CN100429516C - Ultrasonic time-of-flight diffraction detection method based on synthetic aperture focusing technique - Google Patents

Abstract

The present invention relates to a detecting method for ultrasonic transit time based on synthetic aperture focus, which relates to the technical field of the ultrasonic non-destructive detection of blemishes in plates and welded seams. The purpose of the detecting method for ultrasonic transit time based on synthetic aperture focus is to overcome the problem that the blemishes in plates and welded seams can not be accurately positioned and quantitatively analyzed by adopting an ultrasonic transit time method in the prior art. The ultrasonic transit time method is utilized by the present invention to obtain an original B scan image, and then linearization processing is carried out for the B scan image. A mathematical model reestablished by the B scan image of the ultrasonic transit time method based on synthetic aperture focus is established according to the formation process of the B scan image and the geometrical relationship between a probe and a blemish position then. Finally, the reestablishment of linearization-synthetic aperture focus (L-SAFT) of images is realized. The detecting method for ultrasonic transit time based on synthetic aperture focus provides accurate data for the structural integrity, the crack propagation situation and the residual life prediction of constructional elements. Positioning quantitative measurement can be carried out for an open type and a buried-in type blemishes. The longitudinal time resolution of images by adopting the L-SAFT reestablishment of the present invention can reach 0.01 microseconds, and lateral distance resolution can reach the size of arranged minimum scan step length.

Description

Ultrasonic transit time detection method based on synthetic aperture focusing

Technical field

The present invention relates to the Ultrasonic Nondestructive field of defective in sheet material and the weld seam.

Background technology

Ultrasonic transit time method is called for short ultrasonic TOFD (Time of Flight Diffraction), is based on the diffracted signal that receives the defective end defective is positioned quantitative a kind of detection means.This method can be judged the buried depth size and defective self size of defective in sheet material and the weld seam.But, utilize the D scan mode of this method to detect, when the defective end is not on the perpendicular bisector of two probe lines, its location is had very big error.During B scanning, though the defective end at a time is positioned on the perpendicular bisector of two probe lines, but because the acoustic beam of ultrasonic transducer has certain cover width, make defective curved in the B scan image, the lateral resolution of image is low, can't accurately hold the calculating location of flaw indication correspondence, thus the accurate location that can't implement defective with quantitatively.In addition, because the accurate judgement to the defect waves due in has also been disturbed in the existence of the redundant signals such as damped oscillation of sound wave.

Summary of the invention

In order to overcome that the ultrasonic transit time method of available technology adopting can't accurately be located and the quantitative problem of the defective in sheet material and the construction joint thereof, the present invention proposes a kind of ultrasonic transit time detection method based on synthetic aperture focusing (SAFT-Synthetic Aperture Focusing Technique), linearization-synthetic aperture focusing is called for short L-SAFT (Linearization-SAFT).Detection method of the present invention is carried out according to the following steps:

The first step, at first the zone to be detected with sheet material and construction joint thereof places between transmitting probe and the receiving transducer, earlier D scanning is carried out to determine the general lengthwise position of defective in detected zone; Set the scanning step-length of probe, cross the B scanning of defective top, detection system generates pending original B scan image with the detection data of receiving transducer; Wherein, along the horizontal zone on the corresponding B scan image of the lateral wave top that the more shallow degree of depth under the surface of sheet material and construction joint is propagated; Article one parabolic shape zone of above-mentioned horizontal zone below in the upper end diffracted wave correspondence image of defective; The second parabolic shape zone of above-mentioned horizontal zone below in the lower end diffracted wave correspondence image of defective; The horizontal zone of the bottorm echo correspondence image bottom of sheet material and construction joint;

Second step, original B scan image that the first step is obtained carry out linearization process, and the lateral wave horizontal zone of original B scan image is treated to straight line; The parabolic shape zone of the upper and lower end of defective diffracted wave is treated to curve respectively;

Curve after the 3rd step, the selection linearization process is a process object, opens the gray-scale value character matrix of process object, and reads the maximum and the minimum value of every row; Wherein, read the maximal value of every row pixel gray-scale value when the corresponding curve in defective upper end is handled, read the minimum value of every row pixel gray-scale value when the curve of defective lower end correspondence is handled; When reading described maximal value or minimum value each time, carry out following computing:

