JP2007303956A - Technique for noncontact and nondestructive inspection on thermal barrier coating which has deteriorated with age - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To develop a technique for noncontact and nondestructive inspection on degradation with age of thermal barrier coatings of heat-resistant metal materials used for advanced gas turbines etc. <P>SOLUTION: High-frequency transmission characteristics are used to perform quantitative, noncontact, and nondestructive evaluation of the generation and growth of thermo-growth oxides (for example, formed in an interface between a Thermal Barrier Coating (TBC) and a bond coating) which occur in the thermo barrier coating of a heat-resistant material having the thermo barrier coating applied onto a metal substrate. It is thereby possible to grasp the state of degradation prior to the occurrences of peeling and coming-off of the coating, replace parts at appropriate timings, and achieve excellent action effects in safety and economy. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for non-contact and non-destructive evaluation of aged deterioration of a heat shielding coating in a heat-resistant metal material. The present invention relates to a technique for non-contact and non-destructive measurement of the formation / growth of thermally grown oxide generated in a heat-resistant material having a heat shielding coating applied on a metal substrate.

In recent years, the global warming issue has been taken up as one of the most important issues in the world. The global warming issue is a problem in which the atmospheric concentration of greenhouse gases such as carbon dioxide, methane, nitrous oxide, and alternative chlorofluorocarbons has increased in the atmosphere, which has led to an increase in temperature on a global scale. Along with this, sea level rises, changes in precipitation distribution, etc., and there are concerns about the impact on the ecosystem, society / economy, and living environment. About 70% of greenhouse gases are carbon dioxide (CO ₂ ), of which about 80% is said to be generated by fossil fuel consumption. In thermal power generation that uses a large amount of fossil fuel, steady efforts are being made to improve its efficiency. As one of the improvements in efficiency, gas turbines are introduced in power plants and turbine inlet gas temperature (TIT)
Efficiency has been improved, for example, by raising the temperature of the turbine to 1100 ° C, 1300 ° C, and 1500 ° C turbine inlet gas temperatures. In order to withstand such harsh thermal conditions, it was the advancement of high-temperature members used in gas turbine combustors, turbine blades, etc. that contributed to the increase in gas temperature along with air cooling technology. For the turbine blade material, etc., Co-base and Ni-base superalloys with excellent high-temperature strength are used, and the cooling performance is further improved. In part of the 1300 ° C class and 1500 ° C class, moving blades and combustors, etc. Thermal barrier coating or thermal barrier coating (TBC) is applied to the surface of Thermal barrier coating technology is applied for the purpose of preventing the temperature rise of the metal substrate without increasing the cooling air. FIG. 1 shows a structural conceptual diagram of a thermal shielding coating system (TBC system).

In general, the TBC system applies a bond coat on a Ni-base superalloy by low pressure plasma spraying (LPPS), and further, on it, atmospheric pressure plasma spraying (APS) or electronic plasma spraying. The top coat was applied by the electron beam-physical vapor (EB-PVD) method. Recently, high-speed flame spraying (HVOF) using pressurized combustion gas is also used as a thermal spraying method for bond coats. Compared with LPPS, the spraying temperature is lower and the sprayed particles are faster, so the coating formed Is dense with little oxidation and high adhesion to the substrate. The role of the top coat is heat shielding, and yttria partially stabilized zirconia having a low thermal conductivity is used as the material. The bond coat is used to improve the adhesion between the top coat and the base material, and the MCrAlY alloy (M = Co and / or Ni) is adopted. Table 1 shows the role of each element contained in MCrAlY used in the bond coat. M is a basic component, and Co, Ni, or a combination thereof is used to ensure strength at high temperatures or resistance to sulfidation and oxidation. Al forms Al ₂ O ₃ having a good oxidation resistance, Cr compensates the sulfidation corrosion of the Al ₂ O _3, to form a well Cr ₂ O ₃ itself. Y has the role of reinforcing and maintaining the protective oxide film of Al ₂ O ₃ and Cr ₂ O ₃ .

The TBC system can prevent the substrate temperature from rising or increase the usable temperature limit of the substrate. However, when used for a long time in a high-temperature environment, as shown in Fig. 1-7, if the TBC falls off due to peeling or cracking in the TBC, the substrate is exposed to an environment exceeding the service temperature and destroyed. There is a risk that it will eventually lead to a major accident. Therefore, many studies have been conducted on the occurrence of damage such as peeling, cracking or dropping in the TBC system to ensure the reliability of the gas turbine system.
One of the major factors involved in TBC damage is Thermally Grown Oxide (TGO). TGO is generated and grows at the topcoat / bondcoat interface by high temperature oxidation of the bondcoat during gas turbine operation. TBC debonding, cracking and detachment can be referred to as damage to the TBC system, and TGO growth that occurs in the TBC can be referred to as deterioration of the TBC system. TGO growth is thought to promote TBC damage, so suppressing TGO growth is an extremely important issue for improving the reliability and extending the life of high-temperature components in gas turbines. Detailed studies are being carried out on this and its suppression. Research has also been conducted on the correlation between the microstructure of TGO and the lifetime of TBC. TGO in TBC has a long crystal grain size, a small crystal grain size, a uniform layer thickness, and pores. There are reports that there are few.

There is also a report that there is a correlation between the amount of TGO growth and the peel strength of TBC. Ceramic top coat and TGO have a large coefficient of thermal expansion (CTE) value compared to metal bond coat, so a large thermal stress is generated at the interface due to temperature change at start-stop. In particular, since TGO has a small CTE, the thermal stress at the interface between TGO and the bond coat significantly increases due to the growth of TGO. Actually, a lot of delamination is observed at the interface between TGO and bond coat, but this is due to thermal stress at the interface between TGO and bond coat, and other causes are thought to be due to stress caused by volume expansion during the TGO growth phase. It has been. There are reports that the thickness of TGO is an important dominating parameter of damage, stating that the generation of thermal stress due to this TGO layer causes a decrease in interfacial bonding force. Furthermore, if the TGO thickness exceeds a certain value, an extreme decrease in adhesion strength occurs near the TGO / bond coat interface, and the limit film thickness is 5.5-6 μm, 8-10 μm or 13 μm. There are also reports.
Therefore, the growth amount of TGO may be an important index for knowing the degree of damage of TBC. In addition, there is a report that damage occurs in a short time at the final stage of life in TBC constructed by EB-PVD. Therefore, TBC degradation assessment focusing on TGO growth is important for TBC safety and reliability. It is considered to be extremely important for ensuring safety.

Nondestructive inspection methods of TBC include infrared thermography, ultrasonic pulse method, and radiation scattering method (Non-patent document 1: Yoshiya Sawada; Inspection and evaluation of ceramic thermal barrier coating, Non-destructive inspection Vol. 51, No. 7, pp.392-396 (2002); Non-Patent Document 2: Committee on Coatings for High-temperature Structural Materials; Coatings for High-Temperature Structural Materials, National Academy Press, pp.39-43 (1996)). Eddy Current Teting (ECT) is a corrosion resistant coating metal material (CoNiCrAlY) that is equivalent to the TBC bond coat.
Can detect notches with a depth of 0.1 mm, measure corrosion-resistant coating thickness, and metal temperature (Non-patent document 3: Hiroyuki Fukutomi, Akiyoshi Nomoto, Takashi Ogata; Report from the Central Research Institute of Electric Power Industry, T01045 ( 2002)). The ECT method cannot detect cracks in the topcoat ceramic material, which is an insulating material with low electrical conductivity. However, if a crack has reached the bond coat or base material that is a metal, the crack that has reached the metal part is not detected. It is possible to detect and measure the depth from the boundary between TBC and bond coat to the crack tip.

