CN101285796B - Heat barrier coatings damage and its failure procedure acoustic emission real-time detection method - Google Patents

Abstract

The invention discloses a real-time acoustic emission detection method for thermal barrier coating damage and a failure process thereof, which comprises the following steps that: a thermal barrier coating sample and an acoustic emission sensor are connected and coupled; a load is imposed on the thermal barrier coating sample, the acoustic emission sensor receives an acoustic wave signal emitted from the sample; the received acoustic wave signal is carried out through wavelet transform to obtain a wavelet energy spectrum coefficient distribution map, the damage pattern of the signal is determined according to the wavelet energy spectrum coefficient distribution map; the quantitative analysis of the damage is performed according to a relationship curve that the number of acoustic emission signal incidents of each damage pattern changes with the imposed load; a signal component under the scale where the maximum value of the wavelet energy spectrum coefficient is located is extracted, the time difference that the signal component reaches each sensor is calculated by a correlation analysis method, and the position of a damage source is determined by a time difference position method. The method can simply and quickly make a real-time detection to the damage of the thermal barrier coating material and also contributes to the accurate forecast to the life of the thermal barrier coating at the same time.

Description

The acoustic emission real-time detection method of heat barrier coatings damage and failure procedure thereof

Technical field

The present invention relates to the acoustic emission real-time detection method of a kind of heat barrier coatings damage and failure procedure thereof.

Background technology

Thermal barrier coating is a kind of ceramic coat; have low-down heat-conduction coefficient and good resistance to high temperature oxidation energy; it is deposited on the surface of refractory metal parts or alloy by adhesive linkage; effectively high-temperature component (alloy) is completely cut off with high-temperature fuel gas and come; to reduce the surface Working Temperature of high-temperature component; and the protection high-temperature component is avoided the corrosion and the erosion of high-temperature fuel gas; prolong the serviceable life of high-temperature component greatly; and then the raising aeromotor fuel gas temperature and the thermal efficiency; be one of gordian technique of modern aeroengine, good application and development potentiality are arranged.The Nitin of Connecticut, USA university; Gell and Jordan in 2002 at internal authority magazine＜Science go up the importance of describing thermal barrier coating like this: " the dissimilar coating of many kinds is used to protect the Structural Engineering material away from burn into friction and erosion environment, and lubricated and thermal insulation protection effect are provided.Wherein thermal barrier coating system is a kind of structure the most complicated in all coat systems, also is a kind of coating of being badly in need of most being applied in spacecraft and the turbine high-temperature component environment ".Yet thermal barrier coating a series of failure modes such as unforeseen coating cracking, interfacial separation, coating shedding can occur in actual use, why can thermal barrier coating lose efficacy? what is the failure mechanism of thermal barrier coating? how to detect and predict the time and the place of losing efficacy and taking place? how to prevent to lose efficacy? this a series of problem is made in field and the national defence defence in various countries' space flight and aviation and is become increasingly conspicuous, and also is the bottleneck problem that seriously restricts the Thermal Barrier Coating Technologies widespread use.

Because thermal barrier coating complex structure, its failure behaviour are also closely related with the mechanical property of surface effect, size effect, interfacial effect and substrate except outside the Pass the mechanical property with itself has.Inquire into and answer these problems, most important also the most direct research method is exactly the simulation that experimentizes of actual working environment to thermal barrier coating, realizes the real-time detection to the whole failure procedure of thermal barrier coating simultaneously.This be because, with respect to development new, complicated theoretical tool and breaking test method, real non-destructive to the thermal barrier coating failure procedure detects more direct, convenient, it can not only overcome the problem of in theory is analyzed " attending to one thing and lose sight of another ", promptly can not with the factor of influential thermal barrier coating failure behaviour consider fully, and avoided the key messages such as damage formation, evolution and ultimate failure process lost in the traditional breaking test.Therefore, development real non-destructive detection method realizes the Non-Destructive Testing of thermal barrier coating failure procedure, for the understanding and the life prediction of thermal barrier coating inefficacy mechanism provides direct foundation and guidance.Therefore, " the real-time detection technique of heat barrier coatings damage and failure procedure thereof " used thermal barrier coating, reduced catastrophic accident for high-level efficiency, promotes the growth of national economy to become possibility, has important academic significance and engineering using value.