A, read maximum or the pairing time location of minimum gradation value pixel, calculate the buried depth size of defective end according to formula (1),

$d=\frac{1}{2}\sqrt{{\left(T_{D}C\right)}^2+4T_D\mathrm{CS}}---\left(1\right)$

In the above-mentioned formula (1), d represents the buried depth size of defective end, T _DThe time difference of expression lateral wave and defective end diffracted wave, S represents half of two probe acoustic emission dot spacings, and C represents the velocity of sound of ultrasound wave in material;

B, calculate pending A sweep signal X respectively according to formula (2) _nBoth sides participate in the time shift amount of each A sweep signal of SAFT computing,

$\mathrm{\&Delta;}{T}_i=(\sqrt{{(S-(n-i)\mathrm{\&Delta;S})}^2+{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}+\sqrt{{(S+(n-i)\mathrm{\&Delta;S})}^2+d^2}-2\sqrt{{\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+d^2})/C---\left(2\right)$

Δ T in the above-mentioned formula (2) _iExpression A sweep signal X _iThe time shift amount, d represents the buried depth size of defective end, S represents half of two probe acoustic emission dot spacings, Δ S represents the scanning step-length of popping one's head in right, C represents the velocity of sound of ultrasound wave in material;

C, according to formula (3) computing of suing for peace and average,

$X_n\left(T_n\right)=1/(N+1)\underset{i=n-N/2}{\overset{n+N/2}{\mathrm{\&Sigma;}}}{X}_i(T_i-\mathrm{\&Delta;}{T}_i)---\left(3\right)$

In the above-mentioned formula (3), X _n(T _n) be that the diffracted wave due in is T _nA sweep signal X _nThrough the return signal after the SAFT computing, N is for participating in the X of SAFT computing _nThe number of the A sweep signal of both sides, X _i(T _i-Δ T _i) for the diffracted wave due in be T _iA sweep signal X _iReturn signal after time shift;

D, output result of calculation are gathered all result of calculations, and the SAFT that finishes image rebuilds, and above-mentioned curve image is redeveloped into focusedimage.

Principle of work: as shown in Figure 1, the scan mode of ultrasonic TOFD comprises A, B, three kinds of modes of D, and wherein D scanning can be determined the Position Approximate of defective, and B scanning can obtain to comprise the image of the accurate information of defective.B, D scan image all are arranged in order to make up by a series of A sweep signals and form, and the A sweep signal can read acquisition from B, D scan image.The B scan image that the present invention obtained is the image that is contrasted by the light and shade that 256 grades of gray scales constitute as shown in Figure 6.Can Fig. 6 is abstract in image shown in Figure 3 when carrying out theoretical analysis.When waveform phase among Fig. 3 is dextrad, be higher than 128 corresponding to the gradation of image value among Fig. 6; When waveform phase among Fig. 3 is left-hand, be lower than 128 corresponding to the gradation of image value among Fig. 6.Because the decay of diffracted wave energy causes oscillating waveform to disturb temporal resolving power in signal and the image, therefore influenced the accurate positioning and quantitative of defective, so the present invention carries out linearization process to original image earlier.Linearization process is by image array being carried out computings such as subregion, peak value seizure, eliminate the redundant signals such as damped oscillation of sound wave, making as the lateral wave of Time Calculation benchmark and vertical temporal resolution of defective end diffracted wave to be improved; And then utilize the SAFT technology that the curve in the linearized graph picture is focused on, improve the lateral resolution of image.

