CN104197872B - A kind of method that ultrasonic wave measures coat thickness and inner boundary roughness simultaneously - Google Patents

Abstract

A kind of method that ultrasonic wave measures coat thickness and inner boundary roughness simultaneously, belongs to material Ultrasonic NDT and assessment technique field.The invention is using a ultrasonic pulse-echo detecting system for including sample bench, the coarse coat sample of inner boundary, delay block probe, defectoscope, digital oscilloscope and computer, for the thickness and roughness problem of the coarse coat of nondestructive characterisation (NDC) inner boundary, the sound pressure reflection coefficient amplitude spectrum of the coarse coating layer of inner boundary is deduced | r (f；Rq, d) |, correlation operation is carried out to the theoretical sound pressure reflection coefficient amplitude spectrum with experiment in different frequency bands width, obtain coefficient correlation maximum η under each bandwidth_max(Rq, d) corresponding coat thickness d_iWith roughness Rq_i, the thickness of measurement, roughness in different frequency window are averaging and obtained respectivelyWithThe two is required coat thickness and inner boundary roughness.The method has filled up the blank of such coat thickness and roughness nondestructive characterisation (NDC) method.

Description

A kind of method that ultrasonic wave measures coat thickness and inner boundary roughness simultaneously

Technical field

A kind of method for measuring coat thickness and inner boundary roughness simultaneously the present invention relates to ultrasonic wave, it belongs to ultrasound The technical field of Non-Destructive Testing.

Background technology

Practice have shown that interface roughness is to improve the effective measures of interface bond strength, it is also to improve coat mistake under arms The effective measures of impact, corrosion and heat resistant and the service life of being met with stresses in journey.Roughness degree direct influence interface cohesion is strong Degree, interface residual stress and interface oxidation degree etc., and the residual tension that roughness causes may result in coat and break Split and peel off, it means that interface roughness have also been introduced the factor unfavorable to the interface life-span while interface cohesion is strengthened. Therefore, reasonable control interface roughness is the important means for ensureing coat service life, accurately and reliably inner boundary roughness Nondestructive characterisation (NDC) and assessment method have become the urgent engineering demand in the field.

Existing various lossless detection methods can be used for the measurement of surface roughness, can be basically divided into contact type measurement and non- The class of contact type measurement two：Mainly there are comparison method, impression method, tracer method etc. in contact type measurement；In non-contact measurement often There are spectroscopic methodology, atomic force microscopy, ultrasonic reflection spectrometry etc..But the above method be not suitable for coat and The roughness concentration of basal body interface.Coat preparation process can cause the change of Substrate Surface Roughness again, when boundary in coat When face has roughness, the sign of its thickness and roughness is more difficult, and having not yet to see can be thick to coat inner boundary Rough sample carries out the report of Efficient Characterization.

The content of the invention

The invention is analyzed ultrasonic wave and is existed for the thickness and roughness problem of the coarse coating layer of ultrasonic characterization inner boundary Propagation law in layered medium, is deduced the sound pressure reflection coefficient amplitude spectrum of the coarse coating layer of inner boundary first | r (f；Rq,d)|.By relative coefficient formula η, (Rq d), is examined to the sound pressure reflection coefficient amplitude spectrum in different bandwidth with experiment The sound pressure reflection coefficient amplitude spectrum for measuring carries out correlation analysis, maximum correlation coefficient η under Inversion Calculation each bandwidth_max (Rq, d) corresponding coat thickness d_iWith Rq_i.The coat thickness of inverting is averaging with roughness is respectively obtainedWith The two is required coat thickness and inner boundary roughness.The method overcome and draw because coating layer inner boundary is coarse The ultrasonic characterization thickness for rising and roughness problem.

