 shows the thickness of each coating layer, the adhesion strength to breakage, and the type of breakage observed.

[0029]

[Table 5]

In some cases, there was some minimal edge chipping of the coating. The average adhesion strength of the dual coating of the present invention was much better than the average adhesion strength shown in Example 2.

Example 5 The center or gauge section of a fatigue test bar made of a Ti-6Al-4V alloy made as described above was coated with the coating composition of Sample Coating 1, 2 or 3 of Example 1. Coating was performed using the process and gaseous fuel mixture as indicated for sample coatings 1, 2, or 3, respectively. The gauge portion of the further fatigue test bar was coated with an undercoat of the sample coating 1 or 2 coating composition, and the coating composition of sample coating 3 of Example 1 was deposited thereon. These test bars were coated with the coating compositions of Sample Coating 1, 2, or 3 of Example 1 at various thicknesses as shown in Table 6. The test rods were not grit blasted prior to coating, but only coated the test rods with sample coating 3 of Table 2 of Example 1, which were grit blasted with 60 mesh horn particles of alumina. A good adhesion strength between No. 3 and the test bar was promoted. The coated fatigue bars were subjected to a fatigue test as described above, together with several uncoated fatigue bars.
Table 6-8, when the maximum stress and bar indicates the number of cycles endured before breaking tired in gauge portion or the rod individual bars are each subjected does not break after ¹⁰⁷ cycles,
The test was interrupted and marked as runout.

[0032]

[Table 6]

[Table 7]

[Table 8]

The data from Table 3 of Example 2 show that all of the test bars coated with Al ₂ O ₃ broke at the coated bond area, whereas the data from Table 4 of Example 3 showed double data according to the invention. Test bars coated with the coating failed in the epoxy bonded area and did not break in the coated adhesive area.
These tests, Al ₂ O ₃ coating, when used as a single coating, did not effectively adhere to the substrate, tungsten carbide as shown in Example 3 -
Demonstrate effective adhesion to the cobalt undercoat. This makes the Al ₂ O ₃ coating a good wear-resistant cutting surface for the blade, but when used in the intended application, it sometimes fails to adhere to the blade enough to remain adhered to the blade. is there. However, a substrate such as a blade is first coated with a tungsten carbide-cobalt undercoat,
Next, if an Al ₂ O ₃ coating is to be applied,
_{The 2} O ₃ coating adheres to the undercoat layer and provides good adhesion therewith. This can be achieved without the need for grit blasting the substrate, which can cause significant fatigue debit on the substrate. The data in Table 2 shows that coatings applied using a combination of two flammable gases are better than coatings applied using a conventional detonation gun process using only one flammable gas. This shows that the strain and residual stress that led to the fracture were brought about. Table 6-8
The test rod 16 is simply made of a kind of flammable gas using a conventional detonation gun process and has a
1.86 × 10 ^{5 at} KSI (3,500 kg / cm ² )
Cycle survived. Also, a double coating layer of the same coating composition and thickness was applied using a detonation gun process using two flammable gases (test bar 18), which was tested at 1.50 stress at 50 KSI.
It lasted 4166 × 10 ⁷ cycles. Thus, the process used to apply the undercoat to the substrate should be chosen to ensure that the tungsten carbide-cobalt undercoat has a strain to failure greater than 4.3 × 10 ⁻³ inches / inch. is there. Although the invention has been described with respect to specific details thereof, these details should not be considered as limiting the scope of the invention.

Continuation of the front page (56) References JP-A-3-53055 (JP, A) JP-A-52-124010 (JP, A) JP-A-61-26786 (JP, A) JP-A-1-272782 (JP, A) JP-A-56-38403 (JP, A) JP-A-2-156060 (JP, A) JP-A-4-141570 (JP, A) JP-A-55-54105 (JP, A) 2-30704 (JP, A) JP-A-2-70074 (JP, A) Welding technology, 36 [4] (1987), (Showa 62-4-1), Sanho Publishing Co., Ltd., p. 6) Surveyed field (Int.Cl. ⁶ , DB name) C23C 28/00-28/04 B32B 18/00 C23C 4/00-6/00 B32B 15/08

Claims (6)

(57) [Claims] 1. A strain 4.3 × 10 ^-3 inch / inch larger tungsten carbide than to fracture - comprises a topcoat cobalt undercoat and the ceramic material, the ratio of the thickness of the topcoat pair undercoat 6: 1 to 1: 3 der Ri; tungsten carbide - Koba
The rut undercoat has a cobalt content of 7 to 25% by weight,
Elemental content 0.5-5% by weight and tungsten content 70-
92.5% by weight, roughness at least 2.5 microns
(100 micro inches) Ra and thickness at least 0.0
51 mm (2 mil); top coat has low thickness
0.025 mm (1 mil) der Ru even without, duplex coating for use on substrates. 2. The ceramic material comprises alumina; a composition of alumina and at least one member selected from the group consisting of titania, chromia and zirconia; a composition of chromia and alumina; zirconia and magnesia, silica, yttria and calcia. 2. The dual coating of claim 1, wherein the composition is selected from the group consisting of: a composition with at least one member selected from the group; and chromium carbide. 3. The undercoat contains 6% by weight of chromium.
2. The dual coating of claim 1 containing 4. An undercoat having a thickness of 0.051-0.
Duplex coating of Claim 1 which is 76 mm (2 to 30 mils), and and topcoat thickness 0.025~0.51mm (1~20 mils). 5. The dual coating of claim 1 for use on a substrate selected from the group consisting of titanium, steel, aluminum, cobalt, nickel and alloys thereof. 6. The dual coating of claim 5 , wherein the substrate is a coated blade of a chip.