By the way, as mentioned above, it is important to pay attention to the increase in the thickness of TGO, but as a non-destructive inspection method for the increase in the thickness of TGO, the impedance spectroscopy (IS) method is used. Yes (Non-Patent Document 4: Kazuhiro Ogawa, Tetsuo Shoji,
Iwao Abe, Hideo Hashimoto; Material Evaluation, Vol. 58, No. 31, pp.476-481 (2000)). With the IS method, it is possible to quantitatively evaluate the thinning amount of TBC by detecting the magnitude of impedance, phase angle, etc. that change with changes in volume resistivity, dielectric constant, and thickness. There is an advantage that the thickness of the can be quantitatively evaluated. However, as a problem, it is pointed out that it is difficult to detect structural changes such as fine cracks and delamination. This is a partial discharge method that can be used in combination with this IS method (Non-patent document 5: Takashi Watanabe; Study on development of damage assessment method for thermal barrier ceramic coating using partial discharge method 2003 Tohoku University Master's thesis (2004) ) Is expected to enable detection. However, the IS method and the partial discharge method are thought to be effective for detailed monitoring and observation of the growth behavior of TGO, but the measurement requires that the electrode be in contact with the TBC topcoat surface where the surface is uneven. In addition, there are practical concerns that the size of the electrode and the contact area of the electrode greatly affect the measurement results.
Therefore, when considering application to actual equipment, the development of a measurement method capable of measuring changes in TGO film thickness at high speed without contact is eagerly desired.

Sawada Yoshiya; Inspection and evaluation of ceramic thermal barrier coatings, Non-destructive inspection Vol.51, No.7, pp.392-396 (2002) Committee on Coatings for High-temperature Structural Materials; Coatings for High-Temperature Structural Materials, National Academy Press, pp.39-43 (1996) Hiroyuki Fukuzaki, Akiyoshi Nomoto, Takashi Ogata; Report of Central Research Institute of Electric Power, T01045 (2002) Kazuhiro Ogawa, Tetsuo Shoji, Iwao Abe, Hideo Hashimoto; Material Evaluation, Vol. 58, No. 31, pp.476-481 (2000) Takashi Watanabe; Development of damage assessment method for thermal barrier ceramic coating using partial discharge method 2003 Tohoku University Master's thesis (2004)

At present, maintenance and management of TBC construction parts are carried out based on the manufacturer's empirical or engineering management standards. However, TBC repair and TBC construction parts disposal are not necessarily based on the remaining life of parts, and unfortunately scientifically reasonable criteria have not been established. Therefore, establishment of a rational maintenance management method and non-destructive inspection method for gas turbine main high-temperature parts is required from the viewpoint of ensuring reliability and reducing repair costs.
In addition, metal materials with a thermal barrier coating (TBC system) applied to metal substrates (
In other words, in gas turbines that use refractory metal materials), it is important to evaluate the deterioration status of the thermal barrier coating (TBC system) that is being constructed at a stage prior to damage such as peeling or dropping. is there. At this stage, the change in the thickness of the thermally grown oxide layer (TGO) is considered to be an important indicator of the deterioration process of the TBC system. Accordingly, an object of the present invention is to develop a non-contact non-destructive evaluation technique applicable to TBC, and to provide a method and a system / apparatus capable of quantitative evaluation of the TGO film thickness.

In recent years, the application of microwaves or millimeter waves has been studied for the purpose of inspecting insulating materials such as ceramics and polymers, and is expected to be used in various fields. In principle, these are non-contact measurement methods using an electromagnetic field. A microwave is an electromagnetic wave in a frequency band of 300 MHz to 300 GHz (wavelength 1 m to 1 mm), and a wave in a frequency range of 30 to 300 GHz is also referred to as a millimeter wave. Microwaves can propagate well in the air, and the frequency response depends on the electromagnetic property values of the irradiated material, so that non-contact evaluation is possible. Material evaluation by microwaves can be broadly divided into two fields: inspection of defects, etc. and measurement of material and physical properties. The present inventor has focused on the latter application and has so far applied to the evaluation of artificial defects and natural defect shapes in ceramic materials, and has found the possibility of quantitative evaluation of defect shapes (Mikiko Suzuki, Kazuhiro Ogawa, Tetsuo Shoji; The Third US-Japan Symposium on Advancing Applications and Capabilities in NDE, pp.394-400, The American Society for Nondestructive Testing, Inc. Hawaii (2004)) for the purpose of measuring material changes such as TBC system deterioration. As a result of intensive research, we succeeded in quantitative evaluation of TGO film thickness in the thermal shielding coating (TBC system) and deterioration evaluation of TBC system by nondestructive inspection method using high frequency transmission characteristics, and completed the present invention. did.

The present invention provides the following.
[1] A nondestructive evaluation system characterized by noncontact and nondestructive evaluation using a high frequency of 1G to 100 GHz for quantitative evaluation of an oxide film formed on a metal substrate.
[2] Non-destructive evaluation system characterized by non-contact and non-destructive evaluation of generation / growth at the interface of bond coat such as MCrAlY formed on metal substrate and ceramic coating formed thereon .
[3] Nondestructive evaluation sensor characterized by evaluating signals at oxide film or coating interface in response to high frequencies from 1G to 100GH.
[4] Non-contact and non-destructive measurement of the formation and growth of thermally grown oxide formed in the heat-resistant material having a heat shielding coating applied on the metal substrate using high-frequency transmission characteristics. Thermal shielding coating inspection method characterized by

In the present invention, it is possible to quantitatively measure and grasp the degree of deterioration in the TBC system of non-destructive and non-contact members, for example, the generation and growth of TGO, using high-frequency transmission characteristics.
Therefore, it contributes to the establishment of rational maintenance management standards such as gas turbines used in harsh environments. By using the inspection of the present invention, it is possible to perform quantitative life evaluation and nondestructive evaluation for using expensive parts to an appropriate life, and the technology of the present invention performs quantitative nondestructive evaluation. It also gives the advantage that the material can be used up to its original life, such as a guarantee that it can be reasonably continued without repair or disposal.
Other objects, features, excellence and aspects of the present invention will be apparent to those skilled in the art from the following description. However, it is understood that the description of the present specification, including the following description and the description of specific examples and the like, show preferred embodiments of the present invention and are presented only for explanation. I want. Various changes and / or modifications (or modifications) within the spirit and scope of the present invention disclosed herein will occur to those skilled in the art based on the following description and knowledge from other parts of the present specification. Will be readily apparent. All patent documents or references cited herein are cited for illustrative purposes and are not to be construed as an integral part of this specification. is there.

The nondestructive inspection method using the high-frequency transmission characteristics used in the present invention (hereinafter referred to as the high-frequency transmission characteristics method) applies a high-frequency signal to the surface of an object and reflects it as defined by the voltage amplitude ratio between the incident wave and the reflected wave. This is a method for measuring the ratio. In addition to the advantage of non-contact measurement, the high-frequency transmission characteristics method handles electromagnetic signals, so data processing is simple and automatic operation is possible, so it does not depend on the technical level of the measurer, or the structure of the probe There is an advantage that it is simple.
In the present invention, microwaves are used. A microwave is an electromagnetic wave having a relatively short wavelength, a wavelength of about 30 cm to 3 mm, and a frequency band of 1 to 1000 GHz. Microwaves, like light, have properties such as straight travel, reflection, refraction, absorption, and scattering, and use these properties to measure water absorption and thickness of non-metals, voids and porosity in plastics and ceramics. It is used for inspection of inclusions, defect inspection of composite material and honeycomb material joints, and the like. Since microwaves can propagate in the air, they do not require a contact medium like ultrasonic waves, but they have mostly been used in non-metallic materials so far because they hardly penetrate into metals due to the skin effect. .

