JPH03257361A - Ultrasonic flaw detection apparatus - Google Patents

Abstract

PURPOSE:To extract the flaw echo buried in a scattering echo and to enhance the detection accuracy of the flaw contained in an object to be inspected by dividing the spectrum of each time window in a detection signal by the spectrum of a transmission wave and performing the limitation of a band. CONSTITUTION:At first, with respect to the detection signal (b) of the ultrasonic pulse to a standard test piece 13, for example, a time window of predetermined time width of 1 - 10 mus is set to a transmission wave region by the control of a time window control part 20 to calculate the transmission wave spectrum within the time window. Next, the spectrum of the waveform within the time window is calculated with respect to the detection signal (b) from the object 11 to be inspected mounted on an ultrasonic probe 12 while the time window is successively moved from a measuring start point to a completion point and divided by the transmission wave spectrum to be normalized and the limitation of a band is performed on the basis of the width of the frequency component contained in a transmission wave and a flaw is displayed on a display part 28 as a numerical value. By this method, a scattering echo always having a constant spectrum and a flaw echo are clearly discriminated to be capable of being displayed on the display part 28 and, therefore, the flaw echo buried in the scattering echo can be certainly extracted.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an ultrasonic flaw detection device that uses ultrasonic waves to detect defects in a test object, and is particularly suitable for measuring a test object with large crystal grains. The present invention relates to an ultrasonic flaw detection device that efficiently removes the influence of scattered echoes from grain boundaries that occur in crystal grain boundaries.

[Prior Art] An ultrasonic flaw detection apparatus that uses ultrasonic waves to detect defects existing inside or on the surface of a solid object is configured as shown in FIG. 5, for example.

An ultrasonic probe 2 formed of, for example, a piezoelectric element is attached to a subject 1, and a pulse generating circuit 3 and a broadband amplifier 4 are connected to the ultrasonic probe 2.

Then, when a pulse signal is transmitted from the pulse generation circuit 3 to the ultrasound probe 2 in synchronization with the trigger signal output from the trigger signal generator 5, an ultrasound pulse is applied from the ultrasound probe 2 to the subject. be done. Then, this ultrasonic pulse propagates inside the subject 1, is reflected from the bottom surface, and returns to the ultrasonic probe 2 again as an echo. When the ultrasonic pulse encounters a defect while propagating within the object 1, it is reflected by the defect and returns to the ultrasonic probe 2. Ultrasonic probe 2
converts the ultrasonic pulse reflected by the bottom surface or defect and incident again into an electrical signal and outputs it as a detection signal a. The detection signal a output from the ultrasonic probe 2 is transmitted to the broadband amplifier 4.
After being amplified, the signal is converted into a digital value by an A/D converter 6 and input to a waveform analyzer 7. The waveform analyzer 7 analyzes the detection signal a to determine whether or not a defect exists in the object 1 to be inspected, and if so, calculates the scale of the defect.

However, when the object 1 is a metal with strong anisotropy and large crystal grains, the detection signal a contains a large amount of scattered echoes from the crystal grain boundaries, as shown in FIG. Therefore, even if a small defect exists within the object 1, the defect echo due to this defect is completely buried in the scattered echoes as shown in the figure. Therefore, defect detection accuracy is reduced.

In order to solve this problem, as shown in FIG. 7(a), a time window 8 having a predetermined time width T is provided in the detection signal
is provided, and spectrum calculation is performed on the waveform existing within the time window 8 to obtain a spectrum 9 with the frequency f as the horizontal axis shown in FIG. 7(b). Then, the time t of the time window 8 is sequentially moved, and the spectrum 9 of each time window 8 at each time t is calculated. Then, when each time window 8 is moved, the shape of each spectrum 9 is sequentially digitized and displayed with time t as the horizontal axis. As shown in FIG. 7(C), the defect A peak occurs at the time position where . Therefore, the presence of defects can be detected.

This takes advantage of the fact that the frequency characteristics of scattered echoes are almost constant regardless of the time position (power spectrum shape evaluation method based on fractal theory; application to randomness estimation of point scatterer spacing; electronic information Communication Society, Ultrasonic Research Group, Research Presentation Proceedings: 1989.5.28
').

