CN103148815B - Based on the thickness of thin layer supersonic detection method of sound pressure reflection coefficient autocorrelation function - Google Patents

Abstract

Based on a thickness of thin layer supersonic detection method for sound pressure reflection coefficient autocorrelation function, belong to technical field of nondestructive testing.The method uses ultrasonic pulse-echo technology and sound pressure reflection coefficient autocorrelation signal disposal route to carry out thin layer thickness measurement.First use ultrasonic pulse-echo system acquisition one to comprise the reflection echo signal of thin layer upper surface and lower surface, then gather the upper surface echoed signal of a reference block.Then respectively Fast Fourier Transform (FFT) is carried out to the signal collected, obtain thin layer sound pressure reflection coefficient autocorrelation function.Finally in autocorrelation function, read frequency corresponding to each maximum value, and the bond material velocity of sound calculates thickness of thin layer.This method overcomes traditional ultrasonic interference thickness measuring method must can read two accurate limitation obtaining thickness of thin layer of resonance frequency ability in signal spectrum, has the advantages such as measuring accuracy is high, workable, equipment cost is low, wide accommodation.

Description

Based on the thickness of thin layer supersonic detection method of sound pressure reflection coefficient autocorrelation function

Technical field

The present invention relates to a kind of thickness of thin layer supersonic detection method based on sound pressure reflection coefficient autocorrelation function, belong to technical field of nondestructive testing.

Background technology

When using ultrasound impulse echo technique measures thickness of thin layer, due to upper surface echo and the lower surface echoed signal generation aliasing of thin layer, traditional pulse echo signal Time Domain Analysis cannot determine thickness of thin layer.

At present, convectional signals disposal route for using ultrasound impulse echo technique measurement thickness of thin layer is the frequency spectrum analysis method based on principle of interference, as HainesNF etc. uses 20MHz ultrasonic probe in " Theapplicationofbroadbandultrasonicspectroscopytothestud yoflayeredmedia " literary composition, and apply sound pressure reflection coefficient amplitude spectrum and phase spectrum process means acquisition epoxy resin thickness of thin layer.Xu Zhihui, Lin Li etc. in " the surface coating thickness supersonic damage-free detection method based on power spectrumanalysis " literary composition applied power spectral analysis technology to ZrO ₂coating carries out thickness measure.But above-mentioned signal processing method all needs signal spectrum at least to occur two resonance frequencies in ultrasonic probe bandwidth range, then utilize the difference between these two resonance frequencies, relation between the thin layer velocity of sound and thickness of thin layer, calculate thickness of thin layer.The bandwidth requirement of the method to probe is higher, thus also improves the requirement of frequency probe thereupon, causes measuring equipment cost high, and greatly limit its application.

Sound pressure reflection coefficient have good autocorrelation performance and with the features such as noise non-correlation, the situation adopting this method can only read a resonance frequency for signal spectrum in ultrasonic probe bandwidth range still can carry out thin layer thickness measurement, and stands good for the signal-noise ratio signal being low to moderate 20dB.

Summary of the invention

The object of this invention is to provide a kind of thickness of thin layer supersonic detection method based on sound pressure reflection coefficient autocorrelation function.Compared with existing ultrasonic interferometry thickness of thin layer method, when measuring the material of same thickness, the requirement to ultrasonic probe frequency and frequency span can be reduced, or under same ultrasonic probe condition, can the less thin layer of detect thickness.

Technical scheme of the present invention is: a kind of thickness of thin layer supersonic detection method based on sound pressure reflection coefficient autocorrelation function adopts an impulse ultrasound water logging Echo System comprising reflectoscope, ultrasonic pulse immersion probe, Al matter thin layer specimen, tank, digital oscilloscope and computing machine, and its step detected is as follows:

(1) utilize described ultrasonic pulse water logging Echo System to thin layer specimen Vertical Launch ultrasonic signal, and gather the sample signal (aliasing signal) that boundary reflection echoed signal that a boundary reflection echoed signal and thin layer lower surface be made up of thin layer upper surface and water and matrix form forms;

(2) the upper surface reference signal (echoed signal) of a described ultrasonic pulse-echo system acquisition reference block is utilized;

(3) sample signal collected described step (1) and step (2) and reference signal carry out Fast Fourier Transform (FFT) respectively, then use the frequency domain data of frequency domain data divided by reference signal of sample signal, obtain complex function S (f);