Acoustic emission is meant that material is out of shape or when rupturing, because of discharging a kind of phenomenon that (elasticity) energy produces the transient state stress wave rapidly under being subjected to external influence.Gather, analyze these stress wave signals and utilize these signals to infer that the technology of distortion or fracture origin information is called acoustic emission, this technology has dynamically, harmless, real-time detection characteristics.We know, how not complicated the structure of tube material and shape be, no matter how complicated its loaded state and Service Environment be, also no matter how complicated the form of its damage and inefficacy is, in the process that produces distortion and damage, the variation that always is accompanied by energy that is to say and can discharge distortional strain energy, thereby produces stress wave acoustic emission signal just; And, because damage residing position, structure difference in material, the mode difference that forms or expand, always there is difference more or less in the stress wave signal of its generation.Thereby acoustic emission testing technology will be a kind of important laboratory facilities and the method that realizes that the thermal barrier coating failure procedure detects in real time.But, in the face of the complicated and diversified inefficacy of thermal barrier coating, how to discern failure mode, how amount damage develops, and position how to determine source of damage is the huge difficult problem that present acoustic emission testing technology faces.Specifically,, realize that the real-time detection of its failure procedure under the high temperature Service Environment of reality is particularly important to studying its inefficacy mechanism, but also do not report the progress of this respect at present for the thermal barrier coating of being on active service under hot environment.

Summary of the invention

In order to address the above problem, the invention provides the acoustic emission real non-destructive detection method of a kind of heat barrier coatings damage and failure procedure thereof.The present invention can realize that the acoustic emission signal of heat barrier coatings damage under the normal temperature detects, qualitative, the quantitative and positioning analysis of damage; And the acoustic emission signal that can realize heat barrier coatings damage under the hot environment detects and the qualitative and quantitative analysis.

The technical scheme that the present invention solves the problems of the technologies described above may further comprise the steps:

1) the thermal barrier coating sample is connected, is coupled with calibrate AE sensor;

2) to thermal barrier coating sample imposed load, calibrate AE sensor receives the acoustic signals that sends in the sample;

3) acoustic signals that receives is carried out wavelet transformation, obtain Wavelet Energy Spectrum coefficient distribution plan, it can be compared by the spectral coefficient distribution plan with the feature of determining damage mode, judge the damage mode of heat barrier coatings damage acoustic signals;

4) add up the acoustic emission signal event number that each damage produces, draw the experimental formula that event number changes with imposed load.

5) extract component of signal under the pairing wavelet decomposition yardstick of Wavelet Energy Spectrum coefficient maximal value, calculate this component of signal with correlation analysis and arrive mistiming between each sensor, adopt traditional time difference positioning method then, determine the position of source of damage.

It is as follows among the present invention acoustic emission signal to be carried out the principle of wavelet transformation:

If ψ (t) ∈ is L ²(R) its Fourier transform is

When

When satisfying enabled condition:

Claim that ψ (t) is a wavelet basis or female small echo.ψ (t) after flexible and translation, just can be obtained one group of wavelet basis function ψ _{A, b}(t):

${\mathrm{\ψ}}_{a,b}\left(t\right)=\frac{1}{\sqrt{a}}\mathrm{\ψ}\left(\frac{t-b}{a}\right),a,b\∈R,a\≠0---\left(2\right)$

With scale factor a and the discrete a=2 that turns to of shift factor b ^j, b=k2 ^j ₀K, j ∈ Z, then wavelet transformation is discrete dyadic wavelet transform, and wavelet function is 1/2 being a vibration equation of the finite energy at center, and discrete its reconstruction signal of dyadic wavelet transform can be expressed as

$s\left(t\right)=\underset{j}{\mathrm{\Σ}}\underset{k}{\mathrm{\Σ}}{c}_{j,k}{\mathrm{\Ψ}}_{j,k}\left(t\right),---\left(3\right)$

Here c _{J, k}Be wavelet coefficient, ψ _{J, k}(t) be little wave system by wavelet function ψ (t) expansion and translation.