The technical solution adopted for the present invention to solve the technical problems is：A kind of ultrasonic wave simultaneously measure coat thickness with The method of inner boundary roughness, the method using one include sample bench, the coarse coat sample of inner boundary, delay block probe, The ultrasonic pulse-echo detecting system of defectoscope, digital oscilloscope and computer；It is characterized in that：Methods described uses following step Suddenly：

Step 1, in x-z-plane, ultrasonic probe acoustic pressure in transmission be P_I=1 ultrasonic wave impinges perpendicularly on medium 1, medium 2nd, in the second interface structure of the composition of medium 3, medium 1 is delay block medium, and medium 2 is tested coating layer material, and medium 3 is matrix Material, the reflection echo acoustic pressure P that ultrasonic probe is received_RIt is the synthesis of following each reflection echo：The medium 1 and medium 2 The reflection echo P of interface 1₁=r₁₂, the reflection echo at the interface 2 of medium 2 and medium 3Subscript n-1 Reflectivity r is respectively with n-2₂₃With r₁₂N-1 and n-2 powers, n takes 2,3,4,5,6,7,8,9,10 limited number of times in engineer applied, its Middle r₁₂It is the pressure reflection ratio at interface 1, t₁₂And t₂₁It is the acoustic pressure transmissivity at interface 1, subscript 12 is passed for ultrasonic wave from medium 1 Sound pressure reflection or the transmission of medium 2 are multicast to, subscript 21 is sound pressure reflection or the transmission that ultrasonic wave travels to medium 1 from medium 2, under Mark 23 is sound pressure reflection or the transmission that ultrasonic wave travels to medium 3 from medium 2；D is coat thickness；k_2zIt is ultrasonic wave in medium Wave number in 2 along the z-axis direction, wave number is expressed as k_2z=2 π f/c₂, f is ultrasonic frequency, c₂It is coat longitudinal wave velocity；Interface 2 is rough interfaces, the pressure reflection ratio r at interface 2₂₃It is expressed as：

WhereinIt is the pressure reflection ratio under the smooth state of interface 2, Rq is the profile r.m.s. roughness at interface 2；Coating The sound pressure reflection coefficient R of layer is expressed as：

As a plural number, when solving n=2, amplitude spectrum expression formula | the R (f) | of R is formula (3) to R sheets：

Wherein pressure reflection ratio is respectively the function of corresponding Media density ρ and velocity of sound c with transmissivity, it is known that medium 1,2,3 Density p and the velocity of sound c, R (f) | the only function of frequency f, roughness Rq and coat thickness d, be expressed as | R (f；Rq,d)|；

Step 2, the coat sample of interface roughness internal first carry out ultrasound detection, using the two-parameter of correlation analysis Inversion method calculates coat thickness d and roughness Rq simultaneously, and the computing formula of coefficient correlation is：

Rq⁰With d⁰Respectively it is detected the actual roughness value of test specimens and thickness value；|R_the(f；Rq, d) | with | R_exp(f；Rq⁰, d⁰) the sound pressure reflection coefficient amplitude spectrum that theoretical sound pressure reflection coefficient amplitude spectrum is obtained with experiment detection in effective bandwidth is respectively,WithThe acoustic pressure that theoretical sound pressure reflection coefficient amplitude spectrum is obtained with experiment detection respectively in effective bandwidth The average value of reflectance magnitude spectrum；In the range of 2 times that tested test specimens estimate roughness Rq and thickness d, a series of companies are taken The coating bed roughness R of continuous change_qWith thickness d value, corresponding theoretical sound pressure reflection coefficient amplitude spectrum is obtained | R_the(f；Rq,d)| And the sound pressure reflection coefficient amplitude spectrum obtained with experiment detection | R_exp(f；Rq⁰,d⁰) | correlation analysis are carried out, coefficient correlation is maximum Value η_max(Rq, d) corresponding coating bed roughness Rq and thickness d are the actual roughness and thickness value of tested sample；

Step 3, respectively calculating initial spike P_IThe corresponding 5 groups of bands of -8dB of amplitude spectrum, -7dB, -6dB, -5dB, -4dB Width, in each bandwidth | R_exp(f；Rq⁰,d⁰) | data perform the correlation calculations process of step 2, obtain 5 groups of coating thickness Degree d₁、d₂、d₃、d₄、d₅With roughness value Rq₁、Rq₂、Rq₃、Rq₄、Rq₅, respectively to coat thickness in 5 groups of different frequency windows, thick Rugosity is averaging and obtainsWithThe two is required coat thickness and inner boundary roughness.