(4) complex function S (f) that step (3) is tried to achieve is updated in formula 1, obtains thin layer sound pressure reflection coefficient autocorrelation function R _s(φ):

$R_S\left(\mathrm{\φ}\right)=\underset{F\→\mathrm{\∞}}{\lim }\underset{0}{\overset{F}{\∫}}S\left(f\right)S^{*}(f+\mathrm{\φ})df---\left(1\right)$

In formula: R _s(φ) thin layer sound pressure reflection coefficient autocorrelation function is represented, the business that complex function S (f) is sample signal frequency domain data and reference signal frequency domain data;

(5) R asked in step (4) _s(φ) frequency φ corresponding to n maximum value is read in _i, i=0,1,2 ... n, by φ _i, i=0,1,2 ... n and thin layer velocity of sound c substitutes into formula (2) just can try to achieve thickness of thin layer:

$d=\frac{\underset{i=0}{\overset{i=n}{\Σ}}i*c/\left(2{\mathrm{\φ}}_i\right)}{n}---\left(2\right)$

In formula, c is the thin layer velocity of sound (mm/s), d is thickness of thin layer (mm).

Thin layer specimen is placed in tank, after carrying out system connection and instrumental correction, by ultra-sonic defect detector, utilizes ultrasonic pulse immersion probe, launch to the thin layer specimen in tank, receive ultrasound wave, observe waveform by digital oscilloscope and gather Wave data.UT (Ultrasonic Testing) reference block repeats said process, gathers the upper surface echoed signal of a reference block.Respectively Fast Fourier Transform (FFT) is carried out to sample signal and reference signal, obtain the business of sample signal frequency domain data and reference signal frequency domain data, obtain complex function S (f), the S (f) tried to achieve is substituted into formula (1), obtain the sound pressure reflection coefficient autocorrelation function R of thin layer specimen _s(φ).Read R _s(φ) φ that in, each maximum value is corresponding _i, when the known thin layer velocity of sound, utilize thickness of thin layer d and the φ set up _ibetween relational expression (2), thickness of thin layer d can be obtained.

The invention has the beneficial effects as follows: this thickness of thin layer supersonic detection method based on sound pressure reflection coefficient autocorrelation function uses the ultrasonic pulse water logging Echo System be made up of reflectoscope, ultrasonic pulse immersion probe, thin layer specimen, tank, digital oscilloscope and computing machine.First utilize this system to thin layer specimen Vertical Launch ultrasonic signal and gather the aliasing signal that boundary reflection signal that a boundary reflection signal and thin layer lower surface be made up of thin layer upper surface and water and matrix form forms, recycle the upper surface echoed signal of this system acquisition reference block, respectively Fast Fourier Transform (FFT) is carried out to the sample signal collected and reference signal, obtain the business of sample signal frequency domain data and standard signal frequency domain data, obtain complex function S (f), and ask for the autocorrelation function R of complex function S (f) _s(φ); Thickness of thin layer is gone out finally by formulae discovery.Utilize bandwidth be 10.3 ~ 20.5MHz ultrasonic water immersion probe thickness measure is carried out to thin-layered medium, if thin layer is aluminium, now measurable thickness of thin layer is less than 310 μm, and still can Measurement accuracy thickness of thin layer when signal to noise ratio (S/N ratio) is 20dB.Overcome in traditional ultrasonic interference thickness measuring method when signal spectrum only has the problem being difficult to accurately obtain thickness of thin layer when a resonance frequency and low signal-to-noise ratio in ultrasonic probe bandwidth range, device therefor is simple, workable, cost is low, be easy to practical, measuring accuracy is high, measurement range is wide, has larger economic benefit and social benefit.

Accompanying drawing explanation

Below in conjunction with drawings and Examples, the invention will be further described.

Fig. 1 is a kind of systematic schematic diagram of the thickness of thin layer supersonic detection device based on sound pressure reflection coefficient autocorrelation function.

Fig. 2 is the metallograph of aluminium matter thin layer xsect.

Fig. 3 adopts bandwidth to be that the ultrasonic pulse immersion probe of 10.3 ~ 20.5MHz launches ultrasonic signal and the time domain waveform (signal to noise ratio (S/N ratio) is 20dB) that collects to the Al matter thin layer specimen that thickness is 415 μm.

