JP4135512B2 - Ultrasonic signal processing method and ultrasonic measurement apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an ultrasonic signal processing method, and more particularly, to a method for performing flaw detection and thickness measurement on a specimen having a lot of noise due to scattering of ultrasonic waves such as cast material, refractory bricks, and concrete.
[0002]
[Prior art]
Ultrasonic waves (sometimes referred to as elastic waves) are widely used for flaw detection and thickness measurement of specimens because of their good permeability into solids. Usually, the pulse reflection method is generally used for such flaw detection and thickness measurement. In this pulse reflection method, an ultrasonic pulse is incident from the surface of the subject, ultrasonic waves reflected from the defect or the back surface (hereinafter referred to as signal echo) are received, and the flaw detection and thickness are determined based on the propagation time of the signal echo. Is a method of measuring.
[0003]
By the way, when the pulse reflection method is used for specimens such as cast materials, refractory bricks, concrete, etc., ultrasonic waves are scattered by crystal grains, fine voids, aggregates, etc. inside the specimen, and signal echoes are in these tissues. There is a problem that the S / N (S / N ratio) deteriorates because it is hidden by the scattered echo (hereinafter referred to as scattering noise).
[0004]
In order to avoid this, it is effective to lower the frequency of the ultrasonic wave. For example, when steel is used as a target, the frequency is usually about 1 MHz to 50 MHz, but generally 1 MHz for these specimens. Hereinafter, in some cases, the frequency is lowered to about 50 KHz.
[0005]
However, when the frequency is lowered to this level, the directivity angle of the ultrasonic beam becomes very wide. As a result, the signal echo is weakened, and the scattered noise is relatively large with respect to the signal echo, or the tailing of the transmission pulse (which is also a kind of noise) is large in the case of one search. After all, S / N is not improved so much.
[0006]
In order to eliminate the tailing of the transmission pulse of this one method, a two-probe method in which a transmission probe and a reception probe are separately used at low frequencies of 1 MHz or less is often used. However, when the two-probe method is used, the surface wave is directly mixed into the receiving probe and becomes noise. Again, the lower the frequency and the wider the pointing angle, the larger the signal echo becomes.
[0007]
Thus, there is a limit to improving the S / N against scattering noise even if the frequency is simply lowered. Therefore, the following techniques have been proposed as a method for improving S / N in ultrasonic measurement using a highly scattering material.
[0008]
In one technique, when measuring the thickness of a refractory or the like, the reflected wave is received by changing the interval between the position where the elastic wave is transmitted and the position where the reflected wave is received, and the reflected wave is added and averaged. Thus, a decrease in S / N is prevented (see, for example, Patent Document 1).
[0009]
In another technique, when a concrete internal defect is inspected by ultrasonic waves, the probe is scanned at a plurality of locations, and the received waveforms of the plurality of scanning lines are synthesized (for example, see Patent Document 2). ).
[0010]
In this way, signal echoes from the void or back surface of the subject appear in substantially the same beam path even when the probe is moved, and thus are amplified by addition. On the other hand, since the scattered noise is an interference of echoes reflected from various positions, the waveform changes randomly when the probe is moved. Therefore, since the scattered noise is not amplified by addition, S / N can be improved.
[0011]
[Patent Document 1]
Japanese Patent Laid-Open No. 9-61144
[0012]
[Patent Document 2]
Japanese Patent Laid-Open No. 10-288608
[0013]
[Problems to be solved by the invention]
However, in the method of performing averaging, a great number of averages are required to improve the S / N. For example, in the case of concrete, it is said that there is no effect unless the averaging is performed at least 500 times while moving the probe within a range of 200 mm × 200 mm (inspection technique 1999 8 Monthly issue P54-60 “Detecting the inside of concrete with ultrasonic waves (4)”).
[0014]
Further, when using a low frequency, it is desirable to use a probe having a large area so that a narrow directivity angle can be obtained in order to improve S / N. Then, when these methods are used for the measurement of the remaining thickness of the refractory described in Patent Document 1, it is necessary to provide a large opening in the iron skin of a furnace or the like because a probe having a large area is used. However, in practice, it is difficult to provide a large hole in the iron skin of the furnace from the viewpoint of mechanical strength of the furnace or cooling. Therefore, there is a limit to using a large area probe.