The TBC system to be inspected according to the present invention has a top coat layer (ceramic) /
It is composed of TGO layer (oxide) / bond coat layer (metal) / base material (Ni-base superalloy) generated during use, and the thickness of TGO grown on the bond coat is directly subject to inspection. is there. Therefore, in principle, when microwaves are irradiated from the TBC surface side, the microwaves are transmitted through the top coat, which is a ceramic, and reflected at the bond coat interface, which is a metal. When TGO grows, TGO is a kind of ceramics, so basically the same phenomenon occurs,
Since the electrical impedance is different between the topcoat and TGO, the reflected wave changes depending on the growth of TGO, which allows measurement.
In the present invention, the base material of the TBC system includes a heat-resistant alloy, for example, an alloy having a substantial proportion of nickel (Ni) as a base, such as a Ni-base alloy or a Co-base alloy, or cobalt (Co) as a base. And alloys included in the group known in the art as “super alloys”. The term “superalloy” is a technical term generally used to describe something that has very high strength, excellent mechanical properties and corrosion resistance, and a typical superalloy is a unique It is recognized that it has a stable microstructure. Ni-based alloy (Ni) for components such as gas turbines for power generation in thermal power plants
Base alloy) is widely used, and can be cited as a representative example of a base material to which a thermal shielding coating (TBC) is applied in the present invention. The Ni-based alloy has excellent oxidation resistance at high temperatures, excellent high-temperature corrosion resistance, and high mechanical properties at high temperatures. A typical example of the Ni-based alloy is ASTM (American Society for Testing and Materials) standard, for example, ASTM.
B166, ASTM B167, ASTM B168, ASTM B751, ASTM B775, ASTM B829, etc., ISO (International Organization for Standardization) standards, for example, ISO 6207, ISO
JIS (Japanese Industrial Standards) standards such as 6208, ISO 9723-9725, etc., for example, JIS G4901, JIS G4902, etc. can be mentioned, and the product INCONEL 601 (INCONEL or Inconel is a registered trademark of Special Metals Corporation) Are commercially available. In addition to this Ni alloy, INCONEL 600, INCONEL 601GC, INCONEL 617, INCONEL 625, INCONEL 625CF, INCONEL 718, INCONEL X-750, INCONEL 751, INCONEL MA758,
Examples include INCONEL HX, INCOLOY MA956, INCOLOY DS (INCOLOY or Incoloy is a registered trademark of Special Metals Corporation), and examples include ASTM B163, JIS G4903, and JIS G4904.
The alloy composition (wt%) of a typical Ni-base alloy is as follows:
Ni: 58.0-63.0 wt%, Cr: 21.0-25.0 wt%, Al: 1.00-1.70 wt%, Cu: ≦ 1.0 wt%,
Si: ≦ 0.50 wt%, Mn: ≦ 1.00 wt%, C: ≦ 0.10 wt%, and the balance is Fe. Here, the balance of Fe is the amount of Fe excluding impurities accompanying trace amounts. Means. In a typical example, S: ≦ 0.015 wt%.

In the present invention, TBC is an MCrAlY (M is Ni, Co or their alloys) alloy called a bond coat on a superalloy base material such as a Ni-base superalloy base material. High-pressure flame spraying (HVOF) is applied to an appropriate thickness, and ceramics such as yttria stabilized zirconia (YSZ: Yttria Stabilized Zirconia) is used as the top coat. APS formed by applying an appropriate thickness by APS (Air Plasma Spray) or Electron Beam Physical Vapor Deposition (EB-PVD) is included. Typically, the adhesive layer (bond coat layer) that is a bond coat is applied to a thickness of about 100 μm, and the heat shield film layer (heat shield coating layer) that is a top coat has a thickness of 250 μm.
It was constructed to about ~ 300μm.

In the present invention, MCrAlX ¹ alloy (which compensates for oxidation resistance and hot corrosion resistance and exhibits excellent oxidation resistance from the viewpoint of improving adhesion with the top coat layer constituting the thermal barrier coating (TBC)) ( M is one or more metals selected from Fe, Ni and Co, and X ¹ is one or more metals selected from Y, Hf, Ta, Cs, Pt, Zr, La and Th) Is used as a basic bond coat material, and further, an adhesive layer (bond coat layer) composed of an alloy for bond coat blended with a material selected from the group consisting of Ce and Si, It may be formed on the surface of a heat-resistant alloy such as the Ni-base superalloy and further formed with a top coat having excellent heat shielding properties on the adhesive layer (bond coat layer). Generally the alloy for the bond coat of the basic are those having a composition of _{MCr 15 ~ 30 Al 5 ~ 16} X 1 0.1 ~ 1 ( subscript numbers indicate wt% elemental), typically M Are Ni and Co, and X ¹ is yttrium (Y). Typically, the basic bond coat alloy is a CoNiCrAlY alloy whose basic chemical composition is Ni: 29-35 wt%, Cr: 19-23 wt%, Al: 7-9 wt%, Y: 0.4 to 0.6 wt%, the balance being Co (where the balance is Co, excluding impurities accompanying trace amounts)
Means quantity). More preferably, the CoNiCrAlY alloy is Ni: about 32 wt%, Cr: about 21 wt%, Al: about 8 wt%, Y: about 0.5 wt%, the balance being Co (where the balance of Co Is the amount of Co excluding the accompanying impurities in trace amounts). Conveniently, a commercially available alloy for thermal spraying is used. For example, an alloy having a composition of 38.5Co-32Ni-21Cr-8Al-0.5Y is used as an alloy fine powder. Other basic bond coat alloys are widely known in the art, and are disclosed, for example, in US Pat. No. 3,928,026.
In the following, focusing on the reflection characteristics of microwaves irradiated to the TBC system, it is explained that it is possible to non-destructively inspect TGO film thickness changes by quantitatively examining the changes in the reflected waves accompanying the growth of TGO. To do. In particular, the non-destructive inspection method of the present invention uses a network theory used for high-frequency signal transmission characteristic analysis, and uses S-parameters that describe signal propagation characteristics.

[Measurement principle of high-frequency transmission characteristics method]
[Reflection coefficient]
In FIG. 3, the impedance on the transmitting side is Z ₀ , the voltage applied to the load impedance Z _L on the receiving side is V _L , the current flowing into the load is I _L , the incident voltage is V ₁ , the reflected voltage is V ₂ , and the incident voltage is I _1. If the reflected current is I ₂ , the following equation is established.

The relationship between the currents I _L and V _L flowing through the load impedance Z _L is

From this, when obtaining the voltage reflection coefficient ρ _V which is the ratio of the reflected voltage V ₂ to the incident voltage V ₁ ,

It can be expressed as In general, this voltage reflection coefficient ρ _V can be considered as a complex number having a reflection voltage magnitude Γ and a phase difference φ (the phase of reflection with respect to an incident wave). Similarly for current

Further consisting relation of V _L = I _L · Z _L and _{V L = Z L (I 1} + I 2) at the load. From Equation (2-6) and Equation (2-2),
I ₁ · Z ₀ -I ₂ · Z ₀ = Z _L · I ₁ + Z _L · I ₂ From this, the current reflection coefficient ρ _i = (I ₂ / I ₁ ) is calculated as the relationship between the incident current I ₁ and the reflected current I _2.

From the equation (2-7), the voltage reflection coefficient and the current reflection coefficient have the same magnitude and the opposite signs.
(S parameter)
The transmission characteristics of a network can be discussed quantitatively by measuring various circuit parameters. In general electric circuits and electronic circuits, H parameters, Y parameters, Z parameters, etc. are used to represent circuit characteristics. Considering the linear 4-terminal circuit shown in FIG. 4, the H, Y, and Z parameters are defined by the following equations.

To obtain these parameters, the voltage and current values must be measured as a function of frequency at the input or output of the circuit. However, in the high frequency region, when a probe is connected for voltage measurement, the circuit itself changes because the probe itself has an impedance that cannot be ignored. For this reason, it is very difficult to directly measure voltage and current as in a low frequency. On the other hand, power can be accurately measured even in a high frequency region. Therefore, parameters associated with power have been developed to characterize high frequency circuits. This is S
It is a parameter. Compared to the H, Y, and Z parameters described above, the S parameter has many advantages as shown below.
1) Directly associated with gain, loss, reflection coefficient, etc.
2) It can be measured relatively easily and there is no need to connect an undesirable load to the test circuit.
3) Even if a plurality of circuits are provided, the characteristics of the entire system can be predicted if the respective S parameters are known.
4) Any of the H, Y, and Z parameters can be derived from the S parameter.

The principle of S parameter measurement is shown in FIG. Signal a represents a wave incident on the input of the circuit,
The signal b represents each wave going out from the output end. At terminal 1, signal b ₁ is at terminal 1
It is represented by the sum of the reflected signal in the incident wave and the transmission signal incident from the terminal 2 side and passing therethrough. Similarly, the signal b ₂ at the terminal ₂ is represented by the sum of the transmission signal incident from the terminal 1 side and transmitted and the reflected signal in the incident wave on the terminal 2 side. The aι and bι defined in this way are related by the linear equation shown in FIG. A parameter Sij is defined as a correlation coefficient between the signals.
FIG. 6 shows an outline of a method for measuring S parameters S ₁₁ and S ₂₁ . First, when determining S ₁₁ and S ₂₁ , terminal 2 is terminated with Z ₀ that matches the characteristic impedance of the transmission line.
Then, the incident signal, the reflection signal, and the transmission signal are measured, and the input reflection coefficient and the input transmission coefficient can be determined using the equations shown in FIG. In the present invention, to measure a reflection characteristic S _11. From FIG. 6, S ₁₁ represents the ratio of the wave reflected at the input end among the incident waves to the terminal 1. In other words, S ₁₁ represents the reflection coefficient Γ (the amplitude ratio between the incident wave and the reflected wave) in Expression (2-5).