The present invention has set up the mathematical model of rebuilding based on the ultrasonic TOFD method B scan image of SAFT according to the geometric relationship of forming process, probe and the defective locations of B scan image.Illustrate that with defective upper end diffracted wave image reconstruction process its principle is as follows: when transmitting probe 1 and receiving transducer 2 during in A-A ' shown in Figure 2, B-B ', C-C ' position, system generates the A sweep signal X among Fig. 3 respectively _N-1, X _nAnd X _N+1Among Fig. 3, when generating A sweep signal X _nThe time, the upper end of defective 4 two the probe (1 and 2) lines perpendicular bisector on, X _nFor the lateral attitude and the depth location of locating the defective upper end provides information accurately.According to A sweep signal X among Fig. 3 _nLateral wave and the due in of diffracted wave, can calculate the buried depth size d that tries to achieve defective upper end by following formula (1); At this moment, according to probe to the geometric relationship of defective upper end, can try to achieve T time of arrival that the upper end diffracted wave is removed probe delay according to formula (4).

$T=2\sqrt{S^2+{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}/C---\left(4\right)$

In the above-mentioned formula (4), T represents the time of arrival that the upper end diffracted wave is removed probe delay, and S represents half of two probe acoustic emission dot spacings, and d represents the buried depth size of defective upper end, and C represents the velocity of sound of ultrasound wave in material.

For departing from the A sweep signal X that position k scanning step place of two probe line perpendicular bisectors obtains _N+k, its upper end diffracted wave signal is removed T time of arrival of probe delay _{K Δ S}, can try to achieve by following formula (5),

$T_{\mathrm{k\&Delta;S}}=(\sqrt{{(S-\mathrm{k\&Delta;S})}^2+{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}+\sqrt{{(S+\mathrm{k\&Delta;S})}^2+d^2})/C---\left(5\right)$

In the above-mentioned formula (5), T _{K Δ S}The time of arrival that expression upper end diffracted wave signal is removed probe delay, S represents half of two probe acoustic emission dot spacings, and d represents the buried depth size of defective upper end, and Δ S represents the scanning step-length of popping one's head in right, and C represents the velocity of sound of ultrasound wave in material.

At this moment, T _{K Δ S}Δ T time delay of relative T _{K Δ S}Can try to achieve by formula (6),

$\mathrm{\&Delta;}{T}_{\mathrm{k\&Delta;S}}=T_{\mathrm{k\&Delta;S}}-T$

$=(\sqrt{{(S-\mathrm{k\&Delta;S})}^2+{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}+\sqrt{{(S+\mathrm{k\&Delta;S})}^2+d^2}-2\sqrt{S^2+d^2})/C---\left(6\right)$

The retardation that following formula (6) calculates gained is the corresponding time shift amount of pending A sweep signal its both sides A sweep signal when carrying out the SAFT computing.When carrying out the SAFT computing, the pending A sweep signal of hypothesis is that the defective upper end gets in the perpendicular bisector place of two probe lines earlier, the defective buried depth that calculates this A sweep signal correspondence then with and the retardation of both sides A sweep signal defect diffracted wave due in, carry out corresponding time shift according to formula (2).With each A sweep signal after the time shift by formula (3) sue for peace, average calculating operation, finish the SAFT computing of pending A sweep signal.Then next A sweep signal is carried out the SAFT computing.Whole A sweep signals after the S set AFT computing, the SAFT that finishes image rebuilds.Respectively to the A sweep signal X among Fig. 3 _N-lAnd X _nCarry out the time shift computing of SAFT, operation result synoptic diagram such as Fig. 4, shown in Figure 5.When carrying out SAFT computing 000, for the lateral wave signal that remains with the time reference effect with reduce calculated amount as far as possible, adopt the windowing computing.Window width is the number of the A sweep signal of participation summation operation; Window height upper edge is chosen between lateral wave and the defective upper end diffracted wave; The lower edge is chosen between the defective upper and lower end diffracted wave, shown in the empty frame among Fig. 4, Fig. 5.Computing this moment window width is N+1=5.For departing from the A sweep signal X that two probe line perpendicular bisector places obtain _N-1, because above-mentioned hypothesis is false, its SAFT time shift computing makes the defective diffracted wave of each A sweep signal in the computing window arrive in the different moment, as shown in Figure 4.By formula (3) do and, be averaged processing after, the amplitude of flaw indication can be very weak, even go to zero.For the A sweep signal X that obtains at two probe line perpendicular bisector places _n, because above-mentioned hypothesis is set up, its SAFT time shift computing makes the defective diffracted wave of each A sweep signal in the computing window arrive in the identical moment, as shown in Figure 5.By each a-signal work with after being averaged processing, the amplitude of flaw indication is very strong.In view of the above, can catch the accurate information of lateral attitude, defective upper end and depth location.Similar processing is done in the lower end of defective, and different is, is that valley is caught for the seizure of defective lower end diffracted wave due in; The height upper edge of computing window is selected between the defective upper and lower end diffracted wave; The lower edge is chosen between defective lower end diffracted wave and the bottorm echo.