Fig. 4 is the sound pressure reflection coefficient autocorrelation function graph of Al matter thin layer specimen.

In figure: 1, ultra-sonic defect detector, 2, ultrasonic pulse immersion probe (centre frequency 15.4MHz, bandwidth is 10.3 ~ 20.5MHz), 3, Al matter thin layer specimen, 4, tank, 5, digital oscilloscope, 6, computing machine.

Embodiment

After measurement mechanism shown in Fig. 1 carries out system connection and instrumental correction, first by ultra-sonic defect detector 1, (centre frequency is 15.4MHz to the ultrasonic pulse immersion probe 2 of employing wafer diameter 11.4mm, bandwidth is 10.3 ~ 20.5MHz) to tested Al matter thin layer specimen 3, (obtaining its thickness from Fig. 2 is 415 μm, its velocity of sound is 6420m/s) transmit and receive ultrasound wave (Al matter thin layer specimen 3 is as in tank 4), observation and the collection of Al matter thin layer specimen Wave data is completed by DPO4032 digital oscilloscope 5.Then in UT (Ultrasonic Testing) reference block, repeat said process, gather the upper surface echoed signal of a reference block, Figure 3 shows that Al matter thin layer specimen echoed signal (signal to noise ratio (S/N ratio) is 20dB).Utilize computing machine 6 to carry out Fast Fourier Transform (FFT) to sample signal and reference signal respectively, obtain the business of sample signal frequency domain data and reference signal frequency domain data, obtain complex function S (f); Complex function S (f) of trying to achieve is updated to formula (1), obtains the sound pressure reflection coefficient autocorrelation function R of Al matter thin layer specimen 3 _s(f), as shown in Figure 4.Then R is read _s(φ) φ that in, two maximum value is corresponding ₁and φ ₂, be respectively 8MHz and 15.75MHz, as shown in Figure 3.By φ ₁, φ ₂namely can calculate Al matter thickness of thin layer with acoustic velocity of material c substitution formula (2) is 408 μm, and the absolute error between actual (real) thickness 415 μm is 7 μm, and relative error is 1.69%.

Claims (1)

1. the thickness of thin layer supersonic detection method based on sound pressure reflection coefficient autocorrelation function, it is characterized in that: adopt an impulse ultrasound water logging Echo System comprising reflectoscope (1), ultrasonic pulse immersion probe (2), Al matter thin layer specimen (3), tank (4), digital oscilloscope (5) and computing machine (6), its step detected is as follows:

(1) utilize described impulse ultrasound water logging Echo System to thin layer specimen Vertical Launch ultrasonic signal, and gather the sample signal that boundary reflection echoed signal that a boundary reflection echoed signal and thin layer lower surface be made up of thin layer upper surface and water and matrix form forms;

(2) the upper surface reference signal of a described ultrasonic pulse-echo system acquisition reference block is utilized;

(3) sample signal collected described step (1) and step (2) and reference signal carry out Fast Fourier Transform (FFT) respectively, then use the frequency domain data of frequency domain data divided by reference signal of sample signal, obtain complex function S (f);

(4) complex function S (f) that step (3) is tried to achieve is updated in formula 1, obtains thin layer sound pressure reflection coefficient autocorrelation function R _s(φ):

$R_S\left(\mathrm{\φ}\right)=\underset{F\→\mathrm{\∞}}{\lim }\underset{0}{\overset{F}{\∫}}S\left(f\right)S^{*}(f+\mathrm{\φ})df---\left(1\right)$

In formula: R _s(φ) thin layer sound pressure reflection coefficient autocorrelation function is represented _,the business that complex function S (f) is sample signal frequency domain data and reference signal frequency domain data;

(5) R asked in step (4) _s(φ) frequency φ corresponding to n maximum value is read in _i, i=0,1,2 ... n, by φ _i, i=0,1,2 ... n and thin layer velocity of sound c substitutes into formula (2) just can try to achieve thickness of thin layer:

$d=\frac{\underset{i=0}{\overset{i=n}{\Σ}}i*c/\left(2{\mathrm{\φ}}_i\right)}{n}---\left(2\right)$

In formula, c is the thin layer velocity of sound (mm/s), d is thickness of thin layer (mm).