[Changes in high-frequency transmission characteristics due to TGO growth]
As shown in FIG. 7 (a), when a transmission line is arranged in parallel on a test piece and a high frequency signal is applied,
When the TGO in FIG. 7A is not generated, the high-frequency signal from the transmission line is transmitted or absorbed in each layer. A part of the signal is reflected at the interface between the top coat and the bond coat, and a certain reflection coefficient is obtained depending on the thickness and physical properties of each layer. When TGO is generated and grown, the impedance mismatch interface increases as shown in FIG. Depending on the thickness, physical properties, etc. of the TGO layer, the third layer with different impedance penetrates between the top coat and bond coat, so the reflection coefficient also changes. By measuring this change, the growth state of TGO can be evaluated in a non-contact and non-destructive manner. In this method, the transmission line arranged on the specimen serves as a probe.

[Probe]
[Resonance phenomenon]
The technique of the present invention realizes non-contact flaw detection. Therefore, only the probe is connected to the power source, and the test piece is not directly energized. Therefore, it cannot be said that it is a complete transmission line, and the current flowing through the probe is considered to be smaller than that of the complete transmission line. In addition, since various resistance components exist in the high frequency region, the current flowing through the probe is further reduced. However, in order to measure the reflection coefficient with high sensitivity, a certain current or more is indispensable. Therefore, in order to flow a large current through the probe, a high-frequency current is applied to the specimen using a resonance phenomenon. In the technique of the present invention, the probe end disposed on the specimen is electrically open. Considering this state corresponding to the transmission line model of FIG. 8, the load impedance is infinite. Therefore, if Z _L = ∞ is substituted into equation (2-5), Γ _L = 1. This means that all the traveling wave voltages that reach the end are reflected in the same phase and return to the direction of the power supply as reflected wave voltages. On the other hand, the current at the open end must be 0 because Z _L = ∞. Therefore, the traveling wave current that reaches the end is reflected with the phase inverted by 180 °. The high frequency reflected in this way interferes with the incident wave. At this time, when the length of the probe coincides with the half wavelength of the high-frequency current, a standing wave is generated by the incident wave and the reflected wave, and the current flowing through the probe increases. This is called a resonance phenomenon. The relationship between the probe length and frequency when the resonance phenomenon occurs is expressed by the following equation.

λλ is the high frequency wavelength, c is the speed of light, f is the high frequency, and l is the length of the probe. β is called a wavelength shortening rate and depends on the relative dielectric constant ε _S and is expressed by the following equation.

  FIG. 9 shows the time change of the voltage / current distribution on the probe when the resonance phenomenon occurs. When the resonance phenomenon occurs, a standing wave is generated such that the voltage creates a node at the center of the probe and the current forms a node at both ends of the probe.

[Probe shape and size]
One shape of a probe that can be used in the present invention is shown in FIG. For example, the probe may be a transmission line using a copper foil having a length of 30 mm, a width of 1 mm, and a thickness of 35 μm, but is not limited thereto. And it is usually placed parallel to the specimen surface. This can be connected to the inner conductor of the coaxial cable with, for example, a copper foil having a length of 8 mm. In addition, the inner conductor of the coaxial cable and the transmission line are connected so that the high frequency reaching the coaxial cable is transmitted to the probe well.
In one embodiment, the length l of the probe is 30 mm from the formula (2-11), assuming that the speed of light c is 3.0 × 10 ⁸ m / s, the high frequency f is 5.0 GHz, and the wavelength shortening rate β is 1. Can be determined. Actually, the resonance frequency may be slightly different from the theoretical value due to the influence of the copper foil and the test piece connecting the probe and the coaxial cable. It is also preferable.

[Measurement system]
[Network Analyzer]
In the measurement system according to the technique of the present invention, a network analyzer can be used for applying and measuring a high-frequency signal to a test object (for example, including a test piece). In general, a network analyzer is a device that measures transmission characteristics and reflection characteristics of a linear circuit using a sine wave. A configuration diagram of one of the network analyzers is shown in FIG.
An electrical signal can basically be determined by measuring its amplitude ratio (level), phase, and frequency, but a network analyzer can simultaneously measure the amplitude and phase. Actually, the amplitude ratio and phase difference between the reflected signal and the transmission signal and the input signal are evaluated using the S parameter. In the present invention, for example, to measure a reflection characteristic S _11. As shown in FIG. 12, the basic configuration diagram of the measurement system of the present invention is a probe (and a test body) that is the DUT (device under test) in FIG. As a typical network analyzer, for example, HP8753D manufactured by Hewlett-Packard Co., etc. can be mentioned. For example, the measurable frequency range is 30 kHz to 6 GHz. The measured value is expressed in dB and is obtained using equation (2-5).

[Placement of the probe]
As shown in FIG. 13, the probe used in the technique of the present invention can be arranged at a certain distance from the surface of a test piece such as a moving blade and vane for a power generation gas turbine. This distance is called lift-off. By maintaining this lift-off, non-contact measurement is possible. The characteristic impedance of electromagnetic waves in the vicinity of a radio wave source behaves in a complicated manner with distance as a variable. Therefore, in this example, it was noted that the measurement was performed while keeping the lift-off constant.

[Automatic measurement system]
FIG. 14 shows a schematic diagram of one embodiment of an automatic measurement system that can be used in the present invention. The automatic measurement system consists of an XYZ fine movement table that moves the probe at a designated pitch within a designated measurement range, a fine movement table control device system, a personal computer for control and data storage, a network analyzer that is a measurement device, And a probe. The probe moving device may be able to move the probe to any three-dimensional position by controlling the X-axis, Y-axis, and Z-axis direction moving motors with a control personal computer. Is preferred. Using a network analyzer as a measuring device, it is possible to perform measurement of the reflection characteristic S _11. For example, the terminal 1 and the probe of the network analyzer are 50
Can be connected with Ω coaxial cable. For example, it is preferable to arrange the probe so that the line direction and the defect are parallel to each other, and the moving direction is the Y-axis direction. The distance (lift-off) between the probe and the specimen can be measured by selecting the optimum one by confirming the resonance state between about 0.1 mm and 7 mm, for example. The measurement by the automatic measurement system is performed multiple times at each point, for example, about 5 times. At that time, the standard deviation of each measurement data set is calculated, and if it is more than the specified value, the measurement is performed again at the same place. You may do it. If it is less than the specified value, save the average value of the data set and move to the next measurement position. By repeating this operation for a predetermined region, the position and depth of the defect can be evaluated.

According to the technique of the present invention, it is possible to perform nondestructive inspection. Unlike the destructive test, the inspected material and structure can be used as they are after the test. In addition, since 100% inspection is possible, the reliability of inspection is dramatically improved compared to sampling inspection. An automatic flaw detection is preferable for 100% inspection. In addition, human error can be prevented by automatic flaw detection.
The present invention will be described in detail with reference to the following examples, which are provided merely for the purpose of illustrating the present invention and for reference to specific embodiments thereof. These exemplifications are for explaining specific specific embodiments of the present invention, but are not intended to limit or limit the scope of the invention disclosed in the present application. In the present invention, it should be understood that various embodiments based on the idea of the present specification are possible.
All examples were performed or can be performed using standard techniques, except as otherwise described in detail, and are well known and routine to those skilled in the art. .

[Measurement of cracks contained in alumina plates]
1. Shape and dimensions of test specimens In order to evaluate the quantitativeness of defect depth measurement by the high-frequency transmission characteristics method, a test that forms artificial rectangular defects on the surface of alumina, which is a typical ceramic, and includes a closed crack A piece was prepared to examine the effectiveness of the method of the present invention.
As a test piece including an artificial rectangular defect, three disc-shaped alumina plates having a diameter of 157 mm and a thickness of 8.5 mm were prepared, and a defect width of 0.5 mm and a depth of 1 mm, 3 mm, and 5 mm as shown in FIG. Each kind of artificial rectangle defect was introduced. The specimen including the natural defect is a plate-like alumina plate having a length of 315 mm, a width of 42 mm, and a thickness of 8 mm, and has a size of about 20 mm in a position about half as shown in FIG. include. Natural defects are those that occur accidentally during production.