[0015]
In addition, even in the case of concrete that has no problem such as iron-hole opening described in Patent Document 2, it takes a very long time to perform an average of 500 times or more while moving the probe. There is a problem that it takes.
[0016]
The present invention has been made to solve the above-described problems. In ultrasonic measurement of a highly scattering material, a high S / N ratio can be obtained with a small number of data without using a probe having a large area. It is an object of the present invention to provide an ultrasonic signal processing method and an ultrasonic measurement device that are obtained.
[0017]
[Means for Solving the Problems]
  In order to solve the above problems, an ultrasonic signal processing method according to the first aspect of the present invention includes:An ultrasonic signal processing method of an ultrasonic measurement apparatus that performs flaw detection or thickness measurement of an object using ultrasonic waves, and a plurality of time-series signals for each position measured by changing the position of the probe relative to the object When the polarity of at least one signal value among the signal values at the same time of the plurality of time-series signals is different, the signal value at that time is set to 0, otherwise, the plurality of time-series signals For a plurality of signal values at the same point in time, the signal value having the maximum absolute value, the average value of the signal values from the largest absolute value to the predetermined number, the signal value having the second largest absolute value, and all the signal values A new time-series signal is obtained using any one of the average values as the signal value of the time-series signal at that time, and the specimen is subjected to flaw detection or thickness measurement based on the new time-series signal. .
[0018]
  In addition, an ultrasonic signal processing method according to another aspect of the present invention includes:An ultrasonic signal processing method of an ultrasonic measurement apparatus that performs flaw detection or thickness measurement of an object using ultrasonic waves, and a plurality of time-series signals for each position measured by changing the position of the probe relative to the object The signal values at the same time of the plurality of time series signals are classified into two positive and negative polarities, and when the number of signals belonging to a small number of polarity is a predetermined number or more, the signal value at that time is set to 0. In other cases, for a plurality of signal values belonging to many polarities, the signal value having the maximum absolute value, the average value of the signal values from the larger absolute value to the predetermined number, and the absolute value being the second largest A new time series signal is obtained in which any one of the signal value and the average value of all the signal values is used as the signal value of the time series signal at that time, and based on this new time series signal, Detect specimens or measure thickness.
[0019]
  In addition, an ultrasonic signal processing method according to another aspect of the present invention includes:In the ultrasonic processing method described above, the number of a plurality of time-series signals for each measured position is n, the distance to be changed for each measurement of the probe is L, and the ultrasonic beam width is B. The subject is measured so that n and L satisfy the following formula.
  2L ≦ nL ≦ B
[0020]
  An ultrasonic signal processing method according to another aspect of the present invention is the ultrasonic processing method described above,A noise removing step for removing noise included in the new time series signal is further provided.
[0021]
  An ultrasonic signal processing method according to another aspect of the present invention is the ultrasonic processing method described above,The noise removing step removes noise using a bandpass filter.
[0022]
  An ultrasonic signal processing method according to another aspect of the present invention is the ultrasonic processing method described above,In the noise removal step, an integer K is obtained by the following equation using the noise frequency f and the sampling frequency fs for measuring the time series signal, and the signal value is not zero in the new time series signal. When the number of measurement points is equal to or less than K, the signal value of the continuous measurement points is set to zero.
K = [Fs / (2f)]
[0023]
  An ultrasonic signal processing method according to another aspect of the present invention is the ultrasonic processing method described above,The object is a cast material, refractory brick, or concrete.
[0024]
  In addition, other aspects of the present inventionThe ultrasonic measurement apparatus is an ultrasonic measurement apparatus that performs flaw detection or thickness measurement of an object using ultrasonic waves, and a plurality of time-series signals for each position measured by changing the position of the probe with respect to the object. When the polarity of at least one signal value among the signal values at the same time of the acquisition means and the plurality of time-series signals is different, the signal value at that time is set to 0; For a plurality of signal values at the same time point of the time series signal, the signal value having the maximum absolute value, the average value of the signal values from the largest absolute value to the predetermined number, the signal value having the second largest absolute value, all An arithmetic means for obtaining a new time series signal using any one of the average values of the signal values as the signal value of the time series signal at that time, and based on the new time series signal, Equipped with measuring means for flaw detection or thickness measurement It was.