[Problems to be Solved by the Invention] However, even in the ultrasonic flaw detection apparatus which extracts defective echoes from scattered echoes by the procedure shown in FIG. 7, the following problems still exist.

That is, the spectrum 9 of the waveform within the time window 8 set for the detection signal a in FIG. 7 is naturally greatly influenced by the spectrum of the transmitted wave, which is regulated by the waveform of the ultrasound pulse applied to the subject 1. As a result, the spectrum of the transmitted wave is also included when quantifying the shape of the spectrum 9 within the time window 8. Therefore, FIG. 7(e) includes the influence of the spectral shape of the transmitted wave, and the detection accuracy decreases.

Furthermore, although the ultrasonic pulse applied to the subject 1 through the ultrasonic probe 2 generally includes all frequency components, there is a certain band limit. Therefore, when an ultrasonic pulse with a certain band limit is applied, components of the frequency components of the detection signal a due to echoes outside the band of the applied ultrasonic pulse become noise. Therefore, since this noise component 7j is present, the accuracy of analysis of the detection signal a is reduced.

The present invention was made in view of these circumstances, and
By dividing the spectrum of each time window in the detection signal by the spectrum of the transmitted wave and limiting the band, defective echoes buried in scattered echoes can be extracted with high precision.
It is an object of the present invention to provide an ultrasonic flaw detection device that can significantly improve the accuracy of detecting defects present in a test object.

[Means for Solving the Problems] In order to solve the above problems, the present invention applies ultrasonic pulses to a subject and detects detection signals by echoes received from the subject via an ultrasound probe. Each time window when the time window is moved by setting a time window with a predetermined time width for the waveform and performing spectrum calculation on the waveform existing within the time window among the detected signal waveforms. In an ultrasonic flaw detection device that detects defects in a test object from changes in the spectrum of A division unit that divides the spectrum by the transmission spectrum of and a display section that displays the converted values in chronological order.

[Operation] With an ultrasonic flaw detection device configured in this way, the spectrum of each waveform in each time window set for the detection signal obtained from the object is divided by the spectrum of the transmitted wave and normalized. be done. Therefore, the influence of the spectrum of the transmitted wave can be removed from the normalized spectrum of each time window. Therefore, defective echoes can be detected from changes in the spectrum of each time window as in the conventional method. Further, by band-limiting each normalized spectrum, noise components outside the band are removed.

[Example] An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of an ultrasonic flaw detection apparatus according to an embodiment. In the figure, reference numeral 11 denotes an object made of a metal material such as stainless steel with strong anisotropy and large crystal grains, and an ultrasonic probe 12 made of, for example, a piezoelectric element is attached to this object 11. This ultrasonic flaw detection device is equipped with a standard test piece made of ordinary steel with no defects on its surface or inside. It is attached to the tentacle 12.

Further, the trigger signal generation circuit] 4 sends a trigger signal for controlling the timing of starting measurement to the pulse generation circuit 15. The pulse generating circuit 15 applies a pulse signal to the ultrasonic probe 12 in synchronization with the signal input. Then, the ultrasonic probe 12 applies an ultrasonic pulse having a predetermined frequency band to the subject 11. This ultrasonic pulse
It propagates inside, is reflected at the bottom surface, and returns to the ultrasound probe 12 again as an echo. When the ultrasonic pulse encounters a defect while propagating within the object 11, it is reflected by the defect and returns to the ultrasonic probe 12. The ultrasonic probe 12 converts the ultrasonic pulse reflected by the bottom surface or a defect and incident again into an electric signal and outputs it as a detection signal. Note that this ultrasonic probe 12 can transmit and receive ultrasonic waves of 2 to IOM) lz. The detection signal output from the ultrasonic probe 12 is amplified by the broadband amplifier 16 and then sent to the A/D converter 17.
The signal is converted into a digital value at a sampling period of, for example, 40 ns, and then stored in the received waveform memory 18. The detection signal once stored in the received waveform memory 8 is input to the next waveform analyzer 19.

This waveform analysis device 19 is composed of, for example, a micro combinator, and as shown in the figure, a time window control section 209
Time window waveform succession part 21. Spectrum calculation unit 22. Transmission waveform memory 23. Dividing unit 24° Bandwidth control unit 25. Waveform digitization processing section 26. It has functions such as a calculation result memory 27.