2. Result of measurement of artificial rectangular defect Place the probe directly above the defect, set the position where the defect exists to Y = 0, both sides Y =
The region from -20 mm to Y = 20 mm was measured. For the measurement, an automatic measurement system (FIG. 14) including the probe shown in FIG. 10 and a network analyzer (HP8753D manufactured by Hewlett-Packard Co.) having the configuration shown in FIG. 11 was used. Input signal from 1.00 GHz to 5.50 for measurement
The optimum frequency at which the resonance state was maintained was measured in a preliminary experiment, and the change in the reflection coefficient was measured using that frequency. From the preliminary test results, measurements were taken at 4.56 GHz for defective materials and 4.96 GHz for non-defective materials. The defect depth was measured in the order of 5 mm, 3 mm, and 1 mm, and compared with defect-free material (defect depth 0 mm). FIG. 17 shows the result. It can be seen that in any defect, the reflection coefficient is greatly reduced in the vicinity of Y = 0 where the defect exists.
In order to quantitatively grasp the change of defect depth and amplitude ratio, an evaluation parameter of ΔdB was introduced. As shown in FIG. 17, ΔdB is defined as the distance between the intersection and the minimum reflection coefficient by drawing a perpendicular line from the minimum reflection coefficient toward the straight line connecting the left and right minimum points of the minimum reflection coefficient. did. The relationship between ΔdB and the defect depth is shown in FIG. ΔdB decreased almost in proportion to the defect depth, and a good correlation was found between the defect depth and ΔdB. The sensitivity is about 2 dB at a depth of 1 mm.
From this result, possibility that the defect contained in the ceramic surface could be quantitatively evaluated was recognized. The presence of a defect is thought to change the electric field, thereby reducing the current on the probe, thereby reducing the amplitude ratio. When a defect exists, the maximum current density directly above the defect changes, and it is considered that the defect depth can be quantitatively evaluated from the amplitude ratio obtained at this time.

3. Measurement results of natural defects As in the case of artificial rectangular defects, a probe was placed directly above the defect, and the position where the defect exists was set as Y = 0, and 10 mm on each side was measured. When measuring, set the input signal frequency to 1.00 to 5.50.
The value with minimum reflection coefficient was read by changing the frequency in the GHz range. The measurement results are shown in FIG. Similar to the case of the artificial rectangular defect, the reflection coefficient decreased near Y = 0 where the defect exists.
From the result of FIG. 19, ΔdB is about 9 dB. In FIG. 18, the sensitivity to the defect depth of 1 mm is 2 dB.
Therefore, it can be estimated that the natural defect depth is about 4.5 mm.
From the above, when nondestructive inspection method using high frequency transmission characteristics is applied to quantitative evaluation of artificial defects and natural defects in ceramics, the possibility of quantitative evaluation of defects contained in ceramics with a sensitivity of 2 dB per 1 mm depth is shown. This suggests that it may be effective for the TBC system.

[Examination of major factors affecting TBC system affecting voltage reflection coefficient]
Various factors affecting the measured values when the high frequency transmission characteristics method was applied to the thickness measurement of TGO (Thermally Grown Oxide) layers were investigated. In this example, factors considered to affect the measured values were taken up, and sensitivity analysis of each factor was performed. Based on the results, the effectiveness of the TGO thickness measurement technique of the method of the present invention was analyzed.
[TBC measurement system]
FIG. 20 shows an overall conceptual diagram of a TBC system deterioration evaluation method using the high-frequency transmission characteristic method of the present invention. The input signal is a high frequency signal and the output signal is a reflected signal. In actual measurement, the reflection coefficient is measured. The main factors affecting the reflection coefficient include the dielectric constant, film thickness, resistance of the YSZ (Yttria Stabilized Zirconia) layer and TGO layer, resistance of the MCrAlY layer, frequency, and the like. The TGO layer is roughly composed of a mixed oxide layer mixed with oxides of Cr, Ni, and Co and an alumina layer. Possible noise sources include variations in the thickness of the topcoat layer and bond coat layer during TBC construction, lift-off amount during measurement, surface roughness, the influence of surrounding high frequencies, and the fineness of the topcoat due to aging. Can be considered.

[Equivalent circuit]
For the calculation, an equivalent circuit shown in FIG. Considering that capacitance is generated between the transmission line and each layer, capacitance was arranged in the film thickness direction. In addition, each layer is considered to have a resistance component, and the resistance is arranged in a direction parallel to the change in the electric field generated from the transmission line. FIG. 21 (b) is a cross-sectional observation image of an aged TBC system photographed with a scanning electron microscope (SEM). The upper side of the image is the surface side of the TBC system, and the YSZ layer, mixed oxide layer, alumina layer, and MCrAlY layer are arranged in order from the top. The equivalent circuit is also divided into each layer and has a four-layer structure. The YSZ layer, mixed oxide layer, and alumina layer, which are non-conductive materials, are arranged with capacitance and resistance as shown in FIG. 21 (a), and the bond coat, which is a conductive material, has only resistance. The skin depth (signal penetration depth) δ of the Ni-base superalloy, which is the base material under the bond coat, was calculated by the following equation.

When Ni has a specific conductivity σ _r = about 0.26 and a frequency of 5 GHz, the skin depth δ is about 1.8 μm. Assuming that the skin depth of the bond coat (CoNiCrAlY) is about the same, the thickness of the bond coat is 100 μm, so it can be considered that high-frequency signals cannot penetrate into the Ni-base superalloy below it. Therefore, Ni-base superalloy is not considered in the equivalent circuit. As described above, the TGO layer is composed of a mixed oxide (or composite oxide) layer and an alumina layer. The alumina layer grows first, and then the mixed oxide grows on the alumina layer. This is because Al, Co, Ni, and Cr diffuse from the bond coat to the top coat interface in a high temperature environment during operation of the gas turbine. Among them, Al ₂ O ₃ (alumina) formation free energy is lower than Ni, Cr and Co oxide formation free energy, so Al is selectively oxidized, and first Al ₂ O _{3 is formed} on the bond coat side. Grow. However, to form the Al ₂ O ₃ honey remainder of Ni since not a film, Cr, Co mixed oxides combine with O topcoat interface were diffused in the Al ₂ O _3. Also in the calculation, cases were classified according to the stage of deterioration, and three equivalent circuits used for sensitivity analysis were prepared as shown in FIG.

〔Method of calculation〕
As shown in FIG. 22, the TBC system has a top coat layer (YSZ subscript letter y), mixed oxide layer (Mixed Oxide subscript letter o), alumina layer (Al ₂ O ₃ The subscript is a), and the bond coat layer (CoNiCrAlY subscript is m) has a laminated structure, which is classified into three types from the initial stage to the end of deterioration. As shown in Table 2, the level of the impedance parameter of each layer is changed for each level under three conditions. Standard values are given in the reference (Kengo Goto: Elucidation of Aging Mechanism of Thermal Shielding Coating Based on Thermally Grown Oxide Dynamics on the Sea Shield / Bond Coating Sea Surface, pp.53-61, 2001 Master Thesis (2001 ); Kazuhiro Ogawa: Non-destructive inspection, Vol. 51, No. 5, pp.275-280 (2002)) was used. YSZ resistivity ρ _y
The resistivity ρ _o of the mixed oxide and the resistivity ρ _{a of} alumina are set so as to decrease as the level increases. This is because the resistivity decreases with increasing temperature. Capacitance C _y topcoat is obtained from the equation (3-1).

Here, ε _y is the dielectric constant of the top coat layer, and t _y is the film thickness of the top coat layer. The influence of high frequency from the transmission line was assumed to be in the range of width W = 5mm and length L = 30mm. Therefore, C _y is obtained from ε _y and t _y . Similarly, determined using the capacity of the alumina layer C _a, the capacitance C _o of the mixed oxide layer is also of each dielectric constant epsilon _a, epsilon _o and thickness t _a, the t _o. The resistance _{Ry of the} top coat is obtained from equation (3-2).