Suppose the little { ψ of wave system _{J, k}(t), k ∈ Z} is at W _jOne group of orthogonal basis in space.φ (t) is the scaling function of wavelet function ψ (t), and the little { φ of wave system _{J, k}(t), k ∈ Z} (φ _{J, k}(t) expansion and the translation by φ (t) obtains) be V _jOne group of orthogonal basis in space.According to scaling function and wavelet function, can produce a low pass and Hi-pass filter, their effect equation h (k), g (k) are defined as:

h(k)＝<ψ _j(t)，ψ _j-1，k(t)>， (4)

g(k)＝<φ _j(t)，φ _j-1，k(t)>. (5)

SPACE V _J-1So can be decomposed into a sub spaces V _jAnd high-frequency information space W _jAnd, W _jIt then is SPACE V _jAnd SPACE V _J-1Quadrature, and have $V_j=\&Subset;V_{j-1},$ Have like this

$V_{j-1}=V_j{\&CirclePlus;W}_j.---\left(6\right)$

A certain signal is at V _J-1So the quadrature component on the space can be regarded V as _jAnd W _jAnd, V here _j, W _jCan pass through V _J-1Obtain with the convolution of low pass or Hi-pass filter.So, low frequency wavelet coefficient c _{J, k}And high frequency wavelet coefficient d _{J, k}Can represent with following formula:

$c_{j,k}=c_j\left[k\right]=\underset{m}{\mathrm{\&Sigma;}}h(m-2k)c_{j-1}\left[m\right],---\left(7\right)$

$d_{j,k}=d_j\left[k\right]=\underset{m}{\mathrm{\&Sigma;}}g(m-2k)c_{j-1}\left[m\right],---\left(8\right)$

H (m-2k) and g (m-2k) are by insert the reconstruct equation of j-1 0 respectively at coefficient h (k) and g (k).

Definition C _jS (t) and D _jS (t) is respectively the low frequency and the high frequency decomposition amount of j yardstick.Then they can be expressed as

$C_{j}s\left(t\right)=\underset{k}{\mathrm{\&Sigma;}}{c}_{j,k}h(m-2k)---\left(9\right)$

$D_{j}s\left(t\right)=\underset{k}{\mathrm{\&Sigma;}}{d}_{j,k}g(m-2k)---\left(10\right)$

Like this, the J yardstick wavelet decomposition for certain signal S (t) can be expressed as:

$s\left(t\right)=C_{J}s\left(t\right)+\underset{J=1}{\overset{J}{\mathrm{\&Sigma;}}}{D}_{J}s\left(t\right).---\left(11\right)$

Definition E _J ^CS (t) and D _J ^DS (t) is respectively any time t, the cumlative energy of low frequency and high band under the j yardstick.Promptly

$E_J^{C}s\left(t\right)=\underset{\mathrm{\&tau;}=1}{\overset{t}{\mathrm{\&Sigma;}}}{\left(C_{J}s\left(\mathrm{\&tau;}\right)\right)}^2---\left(12\right)$

$E_J^{D}s\left(t\right)=\underset{\mathrm{\&tau;}=1}{\overset{t}{\mathrm{\&Sigma;}}}{\left(D_{j}s\left(\mathrm{\&tau;}\right)\right)}^2,j=\mathrm{1,2},...J---\left(13\right)$

Gross energy can be obtained by following formula

$\mathrm{Es}\left(t\right)=E_J^{C}s\left(t\right)+\underset{j=1}{\overset{J}{\mathrm{\&Sigma;}}}{\left(E_J^{D}s\left(t\right)\right)}^2---\left(14\right)$

Defining the energy of signal under each wavelet scale and the ratio of gross energy is the Wavelet Energy Spectrum coefficient, then has:

$r{E}_J^C=\frac{E_J^{C}s\left(t\right)}{\mathrm{Es}\left(t\right)}$

$r{E}_J^D=\frac{E_J^{D}s\left(t\right)}{\mathrm{Es}\left(t\right)},j=\mathrm{1,2},...J$

(15)

With rE _J ^CAnd rE _J ^DTo wavelet decomposition yardstick j mapping, can obtain Wavelet Energy Spectrum coefficient distribution plan, the pairing wavelet scale of Wavelet Energy Spectrum coefficient maximal value is the characteristic dimension of signal among the figure, and corresponding frequency range is the characteristic spectra of acoustic emission signal.