Then, a transmission waveform memory 23 and a calculation result memory 27
is formed of a memory element such as a RAM, and the time window control unit 201
Time window waveform succession part 21. Spectrum calculation unit 22. Nasal removal part 24. The bandwidth control unit 25 is composed of each data processing procedure in the control program.

Further, a display unit 28 for displaying waveform analysis results is connected to one of the waveform analysis devices.

This ultrasonic flaw detection apparatus is configured to perform defect detection processing on the object 11 according to the procedure shown in the flowchart of FIG.

When the flowchart starts, first, in S (step) 1, the standard test piece 13 is attached to the ultrasonic probe 12. Next, measurements are performed on this standard test piece 13.

That is, the trigger generation circuit 14 is activated and a trigger signal is sent to the pulse generation circuit 15. Then, an ultrasonic pulse is applied from the ultrasonic probe 12 to the standard test piece 12.

Then, the detection signal outputted from this standard test piece 12 is amplified by a broadband amplifier 16 and converted into a digital value by an A/D converter 17. Then, the detection signal of the standard test piece 13 converted into a digital value is temporarily stored in the received waveform memory 18.

Next, in S2, the time window control unit 20 is activated, and the time window 29 shown in FIG. Set to the area of the transmitted wave that appears in the area. Then, the time window waveform reading unit 21 is activated and a time window 2 is set in the region of the transmitted wave.
Read out the transmission waveform existing in 9. In S3, spectrum calculation is performed for the read transmission waveform. Then, the calculated transmission wave spectrum S(ω) is stored in the transmission waveform memory 23.
Store it in

Now that the measurement preparations have been completed, proceed to S4 to prepare the standard test piece 1.
3 is removed, and the subject 11 is attached to the ultrasound probe 12. Then, it is measured under the same conditions as the standard test piece 13. Then, the digitally converted detection signal of the subject 11 is stored in the received waveform memory 18.

Then, in S5, the time window control unit 20 is activated and the time window 29 is set to the measurement starting point t as shown in FIG. Set to Next, in S6, the waveform within the set time window 29 of the detection signal of the subject 11 is read out by the time window waveform reading unit 21, and spectrum calculation processing is performed on the read waveform D (t). is executed to find the vector D(ω). Next, in S7, the spectrum D within the calculated time window 29
(ω) is divided by the transmission wave spectrum S(ω) read from the transmission waveform memory 23, and the normalized spectrum [D (ω
)/S(ω)]. Next, band limitation is performed on the normalized spectrum [D(ω)/S(ω)] as shown in FIG. Note that the frequency width of this band limitation is the width of frequency components included in the transmitted wave. For example, as shown in the figure, it is assumed that the value includes a spectrum waveform of 10% or more of the maximum value P of the transmission spectrum.

At S8, the band-limited normalized spectrum [D (ω
)/S(ω)], for example, is digitized into one data. The procedure for quantifying the waveform of this normalized spectrum [D(ω)/S(ω)] will be explained using FIGS. 3(b) and 3(C).

That is, when the variable is i, the function of the waveform to be quantified is A(i) (-normalized spectrum), and when the number of waveform divisions is N, the unit variable is defined as d2. Next, the calculation variable k is defined as shown in the following equation.

−1 k=log[(Σl A(id+d) −A(1d)
l )/(N/d)] Next, each calculation variable k (d) corresponding to each unit variable d is calculated using the above formula. As shown in FIG. 3(C), each calculated calculation variable k (d) is expressed as
Logarithm 1o of the unit variable d (-1, 2, ..., N)
Graph the gd value on the horizontal axis. In addition, Fig. 3(a)
In the graph at the bottom of the graph, the horizontal axis is displayed not as logd but as a straight line.

Next, digitized data ρ of the corresponding waveform is calculated.

That is, in the characteristics shown in Figure 3(C), the correlation coefficient is 0.
．． A straight line 30 shown by a dotted line is determined within the range of the unit variable d that satisfies 9. Then, the intersecting angle θ between the straight line 30 and the horizontal axis is determined, and the final numerical data D is calculated from this intersecting angle θ using the following formula.

ρ-2-〇 (1≦p≦2) The digitized data ρ thus calculated is stored in the empty area of the calculation result memory 27 together with the information of the time t indicated by the current time window 29.