[0025]
  In addition, other aspects of the present inventionThe ultrasonic measurement apparatus is an ultrasonic measurement apparatus that performs flaw detection or thickness measurement of an object using ultrasonic waves, and a plurality of time-series signals for each position measured by changing the position of the probe with respect to the object. The acquisition means to acquire and the signal values at the same time of the plurality of time series signals are classified into two positive and negative polarities, and when the number of signals belonging to a small number of polarity is more than a predetermined number, the signal value at that time In other cases, for a plurality of signal values belonging to many polarities, the signal value having the maximum absolute value, the average value of the signal values from the larger absolute value to the predetermined number, and the absolute value being 2 Calculating means for obtaining a new time-series signal using any one of the second largest signal value and the average value of all the signal values as the signal value of the time-series signal at that time, and the new time series Performs flaw detection or thickness measurement of the subject based on the signal And a measuring means.
[0026]
  Moreover, an ultrasonic measurement apparatus according to another aspect of the present invention is provided.The ultrasonic measurement apparatus described above further includes noise removal means for removing noise included in the new time series signal, the noise removal means for measuring the frequency f of the noise and the time series signal. Using the sampling frequency fs, an integer K is obtained by the following equation. When the number of consecutive measurement points whose signal value is not 0 is equal to or less than K in the new time series signal, the signal value of the continuous measurement point is calculated. This is an ultrasonic measurement device with zero.
K = [Fs / (2f)]
[0027]
  Moreover, an ultrasonic measurement apparatus according to another aspect of the present invention is provided.In the ultrasonic measurement apparatus described above, the object is a cast material, a refractory brick, or concrete.
[0029]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a configuration diagram showing a first embodiment of the present invention. In the present embodiment, the inspection material 1 is fixed, and the position of one probe 2 is moved and measured at four places. The probe 2 is connected to a transmitter and a receiver not shown in the figure. Then, transmission and reception of ultrasonic waves are performed at each of the four locations, and time series signals of ultrasonic waves at the respective positions are stored in the waveform storage unit 3. Here, the ultrasonic time-series signal refers to an ultrasonic reception signal for each elapsed time. A plurality of time-series signals obtained in this way are sent to the calculation unit 4 and subjected to calculations for reducing scattering noise.
[0030]
FIG. 2 is a diagram for explaining the principle of the present invention.
[0031]
In the flaw detection and thickness measurement, a defect that is a signal echo or a reflected echo from the back surface is measured. Even if the position of the probe 2 is changed, the distance between the interface where the signal echo is generated and the probe 2 is too small. Measurement conditions can be adjusted so as not to change. Then, in this case, the polarities of signal echoes included in a plurality of time-series signals obtained by changing the position of the probe 2 are all aligned.
[0032]
On the other hand, scattering noise is caused by interference of reflected waves from scatterers at various positions. Therefore, if the position of the probe 2 is different, its waveform changes sensitively.
[0033]
As a result of the difference in the nature of the signal, it is considered that the signal echo is always a signal of a certain level or higher at a predetermined position of the received signal even if the position of the probe 2 is changed. Regarding the scattered noise, when the position of the probe 2 is changed, the signal value at a predetermined position of the received signal changes randomly. Therefore, if the signal value having the smallest absolute value among the signal values of the plurality of time series signals at the same time is used as the signal value of the time series signal at that time, the scattering noise is reduced, and the S / N is improved. Can do.
[0034]
The computing unit 4 performs signal processing based on this principle. That is, N time series signals obtained by changing the position of the probe 2 are xj (i) (j = 1 to N), and the absolute value of the signal values at the same time of a plurality of time series signals is The signal value that is the minimum is obtained as time-series signal y (i) at that time using equation (1).
[0035]
y (i) = min | xj (i) | Equation (1)
j = 1 to N
FIG. 3 is a diagram showing experimental results according to the first embodiment of the present invention. This experiment was conducted in order to verify the effectiveness of the calculation of equation (1).