Here, ρ _y is the resistivity of the top coat, and t _y , W, and L are the same as in equation (3-1). Similarly, the resistance R _{a of the} alumina layer and the resistance R _o of the mixed oxide layer are determined using the respective dielectric constants ε _a and ε _o and the film thicknesses t _a and t _o . Next, consider the resistance Rm of the bond coat. The chemical composition of CoNiCrAlY is as shown in Table 3. The conductivity of Co with the largest amount of addition is 17.2 × 10 ⁶ [/ mΩ]. The next highest Ni conductivity is 15 × 10 ⁶ [/ mΩ]. Therefore, the conductivity of CoNiCrAlY is 16 × 10 ⁶ [/ mΩ]
Assume that Therefore, the resistivity is

Assuming that the skin depth of the bond coat at 5 GHz is about 2 μm,

In this way, the capacitance and resistance of each layer were calculated from each parameter, and the reflection coefficient was determined (FIG. 23). As shown in Table 2, each parameter has three kinds of level values, and these parameters are combined and reflected with reference to the orthogonal tables of L18 (first stage) and L36 (second and third stages). Coefficients were calculated and sensitivity analysis was performed. Table 4 shows examples of value combinations in the second stage.

〔Calculation result〕
1. First Stage Calculation Results The first stage reflection coefficient Γ _ym composed of the top coat (YSZ) layer and the bond coat (MCrAlY) layer is as shown in FIG. The vertical axis represents the reflection coefficient and the horizontal axis represents the level. The closer the reflection coefficient is to 0, that is, the higher the value in the figure, the more the reflection coefficient tends to be reflected, and the lower the value, the more likely it is to transmit. From FIG. 24, it can be seen that the film thickness t _{y of the} top coat has the greatest influence on the reflection coefficient, and the reflection coefficient increases as the film thickness t _y increases, and approaches the tendency to reflect. This suggests that variations in the thickness of the topcoat during TBC construction may affect the measured value, but the sensitivity per 1 μm is as small as 0.00585 dB, so the impact on actual experiments is expected to be small. . In FIG. 24, the second largest influence on the reflection coefficient is the top coat resistivity ρ _y . As Table 2 and Figure 24 than the resistivity [rho _y decreases, the reflection coefficient is reduced. A possible cause of the change in resistivity is the densification of TBC. Since densification of TBC occurs when exposed to high temperatures, there is a possibility that the measured value changes depending on the time of thermal aging treatment.

2. Second Stage Calculation Results FIG. 25 and FIG. 26 show the second stage calculation results including the top coat layer, the alumina layer, and the bond coat layer. FIG. 25 shows the reflection coefficient Γ _y− at the interface between the top coat layer and the alumina layer.
_{For am} , FIG. 26 shows the reflection coefficient Γ _ya-m at the interface between the alumina layer and the bond coat layer. Most influential ones in the reflection coefficient in Fig. 25 is a thickness t _y of the topcoat, and a tendency that the reflection coefficient increases as film thickness increases. This is the same tendency as in the first stage. However, the sensitivity is as low as about 0.00116 dB per 1 μm. Further, the most influence on the reflection coefficient in Fig. 26 is a thickness t _a of alumina is a tendency that the reflection coefficient increases as the film thickness increases. Its sensitivity is about 0.03529 dB per μm, about 30 times the film thickness of the topcoat. Therefore, it was suggested that there is a possibility that the growth amount of the thickness of the alumina layer can be evaluated by measuring the change of the reflection coefficient by measuring the change of the reflection coefficient.

3. Calculation Result of Third Stage FIG. 27, FIG. 28, and FIG. 29 show the calculation results of the third stage including the top coat layer, the mixed oxide layer, the alumina layer, and the bond coat layer. 27 shows the reflection coefficient Γ _y-aom at the interface between the top coat layer and the mixed oxide layer, FIG. 28 shows the reflection coefficient Γ _yo-am at the interface between the mixed oxide layer and the alumina layer, and FIG. 29 shows the alumina layer. And the reflection coefficient Γ _yoa-m at the interface of the bond coat layer. In FIG. 27, FIG. 28 and FIG. 29, the largest change is the reflection coefficient at the interface between the topcoat layer and the mixed oxide layer shown in FIG. In FIG. 27, the most influential change in the reflection coefficient is the film thickness t _o of the mixed oxide layer, and shows a tendency that the reflection coefficient increases as the film thickness increases. Its sensitivity is 0.67 dB per μm. On the other hand, in FIG. 27, the reflection coefficient shows a tendency to decrease as the thickness of the alumina layer increases. This is opposite to the influence on the reflection coefficient caused by the increase in the thickness of the alumina layer in the second stage. In the case where a mixed oxide layer is present, it is presumed that an increase in the thickness of the underlying alumina layer decreases the reflection coefficient. 27, 28, and 29, the scale of the vertical axis is the same, but almost no change is seen in FIGS. 28 and 29, and it is considered that other parameters do not substantially affect the reflection coefficient.
As described above, an equivalent circuit was created, the sensitivity analysis of the effect of each parameter on the reflection coefficient was performed, and the measurement system was analyzed for evaluation of aging degradation of the TBC using the high-frequency transmission characteristics method. As a result, it was confirmed that when the TBC system is composed of a top coat layer, an alumina layer, and a bond coat layer, an increase in the thickness of the alumina layer affects an increase in the reflection coefficient. However, in the case of a four-layer structure of a top coat layer, a mixed oxide layer, an alumina layer, and a bond coat layer, an increase in the thickness of the mixed oxide layer increases a reflection coefficient, whereas an alumina layer film It has been shown that increasing the thickness decreases the reflection coefficient.
From the above analysis results, it has been clarified that the growth of the alumina layer and mixed oxide layer, which are thermally grown oxides, causes a change in the reflection coefficient in the aging degradation evaluation of the TBC system using the high frequency transmission characteristics method. . It was shown that the growth thickness of thermally grown oxide (TGO) may be quantitatively evaluated by measuring the change in reflection coefficient.

[Aging degradation evaluation of TBC system using high frequency transmission characteristics method]
The noncontact nondestructive measurement of the TBC system was performed using the high-frequency transmission characteristic method, and the possibility of quantitative evaluation of the TGO growth situation was examined together with the cross-sectional observation with a scanning electron microscope.
1. Test material (base material)
A Ni-base superalloy Inconel 601 having a length of 49 mm × width of 49 mm × thickness of 4 mm was used as the substrate. FIG.
Fig. 4 shows (a) a schematic diagram of the test piece, (b) a photograph before TBC construction, and (c) a photograph after construction. The chemical composition of Inconel 601 is shown in Table 5.

(Sprayed coating)
CoNiCrAlY (manufactured by SULZER METCO: AMDRY9951) was used as the material for the bond coat, and 8 wt% yttria stabilized zirconia (manufactured by SULZER METCO: 204NS) was used as the material for the top coat. Table 6 shows the chemical composition of CoNiCrAlY. After blasting the substrate, it was applied to about 100 μm by low pressure plasma spraying, and YSZ was applied to about 300 μm by atmospheric pressure plasma spraying.

(Thermal aging treatment)
The growth behavior mechanism of TGO is that oxygen diffused in the top coat reacts with aluminum in the bond coat to form alumina, and then mixed oxides such as Co and Cr diffused in the alumina are formed. Depending on the time and temperature exposed to high temperature environment, the thickness of TGO varies. Therefore, in this example, the use of the TBC system in a high temperature environment was simulated, thermal aging treatment was performed for several types of thermal aging times, and test pieces having different TGO growth amounts were produced.
The inlet gas temperature of the advanced gas turbine is about 1500 ° C. On the other hand, in the current moving blade and stator blade material, the maximum operating temperature is 1000 ° C. or lower (850 ° C. to 950 ° C.), and it is used in an environment cooled to about 900 ° C. by TBC or air cooling. In this example, since the thermal aging treatment was performed at 1100 ° C., the oxidation reaction was accelerated. Therefore, the aging time at 1100 ° C (1h, 7h, 25h, 100h, 225h) was converted to the aging time when the actual blade surface temperature was 950 ° C using the Larson-Miller parameter (LMP). LMP is as follows.