Extract signal after the component of signal of characteristic spectra by wavelet transformation, to arrive principle and the method for the time difference between each sensor as follows to draw signal among the present invention acoustic emission signal to be carried out correlation analysis: suppose that a certain signal is received by two sensors, the time-limited component of signal that collects is respectively x (n) and y (n), then the cross-correlation coefficient ρ of these two signals _XyBe defined as:

${\mathrm{\&rho;}}_{\mathrm{xy}}=\frac{\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}x\left(n\right)y\left(n\right)}{{\left[\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}{x}^2\left(n\right)\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}{y}^2\left(n\right)\right]}^{\frac{1}{2}}}---\left(16\right)$

Following formula has been represented the similarity degree of signal x (n) and y (n), when x (n)=y (n), and ρ _Xy=1, two signals are relevant fully; When signal x (n) has nothing to do fully with y (n), ρ _Xy=0; When x (n) and y (n) have when similar to a certain degree, | ρ _Xy| at 0 and 1 intermediate value.

The first half that the sensor that receives signal earlier is signal collected is defined as x (n), and the first half of another this signal of sensor acquisition is defined as y (n), asks y (n) and the related coefficient of x (n) to obtain ρ _Xy(0), then pointwise translation y (n+m) (m=1,2...), with the relevant related coefficient ρ that obtains of length that is equal to x (n) _Xy(m), have

${\mathrm{\&rho;}}_{\mathrm{xy}}\left(m\right)=\frac{\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}x\left(n\right)y(n+m)}{{\left[\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}{x}^2\left(n\right)\underset{n=0}{\overset{N}{\mathrm{\&Sigma;}}}{y}^2(n+m)\right]}^{\frac{1}{2}}}---\left(17\right)$

Work as ρ _XyWhen (m) getting maximal value, signal x (n) is greatly similar to y (n+m), and this moment, the product of m and sampling interval was the mistiming Δ t that acoustic emission signal arrives two sensors.

Technique effect of the present invention is: 1) the present invention utilizes the platinum filament waveguide rod that the thermal barrier coating sample is connected with calibrate AE sensor, the platinum filament waveguide rod can not only well be transmitted acoustic emission signal, guarantee that calibrate AE sensor is within the rated temperature, simultaneously also can easily follow thermal barrier coating moves, do not increase noise signal, the use of waveguide rod has solved thermal barrier coating can't carry out the technical barrier that acoustic emission signal detects under the hot environment.2) acoustic emission signal that detects is carried out wavelet transformation, realized qualitative, the quantitative and location evaluation of heat barrier coatings damage, enrich data for the diagnosis of thermal barrier coating failure mechanism, failure procedure provides, also provide direct foundation simultaneously for the life prediction of thermal barrier coating.Detection method of the present invention and analysis means are simple to operate not only can to carry out Nondestructive Evaluation by the failure procedure to thermal barrier coating in the laboratory, also can the damage evolutionary process of Nondestructive Evaluation thermal barrier coating under the test run environment of reality.

The present invention is further illustrated below in conjunction with accompanying drawing.

Description of drawings

Fig. 1 realizes the structural drawing that heat barrier coatings damage and failure procedure acoustic emission thereof detect in real time for the present invention.

Fig. 2 is the process flow diagram of thermal barrier coating acoustic emission signal wavelet transformation among the present invention.

Fig. 3 is the process flow diagram of the sound emission signal characteristic of a certain damage mode of thermal barrier coating among the present invention.

Fig. 4 is the process flow diagram of heat barrier coatings damage pattern-recognition among the present invention and damage quantitative analysis.

Fig. 5 realizes the process flow diagram of heat barrier coatings damage positioning analysis for the present invention.

Fig. 6 is the distribution plan of the acoustie emission event of thermal barrier coating among the present invention along with the load time.

Fig. 7 is the distribution plan of the acoustie emission event of substrate sample among the present invention along with the load time.

The three kind typical wavelet characters energy spectrograms of Fig. 8 for occurring in the thermal barrier coating stretching failure procedure among the present invention: (a) Surface Vertical crackle (b) Interface Crack, (c) substrate deformation.