[0036]
In this experiment, a φ19 mm ultrasonic probe was used, and a φ1 mm flat bottom artificial defect processed into a steel material having a crystal grain size of about 120 μm was detected by ultrasonic waves having a frequency of 5 MHz. (1) in FIG. 3 is a diagram showing a flaw detection signal at a certain point on the steel material. A signal generated at a position of 3 μs represents a defect signal echo, and a signal generated at other positions is a noise component mixed with scattered noise or reflected echo from the bottom surface.
[0037]
(2) of FIG. 3 is a figure which shows the result of having averaged the flaw detection signal obtained by moving 6 points | pieces on a steel material by 1 mm pitch. In both figures, S / N is low because of the presence of scattered noise around the defect signal echo. (3) in FIG. 3 is a result of performing the minimum value calculation using Expression (1). When (1), (2), and (3) in FIG. 3 are compared, S / N is 6 dB to 10 dB. It can be seen that it has improved.
[0038]
FIG. 4 is a diagram illustrating the relationship between the probe movement pitch, the number of signals, and the S / N reduction effect. This figure has shown the result of having calculated S / N by calculating several times by changing a movement pitch, and calculating Formula (1). In the figure, the horizontal axis indicates the movement pitch, and the mark shape is changed according to the number of times of measurement (number of signals).
[0039]
When the moving pitch was as small as 1 mm, the S / N was higher as the number of signals was larger. It is considered that a defect signal having a certain level or more is output in each flaw detection signal, so that a larger number of signals has an advantageous effect on reducing scattering noise. However, when the moving pitch is as large as 3 mm, the S / N is lowered when the number of signals is large. This is presumably because if the number of signals is increased, the probe is moved away from the defect position, so that the defect signal appearing in the flaw detection signal is reduced, and the S / N is lowered due to the influence.
[0040]
Therefore, it can be seen that there is an optimum combination of movement pitch and number of signals depending on the size and position of the defect. Under the conditions of this experiment, the optimum amount of movement was a movement pitch of 2 mm and a signal number of 6.
[0041]
Next, the results obtained in this experiment will be considered.
[0042]
(1) in FIG. 5 shows the arrangement of the defect and the probe, and (2) in FIG. 5 schematically shows the flaw detection signal at each position. As can be seen from FIG. 5, when the probe 2 is at the position (1) immediately above the defect, a large defect echo is obtained, but as the position (2) and (3) move away from the defect, the defect echo is obtained. Becomes smaller.
[0043]
Therefore, when the signal processing method represented by the expression (1) which is the first embodiment is adopted, the moving distance (= moving pitch × (number of signals N−1)) of the probe 2 is the ultrasonic wave. It can be seen that it should not deviate from a predetermined value determined based on the beam width, and that the number of measurements should be as large as possible within the limits.
[0044]
FIG. 6 is a diagram showing the relationship between the moving distance (= moving pitch × (number of signals N−1)) and S / N based on the result of FIG. In FIG. 6, the horizontal axis represents the moving distance as a multiple of the wavelength λ of the ultrasonic wave, and the vertical axis represents S / N. Also, the marks used in the figure are the same as those used in FIG. The ultrasonic beam width is B, and the positions of 0.8B and 0.6B are shown by dotted lines in the figure. In the present embodiment, B is about 15 mm and λ is about 1.18 mm.
[0045]
According to FIG. 6, the S / N increases until the movement distance is 0 to 8λ, but the S / N starts to decrease at a movement distance exceeding the movement distance. From the definition of the movement distance, it can be seen that the position of the movement distance at which the S / N starts to decrease varies depending on the number of measurements.
[0046]
Therefore, when the S / N is 4 or more, the range of the moving distance recognized as having an S / N improvement effect is 3λ to the beam width B, and when the S / N is 10 or more, the S / N improvement effect is noticeable. It can be seen that the travel distance range is 5λ to beam width B × 0.8.
[0047]
According to the above consideration, when the signal processing method represented by the equation (1) is adopted, if the measurement is performed so as to satisfy the conditions of the moving distance range described above, even if the measurement is several times at most, S / N can be effectively improved.
[0048]
Next, the configuration of the second embodiment according to the present invention will be described. The second embodiment differs from the first embodiment in the processing method of the calculation unit 4.