Invention effect: the invention solves the accurate location and the quantitative problem of defective in sheet material and the construction joint, for structural intergrity, crack propagation situation and the predicting residual useful life of member provides accurate data.Can position quantitative measurment to open-type and inner defective of burying type.Undesired signal has obtained suppressing significantly in the image after L-SAFT of the present invention rebuilds, image resolution ratio is improved, and eliminated the noise of processing region, and make the location of defective end more accurate, reduced the personal error that naked-eye observation brings largely.Adopt vertical temporal resolution of the image of L-SAFT processing of the present invention can reach 0.01 μ s, lateral separation is differentiated the minimum scanning step size that can reach setting.Simultaneously, the present invention has very high automatization level height, and is faster to the location of defective.

Description of drawings

Fig. 1 is B, the D scan mode synoptic diagram of the ultrasonic transit time method of the present invention, and 1 is that transmitting probe, 2 is that receiving transducer, 3 is sheet material and construction joint (being detected zone) among the figure; Fig. 2 is the B scanning motion synoptic diagram of the ultrasonic transit time method of the present invention, and 4 is defective in sheet material and the construction joint 3 among the figure; Fig. 3 is that the B scan image of Fig. 2 generates synoptic diagram, and a represents the side direction ultrasound wave among the figure, and b represents the upper end diffracted wave, and c represents the lower end diffracted wave, and d represents Bottom echo; Fig. 4 is to the A sweep signal X among Fig. 3 _N-1Carry out the time shift result schematic diagram of SAFT computing; Fig. 5 is to the A sweep signal X among Fig. 3 _nCarry out the time shift result schematic diagram of SAFT computing; Fig. 6 is an original B scan image in the specific embodiment of the invention one; Fig. 7 is the B scan image that carries out in the specific embodiment of the invention one after the linearization process; Fig. 8 is the B scan image that carries out in the specific embodiment of the invention one after L-SAFT handles; Fig. 9 is the B scan image of embodiment two; Figure 10 is the B scan image that carries out in the embodiment two after the linearization process; Figure 11 is the B scan image that carries out in the embodiment two after L-SAFT handles; Figure 12 is the process flow diagram that linearization process is carried out in the parabolic shape zone of the second step original B scan image defective lower end correspondence in the embodiment one; Figure 13 carries out the process flow diagram that image L-SAFT rebuilds to corresponding curve processing district, defective upper end during the 3rd of embodiment one goes on foot.

Embodiment

Embodiment one: referring to Fig. 1 to Fig. 8, it is that example illustrates detection method of the present invention that this embodiment is buried the type defective with inside.This embodiment is to be that the inside that processes on the aluminum alloy plate materials of 20mm is buried the type defective and detected to thickness, and it is as follows to detect step:

The first step, at first the zone to be detected with sheet material and construction joint thereof places between transmitting probe and the receiving transducer, earlier D scanning is carried out to determine the Position Approximate of defective in detected zone; Set the scanning step-length of probe, cross the B scanning of defective top, detection system generates pending original B scan image with the detection data of receiving transducer, as shown in Figure 6; Wherein, along the horizontal zone on the corresponding B scan image of the lateral wave top that the more shallow degree of depth under sheet material and the construction joint surface is propagated; Article one parabolic shape zone of above-mentioned horizontal zone below in the upper end diffracted wave correspondence image of defective; The second parabolic shape zone of above-mentioned horizontal zone below in the lower end diffracted wave correspondence image of defective; The horizontal zone of the bottorm echo correspondence image bottom of sheet material and construction joint;