Here, T: temperature (K), C = 20, _tr : time (h). Table 7 shows the conversion results. In Table 7, after 1100 ° C and 100 hours, a considerably long number of years is calculated as compared with the actual repair interval of 2 years. Therefore, it can be judged that aging at 1100 ° C. for 225 hours sufficiently covers the actual use range.

2. Measurement Procedure As shown in FIG. 31, the test pieces were arranged in a line at intervals of about 15 mm, and measurement was performed by scanning the probe in the X-axis direction. When measuring the reflection coefficient, a probe was placed in the center of the non-aged material, the frequency was changed between 1 GHz and 5 GHz, and the measurement was performed under the condition that the value of the reflection coefficient was minimized. The lift-off was fixed at 4 mm, which allows for the most sensitive flaw detection based on the preliminary experiment results. In actual measurement, frequency characteristics and the effect of lift-off were taken into consideration.

〔Measurement result〕
1. Frequency Characteristics FIG. 32 shows the measurement results of the frequency characteristics of the reflection coefficient when the transmission line is located at the center of the test piece. Comparing the results of 0 h and 225 h, it can be seen that the frequency at which the reflection coefficient is minimum is 2.12 GHz for the unaged material and 2.09 GHz for the 225 h aged material, which varies depending on the aging time. In addition, the amplitude ratio at a frequency of 2.12 GHz at which the reflection coefficient of the unaged material is minimum is −44.3 dB, and the 225 h aged material is −21.0 dB, which is a difference of about 23.3 dB. When measuring the change in reflection coefficient with a fixed frequency at 2.12 GHz, this difference becomes the change in reflection coefficient.
FIG. 33 shows the result of measuring the frequency characteristics by changing the lift-off from 4 mm to 2 mm at the center of the unaged material. The frequency with the smallest reflection coefficient was constant at 2.12 GHz for both 2 mm and 4 mm. The minimum reflection coefficient was -44.3 dB at 4 mm, and -26.2 dB and 18.1 dB at 2 mm. From this result, in the measurement method of the present invention, the influence of lift-off on the reflection coefficient was determined to be a very large casting. Therefore, it was confirmed that strict control of lift-off was necessary.

2. Aging Time Dependence of Reflection Coefficient The frequency used for the measurement of the reflection coefficient change was set to a value at which the reflection coefficient becomes the minimum at the center of the non-aged material as in the measurement of FIG. 33, and was fixed at 2.15 GHz. FIG. 34 shows the measurement result of the reflection coefficient change. The aging time is shown by drawing a line in the figure for each boundary of the specimen. It can be seen that the reflection coefficient increases as the aging time increases from 0 h. Since the TGO film thickness increases as the aging time becomes longer, this change may detect an increase in the TGO film thickness. In addition, after 25 hours, the reflection coefficient decreases conversely, and from the examination result of Example 2, it is considered that growth of the mixed oxide, densification of the top coat, and the like occur.

3. FIG. 35 shows the relationship between the minimum value of the reflection coefficient and the aging time. The vertical axis represents the minimum value of the reflection coefficient, and the horizontal axis represents the aging time t. The reflection coefficient increases sharply from aging time 0 h to 25 h, but no significant change is observed thereafter at 100 h and 225 h. The reason for this is presumed that the growth of TGO is diffusion-controlled and follows a parabolic law proportional to the square root of time. The possibility that local top pores were included in the top coat and the influence of noise, etc. are also considered. Therefore, whether this result reflects the thickness of TGO is based on the structure observation in Example 4 below. Study was carried out.

[Observation of structural change by thermal aging of TBC system]
From the measurement result of Example 3, since the possibility that the change in the amplitude ratio detected the growth of TGO was obtained, a field emission scanning electron microscope (FE-SEM (HITACHI S-4700)) was used, The TGO thickness was measured by observing the cross-sectional structure of the aged TBC system. For cross-sectional observation sample preparation, first embed each sample in a resin, then use a rotary polishing table to perform rough polishing in order with water resistant abrasive paper Nos. 320, 600, 1000, 1500, and then particle sizes 15, 6, 3 A mirror finish was obtained by buffing using a diamond paste of 1, 0.5 μm. Thereafter, observation was performed by depositing a Pt film on the surface of the sample using an ion sputtering apparatus (JEOL JFC-1100E).
36 (a) to (f) show cross-sectional SEM images (1000 times) of the aging materials at 0h, 1h, 7h, 25h, 100h, and 225h, respectively. The black belt-like layer having a wave seen in the SEM image other than FIG. 36A is an alumina layer. The gray portion on the alumina layer that is noticeable in (c), (e), and (f) of FIG. 36 is the mixed oxide layer. The mixed oxide layer is scattered and grown on the alumina layer. The reason why the alumina layer is formed in a wavy shape is considered to be due to the surface roughness by shot blasting performed before thermal spraying for the purpose of improving the bonding force between the substrate surface and the bond coat. Since the MCrAlY and the topcoat are sprayed after shot blasting, the shape of the bond coat surface becomes wavy, and the shape of the TGO layer formed on the bond coat also becomes wavy. It can be seen that the thickness of the black and strip-like alumina layer gradually increases from 0h to 225h.

FIGS. 37 (a) to 37 (f) show images obtained by photographing the reflected electron images (2500 times) of the cross section in order to observe the film thickness changes of the alumina layer and the mixed oxide layer in more detail. The black belt-shaped portion is an alumina layer as in FIG. 37 (c) and (f) show that the gray mixed oxide layer is formed in a knot-like shape, and the gray mixed oxide layer is included in the mixed oxide layer. May not be homogeneous and may be distributed unevenly for several types of oxides such as Co, Cr, and Ni.
The average thickness of TGO (alumina layer + mixed oxide layer) was measured from the SEM image of FIG. 36 and shown in FIG. It can be seen that the thickness of TGO changes to follow the parabolic law. This change is considered to correspond to the parabolic law in the proposed TGO layer growth equation. According to this parabolic law, the thickness of the TGO layer formed at the topcoat / bondcoat interface is

It can be expressed as δ represents the thickness of the TGO layer, k _p represents the rate constant, and t represents the heating time. According to Arai et al.6 ⁾ , the reaction index n value takes a constant value of 0.3 in the temperature region of 973K <T ≦ 1173K. According to studies dealing with several TGO layer growth processes, the rate constant is known to show Arrhenius-type temperature dependence, and the logarithmic value of the rate constant k _p and the inverse of the heating temperature (1 / T) and The relationship is

It is represented by Here, R was a gas constant, k ₀ was 98716.0 μm / h ⁿ , and E was 120 kJ / mol. These formulas are proven at aging temperatures from 700 ° C to 1000 ° C. In the experiment of this example, since the aging temperature is 1100 ° C., there is still room for registration whether this equation can be extrapolated, but the TGO thickness calculated for the case where the reaction index n value is 0.3 and 0.4 is calculated by reference. This is also shown in FIG.
FIG. 39 shows the TGO thickness obtained from the measurement of the SEM image and the minimum value of the reflection coefficient measured in the experiment. The rapid increase from 0h to 25h obtained from the high frequency transmission characteristic method and the subsequent saturation phenomenon show the same tendency as the actual time change of TGO thickness. In the sensitivity analysis of Example 2, in the case of three layers of the top coat layer, the alumina layer, and the bond coat layer, it has been confirmed that the increase in the thickness of the alumina layer affects the increase in the reflection coefficient. Therefore, when considered together with the SEM cross-section observation results up to 25 h, it is estimated that the change in the minimum value of the reflection coefficient is due to the increase in the thickness of the alumina layer. However, in the result of 25 hours, there was a large difference between the TGO thickness and the minimum reflection coefficient. In the sensitivity analysis of Example 2, in the case of four layers of the top coat layer, the mixed oxide layer, the alumina layer, and the bond coat layer, it is confirmed that the increase in the thickness of the mixed oxide affects the increase in the reflection coefficient. On the other hand, it was confirmed that the increase in the alumina layer affects the decrease in the reflection coefficient. In this case, since the increase in the reflection coefficient accompanying the increase in the mixed oxide layer is more conspicuous, the reflection coefficient was expected to increase even in the case of four layers. However, FIG. 39 shows that the change in the minimum value of the actual amplitude ratio is slightly reduced from 25h. When attention is paid to (d), (e), and (f) in FIG. 36, it can be seen that the thickness of the mixed oxide layer does not increase so much. For this reason, it is considered that the increase in the thickness of the alumina layer significantly affects the reflection coefficient rather than the increase in the thickness of the mixed oxide layer. From the results so far, it was found that the possibility of quantitative evaluation can be considered sufficiently.