[0049]
As described above, the scattered noise is obtained by measuring reflected waves from scatterers at various positions. Therefore, when the position of the probe 2 is changed, the polarity of the signal value is observed to be positive or negative. Then, using this phenomenon, if the polarities of the signal values at the same time of multiple time series signals are different, the time series signal at that time is estimated to be scattering noise, and in this case the signal value of the time series signal is set to zero. If so, it is expected that the scattering noise is reduced and the S / N is improved.
[0050]
The computing unit 4 performs signal processing based on this principle. That is, the polarities of the signal values at the same time of the plurality of time series signals xj (i) (j = 1 to N) are compared, and the signal value of the time series signal y (i) at that time is based on the equation (2). Ask.
[0051]
y (i) = 0 If even one of xj (i) has different polarity
y (i) = max (xj (i)) When all xj (i) are on the + side
y (i) = min (xj (i)) When all xj (i) are on the negative side
... Formula (2)
Note that the output y (i) of the equation (2) may finally be an absolute value. These calculations can be calculated using a CPU.
[0052]
Here, equation (2) can be generalized and expressed as follows.
[0053]
The polarities of the signal values at the same time of a plurality of time-series signals xj (i) (j = 1 to N) are compared, the N signal values are classified into positive polarity and negative polarity, and either positive or negative Check if there is a lot of polarity. Then, the signal value of the time series signal y (i) at that time is obtained based on the equation (3).
[0054]
When the number of signals with less polarity is M or more
y (i) = 0
When the number of signals with less polarity is less than M
y (i) = max (xj (i)) When the polarity is +
y (i) = min (xj (i)) When the polarity is-
... Formula (3)
Here, it is desirable to obtain the value of M, that is, the value of the ratio M / N for giving M, based on measurement data in an actual machine.
[0055]
Note that the output y (i) of the expression (3) may finally be an absolute value. These calculations can be calculated using a CPU.
[0056]
FIG. 7 is a diagram showing the relationship between the probe movement pitch, the number of signals, and the S / N reduction effect.
[0057]
(1) in FIG. 8 shows the arrangement of defects and probes, and (2) in FIG. 8 schematically shows flaw detection signals at the respective positions. As can be seen from FIG. 8, when the probe 2 is located at the position (1) directly above the defect, a large defect echo is obtained, but the defect echo increases as the positions (2) and (3) move away from the defect. Becomes smaller. Since the position (3) is out of the defect, a signal having a negative polarity at the defect position may be generated due to the influence of scattering noise.
[0058]
For this reason, when the algorithm shown in Expression (2), which is the second embodiment, is applied, if there is a signal with a different polarity, the signal at that position becomes 0, and the defect cannot be detected.
[0059]
Therefore, when adopting the algorithm expressed by the equation (2), the measurement must be performed at such a moving distance that a defect exists in the measurement range. And if a measurement is performed within the movement distance range, the S / N improvement effect can be made remarkable by increasing the number of times of measurement by the algorithm of Expression (2) that employs the maximum value.
[0060]
That is, in the signal processing method of the second embodiment, the movement distance is within the beam width B and the number of measurements is two or more so that a defect exists in the measurement range. Even if it exists, S / N can be improved effectively.
[0061]
FIG. 9 is a block diagram showing a third embodiment of the present invention. In this embodiment, a two-probe method using two types of probes, a transmitting probe and a receiving probe, is used, and the position of the probe is six directions from (1) to (6) in FIG. It is changed to get data.
[0062]
FIG. 10 is a diagram showing the results of measuring the thickness of cast copper, which is a refractory cooling plate, using the third embodiment of the present invention. Here, an experiment was conducted with an iron skin covering the cast copper and a stamp material having a diameter of 50 mm, using two probes having a diameter of 20 mm, a frequency of 200 kHz, and a distance of 30 mm between the probes.
[0063]
(1) to (6) in FIG. 10 are raw reception signals in six directions, (7) in FIG. 10 is a result of calculation according to the algorithm of equation (1), and (8) in FIG. ) Shows the result of operation according to the algorithm. Although the positions of the arrows in (7) and (8) of FIG. 10 represent the back surface position to be measured, the signals from the back surface that cannot be determined from the raw signals by the calculations according to the equations (1) and (2) Became clear.