Second step, original B scan image that the first step is obtained carry out linearization process, and the lateral wave horizontal zone of original B scan image is treated to straight line; The parabolic shape zone of the upper and lower end of defective diffracted wave is treated to curve respectively, as shown in Figure 7;

Curve after the 3rd step, the selection linearization process is a process object, open the gray-scale value character matrix of process object, and the maximum and the minimum value of reading every row, wherein, when being handled, the corresponding curvilinear region in defective upper end reads the maximal value of every row pixel gray-scale value; Read the minimum value of every row pixel gray-scale value when the curvilinear region of defective lower end correspondence is handled, when whenever reading once described maximal value or minimum value, carry out following computing:

A, read maximum or the pairing time location of minimum gradation value pixel, calculate the buried depth size of defective end according to formula (1),

$d=\frac{1}{2}\sqrt{{\left(T_{D}C\right)}^2+4T_D\mathrm{CS}}---\left(1\right)$

In the above-mentioned formula (1), d represents the buried depth size of defective end, T _DThe time difference of expression lateral wave and defective end diffracted wave, S represents half of two probe acoustic emission dot spacings, and C represents the velocity of sound of ultrasound wave in material;

B, calculate pending A sweep signal X respectively according to formula (2) _nBoth sides participate in the time shift amount of the A sweep signal of SAFT computing,

$\mathrm{\&Delta;}{T}_i=(\sqrt{{(S-(n-i)\mathrm{\&Delta;S})}^2+{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">d</figure-callout>}}^2}+\sqrt{{(S+(n-i)\mathrm{\&Delta;S})}^2+d^2}-2\sqrt{{\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="C20051012739300121.png,C20051012739300122.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+d^2})/C---\left(2\right)$

Δ T in the above-mentioned formula (2) _iExpression A sweep signal X _iThe time shift amount, d represents the buried depth size of defective end, S represents half of two probe acoustic emission dot spacings, Δ S represents the scanning step-length of popping one's head in right, C represents the velocity of sound of ultrasound wave in material;

C, according to formula (3) computing of suing for peace and average,

$X_n\left(T_n\right)=1/(N+1)\underset{i=n-N/2}{\overset{n+N/2}{\mathrm{\&Sigma;}}}{X}_i(T_i-\mathrm{\&Delta;}{T}_i)---\left(3\right)$

In the above-mentioned formula (3), X _n(T _n) be that the diffracted wave due in is T _nA sweep signal X _nThrough the return signal after the SAFT processing, N is for participating in the X of SAFT computing _nThe number of the A sweep signal of both sides, X _i(T _i-Δ T _i) for the diffracted wave due in be T _iA sweep signal X _iReturn signal after time shift;

D, output result of calculation are gathered all result of calculations, and the SAFT that finishes image rebuilds, and above-mentioned curve image is redeveloped into focusedimage, as shown in Figure 8.

As Fig. 3 and shown in Figure 12, the process of in second step original B scan image being carried out linearization process is: 1) at first 256 grades of gray level images are carried out subregion, be divided into lateral wave treatment region, upper end diffracted wave treatment region, lower end diffracted wave treatment region and four processing regions of Bottom echo treatment region; 2) character matrix to lateral wave treatment region and lower end diffracted wave treatment region carries out the valley seizure by row, character matrix to upper end diffracted wave treatment region and Bottom echo treatment region carries out the peak value seizure by row, keep peak value and valley corresponding gray, other gray-scale value in the processing region is composed into 128.