Fig. 9 is the distribution plan of various failure modes with stress or time that stretch of thermal barrier coating among the present invention.

Figure 10 stretches in the failure procedure, the acoustie emission event number of the different failure modes of thermal barrier coating and the relation of load time.

The location map of Figure 11 Surface Vertical crackle and Interface Crack acoustic emission source.

Figure 12 thermal barrier coating is in the acoustic emission signal of the cooling and the heating period appearance of thermal cycle.

Various damage mode are at the distribution plan of heating period during Figure 13 thermal cycle.

Various damage mode are at the distribution plan of cooling stage during Figure 14 thermal cycle.

Embodiment

Referring to Fig. 1,1 is the thermal barrier coating sample among the figure, 2 is the platinum filament waveguide rod, and 3 is calibrate AE sensor, and 4 is Acoustic radiating instrument, 5 is the wavelet analysis processing unit, 6 are damage The qualitative analysis display unit, and 7 is acoustic emission signal parameter analytic unit, and 8 is damage quantitative analysis result display unit, 9 is cross correlation analysis and time difference positioning analysis unit, and 10 are damage positioning result display unit.

The present invention includes following steps: whether the environment temperature decision according to 1 military service of thermal barrier coating sample and experiment adopts waveguide rod 2; Adopt spot welding waveguide rod to be connected with calibrate AE sensor 3 with sample with mechanical hook-up, as not needing to use waveguide rod, then directly sample and calibrate AE sensor are coupled by ultrasonic couplant, the number of calibrate AE sensor then per sample shape, size and the requirement of tradition time difference location determine, as to wire, adopt the straight line localization method, can be with 2 sensors.Tabular general employing plane positioning method, available three or four sensors; To the thermal barrier coating imposed load, calibrate AE sensor receives the acoustic emission signal that damage is sent; Sample, handle, show and store by 4 pairs of acoustic emission signals that receive of Acoustic radiating instrument; Utilize 5 pairs of acoustic emission signals that collect of wavelet analysis technology to handle, ask for the distribution plan of Wavelet Energy Spectrum coefficient, determine the damage mode 6 of this signal; Adopt the parameters analysis method 7 of acoustic emission signal, each is damaged pairing acoustic emission signal add up, realize the quantitative test 8 of damage with the Changing Pattern of imposed load; Utilize wavelet analysis technology 5 to extract the component of signal of signal, obtain the mistiming that different sensors receives this component of signal with correlation analysis technology 9 then, realize the positioning analysis 10 of damage at last in conjunction with traditional localization method at characteristic spectra.

The process flow diagram of acoustic emission signal wavelet transformation as shown in Figure 2.At first analyze the feature of thermal barrier coating acoustic emission signal, choose suitable wavelet function; Parameter attribute when gathering according to acoustic emission signal comprises sample frequency f _s, sampling length N and filter length L _f, the out to out J of calculating wavelet decomposition _Max, computing formula is:

$J_{\max }=\min (\mathrm{int}\left({\log }_2\frac{f_s}{20}\right),\mathrm{int}({\log }_2\frac{N}{L_f}+1));---\left(18\right)$

Then signal is carried out wavelet decomposition, obtain the reconstruction signal and the corresponding frequency spectrum figure of signal time domain under each yardstick; Calculate the energy of signal under each yardstick and the ratio of gross energy, draw Wavelet Energy Spectrum coefficient distribution plan, finish the wavelet analysis of acoustic emission signal.

In thermal barrier coating system, modal damage type has the substrate plastic yield, in the ceramic layer perpendicular to the crackle of thermal barrier coating system median surface and ceramic layer and intermediate layer at the interface and along the Interface Crack of this interface expansion.Realize the qualitative analysis of heat barrier coatings damage, key is to identify the sound emission signal characteristic of these three kinds of damage mode.The frequecy characteristic and the signal of different its acoustic emission signals of damage mode are different in the energy distribution of each frequency band, and the frequecy characteristic of signal and energy distribution situation are almost irrelevant with size, the magnitude of load of damage.For this reason, can carry out single damage mode respectively or have only a kind of unfounded acoustic emission experiment of sound emission signal characteristic of damage mode, obtain a large amount of damage mode acoustic emission signal of cicada, the sound emission signal characteristic of this pattern is analyzed and set up to these signals, and its process flow diagram as shown in Figure 3.