The invention has the advantages that：The invention overcomes traditional roughness non-destructive measuring method and cannot detect coat The problem of inner boundary roughness, and first by the sound pressure reflection coefficient amplitude spectrum of the coarse coating layer of inner boundary | R (f；Rq,d) | combined with correlation analysis, realized the thickness and roughness ultrasonic measurement of the coarse coating layer of inner boundary.And to coating Layer does not require that it is metal or nonmetallic sample that can measure matrix material with the optics of matrix, electricity and magnetic performance.Should Inventing measurable rough interfaces type includes：The randomness that the processing technologys such as shot-peening, burn into spraying and grinding are caused is coarse Degree.Multigroup ultrasonic delay line probe can be accurately measured the sample of different coat thickness and roughness.The present invention Coat thickness and inner boundary roughness can be easily and flexibly measured, device therefor is light and handy, portable, workable, low cost, Certainty of measurement is high, scope is wide, and Site Detection result is reliable and stable, with larger economic benefit and social benefit.

Brief description of the drawings

Patent of the present invention is described further with reference to the accompanying drawings and examples.

Fig. 1 is that ultrasonic wave propagates schematic diagram in medium 1, medium 2, the dielectric structure of matrix 3 three.

Fig. 2 is ultrasonic pulse-echo detecting system.

Fig. 3 is initial spike P_EnterTime domain waveform and amplitude spectrum.

Fig. 4 is the reflection echo acoustic pressure P of 28 μm of samples of nominal roughness_InsteadTime domain waveform and sound pressure reflection coefficient amplitude spectrum |R_exp(f；Rq⁰,d⁰)|。

Fig. 5 be 28 μm of samples of nominal roughness different roughness Rq coefficient correlation ηs corresponding with the combination of thickness d value (Rq, d)。

Fig. 6 is the rough surface metallograph and fluctuating pattern of 28 μm of samples of nominal roughness.

Fig. 7 is the inner boundary coarse structure physical model of delay block/AlN coats/steel matrix composition.

Fig. 8 is the echo-signal of the excitation sound source with different roughness model measurement of numerical simulation.

Fig. 9 is the theoretical and measuring amplitude spectrum of coarse Rq=30.0 μm of physical model.

In figure：(1) sample bench；(2) the coarse coat sample of inner boundary；(3) delay block probe；(4) defectoscope；(5) number Word oscillograph；(6) computer.

Specific embodiment

A kind of method that ultrasonic wave measures coat thickness and inner boundary roughness simultaneously, using a bag shown in Fig. 2 Include the coarse coat sample of sample bench (1), inner boundary (by Ni based high-temperature alloys shot blasting on surface prepare, thickness of thin layer about 0.5~ 0.7mm, nominal roughness value is respectively 8 μm, 14 μm of 3 pieces of tested test specimens with 28 μm, and the sample back side is ground with 2000# sand paper Scabble whole, and be processed by shot blasting, eliminate the back side and rise and fall influence) (2), nominal frequency 20MHz delay blocks probe, wafer diameter The ultrasonic pulse-echo detection system of 6mm (3), USIP40 defectoscopes (4), DPO4O32 digital oscilloscopes (5) and computer (6) System；

The measuring process that it is used is as follows：

(1) delay block probe is connected and is rectified an instrument with system.The anti-of delay block probe is gathered by digital oscilloscope Penetrate acoustic pressure and be denoted as P_Enter, see in Fig. 3 (a).Its sound pressure reflection coefficient amplitude spectrum is obtained by Fourier transform, is seen in Fig. 3 (b), should - the 8dB of amplitude spectrum, -7dB, -6dB, -5dB, the corresponding bandwidth of -4dB amplitudes be respectively [12.0MHz, 25.0MHz], [12.8MHz,24.4MHz]、[13.4MHz,23.8MHz]、[14MHz,23.4MHz]、[14.8MHz,22.8MHz]。