Be that the example key diagram is as linearizing detailed process, as shown in figure 12 with original image curve lower end diffracted wave below.001 step, to select lower end diffracted wave treatment region be process object, and open (a * b) of original image pixels character matrix X in the treatment region; 002 step, the definition first loop variable j=1,0≤j≤b; 003 step, the definition second loop variable i=1,0≤i≤a; 004 step, judge described character matrix X (pixel X (i among a * b), whether gray-scale value j) is the minimum value of column, if pixel X is (i, j) gray-scale value is the minimum value of column, then carries out 005 step reservation pixel X (i, gray-scale value j), if pixel X is (i, j) gray-scale value is not the minimum value of column, then carry out 006 the step with pixel X (i, gray-scale value j) composes into 128; 007 the step, execute 005 the step or 006 the step after, the second loop variable i is added 1; 008 step, whether judge the second loop variable i,, then repeat 004 and went on foot for 008 step, go on foot the first loop variable j is added 1 if the second loop variable i, then carries out 009 greater than a if the second loop variable i is not more than a greater than a; 010 step, whether judge the first loop variable j,, then repeat 003 and went on foot for 010 step, if the first loop variable j, then finishes valley seizure greater than b if the first loop variable j is not more than b greater than b.When catching,, " minimum value " in 004 handle as long as being changed into " maximal value " for peak value.

The process of reconstruction of image mainly is divided into the reconstruction in the curve processing district of defective upper end and defective lower end correspondence in the 3rd step.As shown in figure 13, the detailed process that corresponding curve processing district, described defective upper end rebuilds is carried out according to the following steps: 011 step, open the image pixel numeral matrix Y (a * b) in corresponding curve processing district, defective upper end, import following correlation parameter: half of two probe acoustic emission dot spacings, the velocity of sound of compressional wave in material, the scanning step-length of transmitting probe and receiving transducer, the upper and lower parameter of processing region along time corresponding value, decision processing region width; 012 step, definition the 3rd loop variable w=1,0≤w≤b; 013 step, definition the 4th loop variable v=1,0≤v≤a; 014 step, judge described character matrix Y (pixel Y (v among a * b), whether gray-scale value w) is the maximal value of column, if this pixel Y is (v, w) be maximal value, then carry out 015 this pixel of step record Y (v, w), if this pixel Y (v, w) be not maximal value, then carry out 020 the step the 4th loop variable v is added 1; 016 goes on foot, executes 015 step back calculates the buried depth size of defective upper end according to described formula (1); 017 step, according to the time shift amount of each A sweep signal in described described formula (2) the calculation operations window; 018 step, according to described formula (3) computing of suing for peace and average, and carry out 019 step output result, carried out for 020 step then the 4th loop variable v added 1; 021 step, execute 020 step back and whether judges the 4th loop variable v,, then carried out for 022 step the 3rd loop variable w is added 1,, then repeat 014 and go on foot 021 and go on foot if the 4th loop variable v is not greater than a if the 4th loop variable v is greater than a greater than a; 023 the step, execute 022 the step after, whether judge the 3rd loop variable w greater than b, if the 3rd loop variable w is not more than b, then repeat 013 go on foot 023 the step, if the 3rd loop variable w then finishes computing greater than b.To diffracted wave curve processing district, lower end carry out SAFT when rebuilding, only " maximal value " in 014 step need be changed into " minimum value " and handle and get final product.

Adopt this embodiment as shown in the table to the measurement result of the type that buries defective with upper and lower end:

Embodiment two: referring to Fig. 9 to Figure 11, adopting the method for embodiment one is that the type that the buries defective of processing on the aluminum alloy plate materials of 20mm detects to another thickness, and the result of detection is as shown in the table:

Obtaining all of original image finished in the Ultrasonic C-Scan detection system that U.S. physical acoustics company (PAC) produces in the above-mentioned embodiment one and two.The measurement result of original image is 3 mean values of measuring; Measurement result in the reconstructed image is the one-shot measurement gained.According in above-mentioned embodiment one and the two shown testing result data as can be seen, when the systematic sampling frequency is 100MHz, in the image after reconstruction, the measuring error of defective end depth of burial is no more than 0.2mm; When scanning step was 0.2mm, the lateral separation measuring error of defective end was no more than 0.2mm; Measuring error to the defective angle of inclination is no more than 5 °.