After the sound emission signal characteristic of various damage mode in the thermal barrier coating is set up, just can finish the identification of a unknown damage mode acoustic emission signal easily, finish the quantitative test of damage simultaneously, its process flow diagram is as shown in Figure 4.To an acoustic emission signal that records the unknown pattern of load time or loading stress information, obtain Wavelet Energy Spectrum coefficient distribution plan by wavelet analysis, with it judgement of comparing with the feature energy spectral coefficient distribution plan of determining damage mode, if can spectral coefficient similar then acoustic emission signal counter this damage mode adds 1 to a certain pattern wherein, similar then be not classified as other if having.Statistical figure with the acoustic emission signal counter of various damages are ordinate, load time or stress with the signal correspondence are the horizontal ordinate mapping, obtain damaging the acoustic emission signal that produced relation curve with plus load, this curve is carried out numerical fitting, can obtain both experimental formulas, realize the quantitative test of damage.For the acoustic emission signal that belongs to other damage mode, also need the pattern that other is new that how much judges whether according to its quantity, this moment can be by analytical approachs such as spectrum analyses.

Fig. 5 is a process flow diagram of realizing the heat barrier coatings damage positioning analysis.For an acoustic emission signal, at first it is carried out wavelet decomposition, and extract signal in the characteristic spectra component of signal of the pairing wavelet scale of Wavelet Energy Spectrum coefficient maximal value just, also extract the component of signal of other sensor in corresponding band, utilize correlation analysis to solve the mistiming of different sensors received signal, in conjunction with traditional time difference positioning method source of damage is positioned again.As sample (being wire) for an one dimension size, known two sensors is placed in the two ends of sample respectively, be that distance between the two sensors is the length L of sample, behind the mistiming Δ t that obtains the two sensors signal with top method, then can calculate the distance that acoustic emission source (source of damage) distance receives the sensor of signal earlier $l=\frac{L-\mathrm{V\&Delta;<figure-callout\; id="2"\; label="t"\; filenames="S2008100311804E00011.png"\; state="\{\{state\}\}">t</figure-callout>}}{2},$ Wherein V is the velocity of propagation of sound wave in material, can table look-up, and also can measure.

Specific embodiments of the invention 1:

Stretch the at normal temperatures acoustic emission of failure procedure of thermal barrier coating detects in real time.Deng the base material from spraying thermal barrier coating sample is GH3030, and the middle transition layer material is NiCr ₂₂Al ₇Y _0.2(wt.%), ceramic layer material is ZrO ₂-8wt.%Y ₂O ₃, the thickness of intermediate layer and ceramic layer is about 200 μ m.Experiment loads sample with omnipotent drawing machine, simultaneously two calibrate AE sensors is placed in the two ends of sample, receive acoustic emission signal, and by the Acoustic radiating instrument collection, sample frequency is 1MHz, and sampling length is 1K, and threshold value is provided with 36dB.Experiment has been carried out stretching and acoustic emission detection to base material and the thermal barrier coating that does not have coating respectively.

In the fail in tension process, the acoustie emission event of thermal barrier coating and substrate sample along with the distribution of load time shown in Fig. 6,7, the acoustie emission event distribution curve can be divided into four zones: (I) less in the starting stage acoustie emission event that loads, in this stage, substrate is in elastic deformation; (II) along with the surrender of base material, acoustie emission event increases rapidly.In this stage, the crackle in substrate plastic yield and the coating exists jointly, does not see coming off of ceramic layer in this stage; (III) acoustie emission event significantly reduces again, and in the moment that this stage finishes, ceramic layer comes off from substrate substantially, thus acoustic emission signal rapid increase should be owing to the fracture of thermal barrier coating base material in (IV) stage.Substrate is in the process of fail in tension, and acoustic emission signal occurs seldom before surrender, thereby can learn that the acoustic emission signal in (I) stage derives from the crack growth or the expansion of coating.