(2) it is 28 μm of burnishing surfaces of sample delay block probe to be coupled into nominal roughness, and mat surface is soaked in water, is constituted One delay block/thin metal layer/dielectric structure of water base body three.One coat reflection echo acoustic pressure is gathered by digital oscilloscope P_Instead, see in Fig. 4 (a).By incident sound pressure P_ITo reflecting acoustic pressure P_RDo business's treatment, the sound pressure reflection coefficient amplitude after being normalized Spectrum | R_exp(f；Rq⁰,d⁰) |, see in Fig. 4 (b).Selection -8dB bandwidth [2.96MHz, 7.07MHz] is corresponding | R_exp(f；Rq⁰,d⁰) | data carry out correlation analysis.Known delay block is epoxy resin, and ultrasonic method measurement longitudinal wave velocity is 2316m/s, Archimedes Drainage measurement density is 1045kg/m³.The longitudinal wave velocity of Ni base thin layers is 5535m/s, and measurement density is 8908kg/m³.Water Longitudinal wave velocity is 1479m/s, and density is 1004kg/m³.Thickness of thin layer scope 0~1.2mm of selection, roughness range selection 0~ 60 μm, the thickness of thin layer d and roughness Rq values for taking a series of consecutive variations bring formula (3) into, obtain corresponding theoretical acoustic pressure anti- Penetrate coefficient amplitude spectrum | R_the(f；Rq,d)|.Will | R_the(f；Rq, d) | with | R_exp(f；Rq⁰,d⁰) | bring formula (4) into, obtain difference Thickness of thin layer d coefficient correlation ηs corresponding with the combination of roughness Rq values (Rq, d).Read maximum correlation coefficient η_max(Rq, d)= Rq=25.6 μm of 0.999 corresponding thickness d=0.779mm and roughness, are shown in Fig. 5 in (a) and 5 (b).

(3) in the corresponding another 4 groups of bandwidth of p- 7dB, -6dB, -5dB, -4dB | R_exp(f；Rq⁰,d⁰) | data perform step (2) correlation operation process, calculates corresponding coating layer thickness and roughness, respectively to 5 groups of coating layer thicknesses of calculating, roughness It is averaging and obtainsWithOther nominal roughness are 8 μm and are shown in Table 1 with the testing result of 14 μm of samples. Then 3 pieces of roughness values of coarse sample are measured using laser confocal microscope, is shown in Table 2.28 μm of samples of nominal roughness Rough surface metallograph is shown in (a) and (b) in Fig. 6 in Fig. 6 respectively with fluctuating pattern.Coarse sample is measured using spiral micrometer Thickness value, Mechanical Method thickness in being shown in Table 1.Ultrasonic measurement result is surveyed into roughness Rq results with Mechanical Method thickness d, microscopic method to enter Row contrast and analytical error, ultrasonic thickness measurement relative error is respectively less than 5.5%, and is minus deviation, mainly due to spiral micrometer Instrument measurement thickness of thin layer value is the maximum corresponding thickness value that rises and falls of rough surface；Ultrasonic measurement roughness relative error no more than 8.5%.

The thickness d of the different roughness sample of table 1 and roughness Rq results

(4) it is more difficult due to making a series of different coat sample of inner boundary roughness, patent of the present invention for The validity that coat thickness is characterized with roughness is using based on Fdtd Method (Finite Difference of Time Domain, FDTD) method for numerical simulation further illustrate.Fig. 7 gives the interior of delay block/AlN coats/Ni substrate composition Interface roughness structure physical model.Coat builds Gaussian random roughness with Ni-based body interface, is respectively provided with roughness Rq =6.3 μm, 11.7 μm, 20.7 μm and 30.0 μm of 4 groups of physical models.The a width of 10mm of setting model, delay block thickness is 5mm, is applied Coating geometric average thickness is 0.70mm, and Ni-based body thickness is infinity.Other parameters,acoustics used in simulation are shown in Table 2, material Material longitudinal wave velocity is measured by Experimental Ultrasonic, and density is measured using Archimedes's drainage.

The material acoustics parameter that table 2 is used in simulating

In Fig. 8 the Gaussian pulse of centre frequency 10MHz shown in (a) as numerical simulation excitation sound source.In order to carry Computational accuracy high, the 1/20 of the size of mesh opening selection minimum of computation wavelength of FDTD.To avoid reflection from model boundary and spreading out It is emitted back towards wave action, model or so and is set to compressional wave fixed boundary, setting up and down is absorbing boundary.4 groups of physical models are by simulation The echo-signal being calculated is shown in Fig. 8 (b).