Claims (2)

1,, it is characterized in that described detection method carries out according to the following steps based on the ultrasonic transit time detection method of synthetic aperture focusing:

The first step, at first the zone to be detected with sheet material and construction joint thereof places between transmitting probe and the receiving transducer, earlier D scanning is carried out to determine the Position Approximate of defective in detected zone; Set the scanning step-length of probe, cross the B scanning of defective top, detection system generates pending original B scan image with the detection data of receiving transducer; Wherein, along the horizontal zone on the corresponding B scan image of the lateral wave top that the more shallow degree of depth under the surface of sheet material and construction joint is propagated; Article one parabolic shape zone of above-mentioned horizontal zone below in the upper end diffracted wave correspondence image of defective; The second parabolic shape zone of above-mentioned horizontal zone below in the lower end diffracted wave correspondence image of defective; The horizontal zone of the bottorm echo correspondence image bottom of sheet material and construction joint;

Second step, original B scan image that the first step is obtained carry out linearization process, and the lateral wave horizontal zone of original B scan image is treated to straight line; The parabolic shape zone of the upper and lower end of defective diffracted wave is treated to curve respectively;

Curve after the 3rd step, the selection linearization process is a process object, opens the gray-scale value character matrix of process object, and reads the maximum and the minimum value of every row; Wherein, when being handled, the corresponding curve regions in defective upper end reads the maximal value of every row pixel gray-scale value, when being handled, the curve regions of defective lower end correspondence reads the minimum value of every row pixel gray-scale value, when whenever reading once described maximal value or minimum value, carry out following computing:

A, read maximum or the pairing time location of minimum gradation value pixel, calculate the buried depth size of defective end according to formula (1),

$d=\frac{1}{2}\sqrt{{\left(T_{D}C\right)}^2+4T_D\mathrm{CS}}---\left(1\right)$

In the above-mentioned formula (1), d represents the buried depth size of defective end, T _DThe time difference of expression lateral wave and defective end diffracted wave, S represents half of two probe acoustic emission dot spacings, and C represents the velocity of sound of ultrasound wave in material;

B, calculate pending A sweep signal X respectively according to formula (2) _nBoth sides participate in the time shift amount of each A sweep signal of synthetic aperture focusing computing,

$\mathrm{\&Delta;}{T}_i=(\sqrt{{(S-(n-i)\mathrm{\&Delta;S})}^2+d^2}+\sqrt{{(S+(n-i)\mathrm{\&Delta;S})}^2+d^2}-2\sqrt{S^2+d^2})/C---\left(2\right)$

Δ T in the above-mentioned formula (2) _iExpression A sweep signal X _iThe time shift amount, d represents the buried depth size of defective end, S represents half of two probe acoustic emission dot spacings, Δ S represents the scanning step-length of popping one's head in right, C represents the velocity of sound of ultrasound wave in material;

C, according to following formula (3) computing of suing for peace and average,

$X_n\left(T_n\right)=1/(N+1)\underset{i=n-N/2}{\overset{n+N/2}{\mathrm{\&Sigma;}}}{X}_i(T_i-\mathrm{\&Delta;}{T}_i)---\left(3\right)$

In the above-mentioned formula (3), X _n(T _n) be that the diffracted wave due in is T _nA sweep signal X _nThrough the return signal after the synthetic aperture focusing processing, N is for participating in the X of synthetic aperture focusing computing _nThe number of the A sweep signal of both sides, X _i(T _i-Δ T _i) for the diffracted wave due in be T _iA sweep signal X _iReturn signal after time shift;

D, output result of calculation are gathered all result of calculations, finish the synthetic aperture focusing of image and rebuild, and above-mentioned curve image is redeveloped into focusedimage.

2, the ultrasonic transit time detection method based on synthetic aperture focusing according to claim 1, it is characterized in that the process of in second step original B scan image being carried out linearization process is: 1) at first 256 grades of gray level images are carried out subregion, be divided into lateral wave treatment region, upper end diffracted wave treatment region, lower end diffracted wave treatment region and four processing regions of Bottom echo treatment region; 2) character matrix to lateral wave treatment region and lower end diffracted wave treatment region carries out the valley seizure by row; Character matrix to upper end diffracted wave treatment region and Bottom echo treatment region carries out the peak value seizure by row, keeps peak value and valley corresponding gray, and other gray-scale value in the processing region is composed into 128.

Images