At first the acoustic emission signal of base material when stretching carried out wavelet decomposition, and determine the feature of its Wavelet Energy Spectrum coefficient, shown in Fig. 8 a; Then a large amount of acoustic emission signals before the substrate surrender are carried out wavelet analysis, determine the feature of the Wavelet Energy Spectrum coefficient of Surface Vertical crackle, shown in Fig. 8 b; From the acoustic emission signal that gather substrate surrender back, find out a large amount of acoustic emission signals that are different from the two kinds of damage mode in front and carry out wavelet decomposition once more, determine the Wavelet Energy Spectrum coefficient characteristics of Interface Crack, shown in Fig. 8 c.

After determining these three kinds of damage mode, just can carry out wavelet analysis in the fail in tension process to the acoustic emission signal that the thermal barrier coating sample is collected, draw the Wavelet Energy Spectrum characteristic coefficient distribution plan of each signal, and compare with three kinds of Wavelet Energy Spectrum characteristic coefficient distribution plans among Fig. 8, determine the damage mode of each signal.Count the event number of the acoustic emission signal of each damage generation, and be ordinate with the acoustic emission signal event number of various damages, with the load time is that horizontal ordinate is done histogram 9, determine thermal barrier coating and under the drawing stress effect, experienced the coating failure procedure that Surface Vertical crackle, Interface Crack and ceramic layer come off, and known that the start-stop stress that the damage of Surface Vertical crackle, Interface Crack are damaged is respectively 57MPa and 436MPa, 363MPa and 513MPa.Acoustic emission signal event number with a certain damage is an ordinate, with the load time is that the horizontal ordinate mapping draws the evolution rule of this damage with load time (being directly proportional with loading stress), as Figure 10, in the case, Surface Vertical crackle and Interface Crack all are power function relationship with the load time, and experimental formula is N=a (t-t ₀) ^x(a, x are fitting constant, t ₀Relevant with the zero-time that damage forms).

Use localization method of the present invention, a certain thermal barrier coating sample is drawn the acoustic emission signal analysis of coating for each damage before coming off fully, the result as shown in figure 11.The statistical distribution of acoustic emission signal can provide the position of the various damages of thermal barrier coating accurately.Ordinate is represented the time that acoustic emission signal occurs among the figure, and horizontal ordinate is represented the position of source of damage From Left sensor.(Figure 11 a) to surperficial vertical crack, can see the distribution that acoustic emission signal is mixed and disorderly along the longitudinal direction, but anatomize, but can significantly find out the rule that acoustic emission signal distributes, the acoustic emission signal of Chu Xianing occurs in the position of about 30～37mm of range sensor and the about 10.5mm of distance probe the earliest, along with the increase of time, the acoustic emission signal in these two zones becomes intensive; When the time further increased, the acoustic emission signal at the about 19～22mm of range sensor and 56～60mm place became intensive; Time, acoustic emission signal almost spreaded all over each position of sample longitudinal direction when increasing again, but can also see in some place intensively, and some place is sparse.The position consistency at the visible surface vertical crack place at 10.5mm, 21mm, 34mm, 36mm and 58mm place on these results and the thermal barrier coating surface topography photo sees that also tiny Surface Vertical crackle almost is distributed on the whole draw direction of sample simultaneously.Because visible disbonding seldom, the quantity of Interface Crack is also less relatively, and these signals are positioned analysis, and the result is shown in Figure 11 b.Can see also at random the draw direction that is distributed in sample of signal, but at 32～45mm place and 52～67mm place relatively want early time that acoustic emission signal occurs, the distribution of signal is intensive relatively, and is consistent with the position of as seen peeling off on the thermal barrier coating pattern photo.

Specific embodiments of the invention 2:

The acoustic emission of thermal barrier coating elevated temperature heat fatigue failure detects in real time.Thermal barrier coating has been carried out elevated temperature heat round-robin acoustic emission detection experiment, and thermal cycle is included in 800 ℃ the high temperature furnace heating 5 minutes and 10 minutes two processes of natural cooling in air.In the experiment, use the substrate surface of platinum filament waveguide rod spot welding thermal barrier coating sample, and be connected with calibrate AE sensor with the sound wave couplant, realize the detection of thermal barrier coating in whole thermal cycle process damage by mechanical hook-up.