3 Since the analysis process for the waveform within the time window 29 at one time position t has been completed, in S9 the time window 29 is moved by Δt in the direction in which the time t increases. 31.0
When it is confirmed that the time window 29 after the movement has not reached the measurement end point tF, the process returns to S6 and the analysis process for the waveform within the corresponding time window 29 is started.

When the two time windows reach the measurement end point t6 in SIO, the entire time domain (t o =
Since the waveform analysis process at t' E ) has been completed,
In S11, the normalized spectrum [D
The calculation result (numerized data ρ) of (ω)/S(ω) is read out and displayed as time t on the horizontal axis as shown in the top column of FIG. 3(a).

Note that the width of the time window 29, the amount of movement, and the score for spectrum calculation are set depending on the compositional situation of the subject 11.

According to the ultrasonic flaw detection apparatus configured in this way, as shown in FIG. Several time windows 29 are provided, a waveform vector in each time window 29 is obtained, and each spectrum is normalized by the spectrum of the transmitted wave. The display unit 26 displays changes over time in the numerical values of the normalized spectrum.

Therefore, the scattered echoes and defective echoes, which always have a constant spectrum, can be clearly distinguished and displayed, so that the defective echoes buried in the scattered echoes can be reliably extracted.

Furthermore, when converting the waveform into numerical values, the influence of the spectrum of the transmitted wave can be removed.

Furthermore, since the spectrum of each time window normalized by the spectrum of the transmitted wave is band-limited, for example, by the bandwidth of the spectrum of the transmitted wave, noise outside the required bandwidth can be removed. Therefore, measurement accuracy can be further improved.

FIG. 4 is a diagram showing a comparison between the calculation results obtained with the conventional apparatus and the calculation results obtained with the embodiment apparatus using the same subject 11. As can be understood from this figure, defects can be detected more accurately and quantitatively with the calculation results of the apparatus of the embodiment than with the conventional apparatus.

[Effects of the Invention] As explained above, according to the ultrasonic flaw detection device of the present invention,
By dividing the spectrum of each time window in the detection signal by the spectrum of the transmitted wave and limiting the band, defective echoes buried in scattered echoes can be extracted with high accuracy. Therefore, the accuracy of detecting defects present in the object can be greatly improved.

[Brief explanation of drawings]

1 to 4 show an ultrasonic flaw detection device according to an embodiment of the present invention, in which FIG. 1 is a block diagram showing a schematic configuration, FIG. 2 is a flowchart showing the operation, and FIG. a) is a diagram of each signal waveform showing the operation, FIGS. 3(b) and 3(C) are diagrams for explaining the procedure for converting waveforms into numerical values, and FIG. 4 is a diagram showing calculation results in comparison with a conventional device. FIG. 5 is a block diagram showing a conventional ultrasonic flaw detection device, FIG. 6 is a waveform diagram of a general detection signal, and FIG. 7 is a waveform diagram of each signal showing the operation of the conventional device. 11... Subject, 12... Ultrasonic probe, 13...
・Standard test piece, 15...Pulse generation circuit, 17...A
/D converter, 18... Received waveform memory, 19... Waveform analyzer, 20... Time window control section, 21... Time window waveform reading section, 22... Spectrum calculation section, 23.・
- Transmission waveform memory, 24...Dividing unit, 25...Band limiting unit, 26...Waveform digitization processing unit, 27...Computation result memory, 28...Display unit, two... time window.

Claims (1)

[Claims] Applying ultrasonic pulses to a subject and setting a time window with a predetermined time width for the waveform of a detection signal based on an echo received from the subject via an ultrasound probe. Then, spectral calculation is performed on the waveform existing within the time window among the detection signal waveforms, and from the change in the spectrum of each time window when the time window is moved, it is possible to calculate the spectrum that exists in the subject from the change in the spectrum of each time window when the time window is moved. An ultrasonic flaw detection device for detecting defects includes: a transmission waveform memory that stores a spectrum of a transmission wave included in a detection signal waveform as a transmission spectrum; and a division unit that divides the spectrum of each time window by the transmission spectrum of the transmission waveform memory. , a band limiter that limits the band of the spectrum of each time window divided by this divider, and a value that quantifies the shape of the spectrum of each time window band-limited by this band limiter in time series. An ultrasonic flaw detection device equipped with a display section.