(5) to Rq=6.3 μm in Fig. 8 (b), 11.7 μm, 20.7 μm enter with the echo-signal of 30.0 μm of four groups of physical models The spectrum analysis of row step (2), by theoretical sound pressure reflection coefficient amplitude spectrum | R_the(f；Rq, d) | the acoustic pressure with corresponding model is anti- Penetrate coefficient amplitude spectrum | R_mod(f；Rq⁰,d⁰) | bringing formula (4) into carries out correlation analysis, designs Rq=30.0 μm of model of roughness Coefficient correlation maximum η_max(Rq, d)=1.000 the corresponding theoretical amplitude spectrum with model see Fig. 9.Coefficient correlation maximum Rq=29.7 μm of corresponding roughness, thickness d=0.707mm.Other roughness are 6.3 μm, 11.7 μm with 20.7 μm of models Measurement result is shown in Table 3.The ultrasonic measurement result of model and design thickness d, roughness Rq contrast and analytical error, ultrasound Thickness measuring relative error is no more than 1.0%；Ultrasonic measurement roughness relative error is no more than 2.6%.For the velocity of sound in 5000- The most metals coat (steel, aluminium, nickel, copper etc.) of 7000m/s, the coat thickness that 10MHz probes can be detected is more than 350 μ M, the valid analysing range to roughness is about [10 μm, 70 μm].20MHz probes can detect the coating thickness for applying more than 175 μm, Valid analysing range to roughness is about [5 μm, 35 μm]；For the velocity of sound the tungsten carbide of 3000-5000m/s, aluminium nitride with And the coat such as zirconium oxide, the coat thickness that 10MHz probes can be detected is more than 250 μm, to the valid analysing range of roughness About [6 μm, 50 μm].20MHz probes can detect that the coating thickness of painting is more than 125 μm, to the valid analysing range of roughness about It is [3 μm, 25 μm].

The coat thickness d of the different roughness model of table 3 and roughness Rq results

Claims (1)

1. a kind of method that ultrasonic wave measures coat thickness and inner boundary roughness simultaneously, the method includes sample using one The ultrasonic pulse-echo of the coarse coat sample of platform, inner boundary, delay block probe, defectoscope, digital oscilloscope and computer Detecting system；It is characterized in that：Methods described uses the following steps：

Step 1, in x-z-plane, ultrasonic probe acoustic pressure in transmission be P_I=1 ultrasonic wave impinges perpendicularly on medium 1, medium 2, is situated between In the second interface structure of the composition of matter 3, medium 1 is delay block medium, and medium 2 is tested coating layer material, and medium 3 is matrix material Material, the reflection echo acoustic pressure P that ultrasonic probe is received_RIt is the synthesis of following each reflection echo：The boundary of the medium 1 and medium 2 The reflection echo P of face 1₁=r₁₂, the reflection echo at the interface 2 of medium 2 and medium 3Subscript n-1 with N-2 is respectively reflectivity r₂₃With r₁₂N-1 and n-2 powers, n takes 2,3,4,5,6,7,8,9,10 limited number of times in engineer applied, wherein r₁₂It is the pressure reflection ratio at interface 1, t₁₂And t₂₁It is the acoustic pressure transmissivity at interface 1, subscript 12 is propagated for ultrasonic wave from medium 1 Sound pressure reflection or transmission to medium 2, subscript 21 are sound pressure reflection or the transmission that ultrasonic wave travels to medium 1 from medium 2, subscript 23 is sound pressure reflection or the transmission that ultrasonic wave travels to medium 3 from medium 2；D is coat thickness；k_2zIt is ultrasonic wave in medium 2 In wave number along the z-axis direction, wave number is expressed as k_2z=2 π f/c₂, f is ultrasonic frequency, c₂It is coat longitudinal wave velocity；Interface 2 It is rough interfaces, the pressure reflection ratio r at interface 2₂₃It is expressed as：

$r_{23}=r_{23}^0\exp \left(\frac{-8{\mathrm{\π}}^2{\mathrm{Rq}}^2{f}^2}{c_2^2}\right)---\left(1\right)$

WhereinIt is the pressure reflection ratio under the smooth state of interface 2, Rq is the profile r.m.s. roughness at interface 2；Coat Normalization sound pressure reflection coefficient R is expressed as：

$R=\frac{P_R}{P_I}=r_{12}+t_{12}{r}_{23}{t}_{21}\exp \left(2{\mathrm{ik}}_{2z}d\right)\underset{n=0}{\overset{\mathrm{\∞}}{\Σ}}{\[r_{23}{r}_{21}\exp \left(2{\mathrm{ik}}_{2z}d\right)\]}^n---\left(2\right)$