Thermal barrier coating found that in the acoustic emission detection of whole heat fatigue, acoustic emission signal both can occur also occurring in the heating period at cooling stage, as shown in figure 12, this explanation thermal barrier coating can form at cooling stage in its damage of elevated temperature heat circulation time, also can form, thereby the cooling stage that only detects the thermal cycle process is inaccurate to the prediction of heat barrier coatings damage state in the heating period.

Utilize the acoustic emission signal that acoustic emission signal analytical approach provided by the invention lost efficacy to thermal cycle to carry out wavelet analysis, draw the wavelet character energy spectral coefficient distribution plan of signal, and can compare by spectral coefficient distribution plan (Fig. 9) with the feature of three kinds of patterns in embodiment 1, having determined, judge the damage mode of signal.The result that the acoustic emission signal that heating and cooling stage are occurred is carried out pattern-recognition is shown in Figure 13,14, and the damage that forms in heating process of thermal barrier coating mostly is the Surface Vertical crackle as can be seen, and the damage that forms in the cooling procedure mostly is Interface Crack.

Claims (6)

1. the acoustic emission real-time detection method of heat barrier coatings damage and failure procedure thereof may further comprise the steps:

1) the thermal barrier coating sample is connected, is coupled with calibrate AE sensor;

2) to thermal barrier coating sample imposed load, calibrate AE sensor receives the acoustic signals that sends in the sample;

3) acoustic emission signal that receives is carried out wavelet transformation, obtain Wavelet Energy Spectrum coefficient distribution plan, it can be compared by the spectral coefficient distribution plan with the feature of determining damage mode, judge the damage mode of heat barrier coatings damage signal;

4) add up the acoustic emission signal event number that each damage produces, draw the experimental formula that event number changes with imposed load;

5) extract component of signal under the pairing wavelet decomposition yardstick of Wavelet Energy Spectrum coefficient maximal value, calculate this component of signal with correlation analysis and arrive mistiming between each sensor, adopt traditional time difference positioning method then, determine the position of source of damage.

2. the acoustic emission real-time detection method of heat barrier coatings damage according to claim 1 and failure procedure thereof, described step 1) is: during at high temperature to the thermal barrier coating imposed load, one end of platinum filament waveguide rod is connected with the spot welding of thermal barrier coating sample, the other end is connected with calibrate AE sensor by mechanical hook-up, and by ultrasonic couplant and calibrate AE sensor coupling; When normal temperature loads down, directly calibrate AE sensor is connected, is coupled with sample by mechanical hook-up and ultrasonic couplant.

3. the acoustic emission real-time detection method of heat barrier coatings damage according to claim 1 and failure procedure thereof, the small echo of using in the wavelet transformation in the described step 3) is meant discrete dyadic wavelet, its maximum decomposition scale J _MaxSample frequency f according to acoustic emission signal _s, filter length L _fAnd sampling length N determines that concrete computing formula is:

4. the acoustic emission real-time detection method of heat barrier coatings damage according to claim 1 and failure procedure thereof, wavelet character power spectrum coefficient is the energy of signal each component after wavelet decomposition and the ratio of signal gross energy in the described step 3).

5. the acoustic emission real-time detection method of heat barrier coatings damage according to claim 1 and failure procedure thereof, described step 4) is: the acoustic emission signal of determining each damage according to step 3), add up and draw the relation curve that each acoustic emission signal event number of being produced of damage changes with plus load one by one, curve is carried out match draw between the two experimental formula.

6. the acoustic emission real-time detection method of heat barrier coatings damage according to claim 1 and failure procedure thereof, described step in step 5) is:

5a) according to the Wavelet Energy Spectrum coefficient distribution plan of each sensor signal, verify that the signal that each sensor receives is that same source of damage is sent, and do not have emergence pattern to change;

5b) find out the peaked wavelet decomposition yardstick of Wavelet Energy Spectrum coefficient, and extract each sensor and receive the Wavelet Component of this signal under this yardstick;

5c) table look-up or measure and determine the velocity of propagation of sound wave in sample;

5d) according to correlation analysis, solve each sensor and receive mistiming by the determined Wavelet Component of step 5b;

5e) according to traditional time difference positioning method, be that source of damage positions to acoustic emission source.

Images