[Effect of densification of top coat on reflection coefficient]
In Example 4, evaluation of thermally aged TBC was performed using the high frequency transmission characteristic method, and the possibility of TGO film thickness evaluation was shown. However, the change in TGO film thickness due to thermal aging time was not consistent with the change in reflection coefficient. One reason for this is thought to be a dense top coat.
The densification of the top coat is caused by the progress of sintering due to thermal aging. This densification leads to an increase in thermal conductivity and affects the thermal insulation performance of TBC. Furthermore, a change in electromagnetic properties due to densification is also expected, and it is possible that the reflection characteristics are also changing. Therefore, in Example 5, a test piece having only a top coat was prepared, and the influence on the measurement result of the thermal aging time was examined.
1. Specimen As shown in Fig. 40, it was cut by wire electrical discharge machining until a metal part with a thickness of about 0.1 mm was left under the top coat, and then washed into aqua regia (HCl: HNO ₃ = 3: 1). The soaked metal part was dissolved to obtain a ceramic top coat alone. FIG. 41 shows the photograph. Furthermore, thermal aging was applied to the test piece of the top coat alone at 1100 ° C. for 1 h, 7 h, and 100 h. The measured porosity is shown in Table 8 for reference.

2. Measurement Method Measurement was performed by arranging the test pieces in a line at intervals of about 15 mm as shown in FIG. 42, and scanning the probe in the X-axis direction. When measuring the change in the reflection coefficient, place a probe in the center of the unaged material, change the frequency from 1 GHz to 5 GHz, determine the frequency at which the value of the reflection coefficient is minimum, Hereinafter, measurement was performed at the frequency. In addition, the lift-off was kept at 4 mm, which allows for the most sensitive flaw detection in preliminary experiments.

3. Frequency characteristics FIG. 43 shows the measurement results of the frequency characteristics of the reflection coefficient when the transmission line is located at the center of the test piece. The frequency showing the minimum reflection coefficient was 1.83 GHz. As can be seen from FIG. 43, no change in frequency characteristics due to the difference in aging time is observed. Therefore, it is considered that the influence of densification of the top coat on the reflection characteristics used in the present invention is small.
In FIG. 32, the frequency at which the amplitude ratio is minimum is 2.12 GHz, and a difference of 0.29 GHz is generated as compared with 1.83 GHz in the case of only the top coat. From this, it is considered that there is a difference in reflection characteristics depending on the presence or absence of the metal part.

4). Reflection coefficient change Measurement of the reflection coefficient change was performed using 1.83 GHz, which shows the minimum reflection coefficient in FIG. The measurement result of the reflection coefficient change of FIG. 44 is shown. The aging time is shown by drawing a line in the figure for each boundary of the specimen. In the measurement result of FIG. 44, a substantial change in the reflection coefficient due to the difference in aging time is not recognized. The variation error of the amplitude ratio is considered to be due to the variation in the thickness of the top coat, the variation in the gap distance between the test piece and the base alumina plate, or noise.

As described above, the possibility of quantitative evaluation of TGO thickness by measuring various thermal aging specimens using the high-frequency transmission characteristic method capable of non-destructive non-contact measurement of TGO growth in the TBC system As a result of verification, the following became clear.
1) It was clarified that the reflection coefficient increases according to the parabolic law, which is the growth law of TGO, depending on the thermal aging time.
2) As a result of cross-sectional observation by SEM, it was confirmed that the film thickness of TGO also increased according to the parabolic law.
3) It has been clarified that there is a high correlation between the TGO thickness and the tendency of change in the minimum value of the reflection coefficient.
4) The reflection coefficient is easily affected by lift-off, and accurate control of lift-off is required for measurement.
5) It was clarified that the reflection coefficient of the top coat alone did not change regardless of the aging time. Therefore, it is considered that the densification of the top coat does not affect the measured value.
From the above examination results, it was clarified that the TGO thickness generated and grown inside the TBC system could be measured in a non-contact and non-destructive manner by the high-frequency transmission characteristic method proposed in this study.

In the present invention, a gas used under severe conditions such as a gas turbine for a thermal power plant in which the operating temperature is raised for reducing carbon dioxide, which is a cause of global warming, and for improving efficiency. It is possible to develop inspection technology for non-destructive and non-contact evaluation of deterioration of turbine blade thermal barrier coating (TBC) systems. In the present invention, it is possible to quantitatively evaluate the generation / growth thickness of thermally grown oxide (TGO) formed in the TBC system using high-frequency signal transmission characteristics. Accurate maintenance management for expensive parts such as high-temperature parts becomes possible, and not only safety but also economical advantages are great. A non-destructive evaluation can be performed quantitatively, guaranteeing that reasonably expensive parts and materials can be used without repair or disposal, and the materials can be used to their original lifetime.
It will be apparent that the invention may be practiced otherwise than as particularly described in the foregoing description and examples. Many modifications and variations of the present invention are possible in light of the above teachings, and thus are within the scope of the claims appended hereto.

The structure of the TBC system. TBC degradation and damage. Voltage reflection and current reflection. Linear two-terminal circuit. S-parameter measurement principle. Summary of S ₁₁ and S ₂₁ measuring method. Changes in high frequency current due to the presence of defects in the TBC system. Transmission line model. Time variation of voltage / current distribution on the probe. The shape of the probe. The block diagram of a network analyzer. The basic composition figure of the measurement of the method of the present invention. Probe placement. The automatic measurement system of the method of the present invention. Schematic diagram and appearance photograph of ceramic alumina test piece containing artificial rectangular defect. Schematic diagram and appearance photograph of ceramic alumina test piece containing natural defects. The measurement result of the artificial rectangular defect of the test piece shown by FIG. Relationship between ΔdB and defect depth. The measurement result of the natural defect of the test piece shown in FIG. The conceptual diagram of the TBC system degradation evaluation method by the high frequency transmission characteristic method of this invention. SEM observation image and equivalent circuit of cross section of aged TBC system. Equivalent circuit considering TGO layer growth process. How to find the reflection coefficient. The reflection coefficient Γ _ym . The reflection coefficient Γ _y-am . The reflection coefficient Γ _ya-m . The reflection coefficient Γ _y-aom . The reflection coefficient Γ _yo-am . The reflection coefficient Γ _yoa-m . Schematic diagram of the base material and an example of overview photography before and after TBC construction. Measurement of change in amplitude ratio. Frequency characteristics at the center of the test piece of 0h aging material and 225h aging material. Change in frequency characteristics due to lift-off. Change in reflection coefficient due to difference in aging time. Relationship between minimum reflection coefficient and aging time. SEM image of cross section of tissue in TBC system (1000 times). Backscattered electron image of tissue cross section in TBC system (2500 times). Relationship between TGO thickness obtained by measurement and calculation and aging time. Relationship between minimum TGO thickness and reflection coefficient and aging time. A test piece preparation method for analyzing the influence of densification of the top coat on the reflection coefficient. The base material after cutting and the produced top coat test piece. State of measurement. 0h, 1h, 7h, 100h Frequency characteristics at the center of test piece of aging material. Change in topcoat reflection coefficient due to differences in aging time.

Claims (4)

A non-destructive evaluation system characterized by non-contact and non-destructive evaluation using a high frequency of 1G to 100 GHz for quantitative evaluation of an oxide film formed on a metal substrate. Non-destructive evaluation system characterized by non-contact and non-destructive evaluation of generation / growth at the interface between a bond coat such as MCrAlY formed on a metal substrate and a ceramic coating formed thereon. Non-destructive evaluation sensor that evaluates the signal of oxide film or coating interface in response to high frequency of 1G to 100GH. It is characterized by non-contact and non-destructive measurement of the formation and growth of thermally grown oxide generated in a heat-resistant material having a thermal shielding coating applied on a metal substrate using high-frequency transmission characteristics. Heat shield coating inspection method.