As a plural number, when solving n=2, amplitude spectrum expression formula | the R (f) | of R is formula (3) to R sheets：

$\left|R\left(f\right)\right|={\[\frac{r_{12}^2+{\left(\right(1-r_{12}^2\left)r_{23}^0\exp \right(\frac{-8{\mathrm{\π}}^2{\mathrm{Rq}}^2{f}^2}{c_2^2}\left)\right)}^2+2(r_{12}-r_{12}^3)r_{23}^0\exp \left(\frac{-8{\mathrm{\π}}^2{\mathrm{Rq}}^2{f}^2}{c_2^2}\right)cos\left(\frac{4\mathrm{\π}fd}{c_2}\right)}{r_{12}^2}\]}^{1/2}---\left(3\right)$

Wherein pressure reflection ratio is respectively the function of corresponding Media density ρ and velocity of sound c with transmissivity, it is known that medium 1,2,3 it is close Degree ρ and velocity of sound c, the function of | R (f) | only frequency f, roughness Rq and coat thickness d is expressed as | R (f；Rq,d)|；

Step 2, the coat sample of interface roughness internal first carry out ultrasound detection, using the two-parameter inverting of correlation analysis Method calculates coat thickness d and roughness Rq simultaneously, and the computing formula of coefficient correlation is：

$\mathrm{\η}(R_q,d)=\frac{\underset{i=1}{\overset{N}{\Σ}}\[\left|R_{the}(f_i;Rq,d)\right|-\stackrel{\‾}{\left|R_{the}(Rq,d)\right|}\]\[\left|R_{exp}(f_i;{\mathrm{Rq}}^0,d^0)\right|-\stackrel{\‾}{\left|R_{exp}({\mathrm{Rq}}^0,d^0)\right|}\]}{\sqrt{\underset{i=1}{\overset{N}{\Σ}}{\[\left|R_{the}(f_i;Rq,d)\right|-\stackrel{\‾}{\left|R_{the}(Rq,d)\right|}\]}^2\underset{i=1}{\overset{N}{\Σ}}{\[\left|R_{\mathrm{ex}p}(f_i;{\mathrm{Rq}}^0,d^0)\right|-\stackrel{\‾}{\left|R_{exp}({\mathrm{Rq}}^0,d^0)\right|}\]}^2}}---\left(4\right)$

Rq⁰With d⁰Respectively it is detected the actual roughness value of test specimens and thickness value；|R_the(f；Rq, d) | with | R_exp(f；Rq⁰, d⁰)| The sound pressure reflection coefficient amplitude spectrum that theoretical sound pressure reflection coefficient amplitude spectrum is obtained with experiment detection respectively in effective bandwidth,WithThe acoustic pressure that theoretical sound pressure reflection coefficient amplitude spectrum is obtained with experiment detection respectively in effective bandwidth The average value of reflectance magnitude spectrum；In the range of 2 times that tested test specimens estimate roughness Rq and thickness d, a series of companies are taken The coating bed roughness R of continuous change_qWith thickness d value, corresponding theoretical sound pressure reflection coefficient amplitude spectrum is obtained | R_the(f；Rq, d) | And the sound pressure reflection coefficient amplitude spectrum obtained with experiment detection | R_exp(f；Rq⁰, d⁰) | correlation analysis are carried out, coefficient correlation is maximum Value η_max(Rq, d) corresponding coating bed roughness Rq and thickness d are the actual roughness and thickness value of tested sample；

Step 3, respectively calculating initial spike P_IThe corresponding 5 groups of bandwidth of -8dB of amplitude spectrum, -7dB, -6dB, -5dB, -4dB, to every In individual bandwidth | R_exp(f；Rq⁰, d⁰) | data perform the correlation calculations process of step 2, obtain 5 groups of coat thickness ds₁、d₂、 d₃、d₄、d₅With roughness value Rq₁、Rq₂、Rq₃、Rq₄、Rq₅, flat is asked to coat thickness, roughness in 5 groups of different frequency windows respectively ObtainWithThe two is required coat thickness and inner boundary roughness.