[0064]
(9) of FIG. 10 is a result of measurement by one probe method using one probe of φ40 mm. In general, the larger the probe size, the better the S / N against scattered noise. However, because the probe size is large in this experiment, the probe is affected by the sticking of the stamp material remaining on the cast copper. The vibration surface of the child could not contact the cast copper uniformly. As a result, scattering noise was conspicuous, and echo from the back surface could not be detected with good S / N.
[0065]
On the other hand, since the contact area is small when a φ20 mm probe is used, it is possible to uniformly contact the cast copper while avoiding the protrusion of the stamp material remaining on the cast copper, and the measurement can be performed as described above. did it.
[0066]
As described above, even when a probe having a small cross section is used in a limited opening area, a high S / N ratio can be obtained by using the signal processing method of the present invention. This effect is not limited to cast copper refractories, but also applies to the measurement of refractories such as carbon bricks.
[0067]
FIG. 11 shows an example in which the signal shown in (8) of FIG. 10 is subjected to filter processing using a bandpass filter (100 to 300 Hz), and pulse-like noise appearing in the signal is removed. S / N was improved.
[0068]
FIG. 12 is a block diagram showing a fourth embodiment of the present invention. In the present embodiment, mechanical movement is not required by electrically switching a plurality of probes 2.
[0069]
Each probe 2 is connected to a transmission unit and a reception unit (not shown). The transmission unit and the reception unit operate in order by time division, and transmission and reception of ultrasonic waves are performed at each of the four positions. The received signals are sequentially read and stored in the waveform storage unit 3 by the multiplexer 5. When the series of signals is obtained, the calculation unit 4 performs calculation using the formula (1), the formula (2), or the formula (3). Thereby, S / N good data can be obtained in a short time without mechanically moving the probe.
[0070]
FIG. 13 is a diagram illustrating a method of removing pulse noise according to the fifth embodiment of the present invention.
[0071]
In the calculation of Expression (2), if even one of the signals has a different polarity, the calculation is forcibly made zero, so the signal becomes zero in a considerable part of the scattering noise whose phase changes. However, if the polarity of a plurality of signals happens to be all at one point, they will remain as pulse noise as indicated by arrows in FIG.
[0072]
This can be removed by the following calculation. That is, the pulse width of the pulse-like noise appears only as one point or at most several consecutive points in time series. For this reason, when the data on both sides of one point or a plurality of points in the time-series data is zero, it is determined as pulse-like noise, and can be removed if the data at one point or a plurality of points is set to zero.
[0073]
On the other hand, the target signal has a certain frequency component, and the number of points that are not continuously zero data is several times greater than that of pulsed noise. Therefore, when the predetermined score or more is not continuously 0, it is determined that the noise is not pulse-like noise. (2) in FIG. 13 is obtained by removing pulse-like noise by the above algorithm.
[0074]
Here, the predetermined score K for determining as noise is expressed by Equation (4), where f is the noise frequency and fs is the sampling frequency.
[0075]
K = [fs / (2f)] (4)
Note that since the pulse-like noise has a frequency different from that of the target signal, it can be removed using an appropriate band-pass filter.
[0076]
FIG. 14 is a diagram showing a configuration of an ultrasonic measurement apparatus to which the signal processing method according to the present invention is applied.
[0077]
This ultrasonic measurement apparatus receives and amplifies signals from probes 2a and 2b for inspecting the inspection material 1, a transmitter 6 for transmitting ultrasonic pulses to the probe 2a, and the probe 2b. A receiving unit 7, a multiplexer 5 for switching a plurality of received signals, an AD converter 8 for converting the received signals into digital signals, a processing device 9 for processing the converted digital signals, and a movement for moving the probes 2a and 2b The apparatus 13 is configured.
[0078]
The processing device 9 includes a control unit 10, a waveform storage unit 3, a calculation unit 4, a noise removal unit 11, and a determination unit 12. The calculation unit 4 includes a minimum value calculation unit 4a and a polarity calculation unit 4b. It has been.
[0079]
Here, the control unit 10 comprehensively controls each unit of the ultrasonic measurement apparatus. Therefore, in this description, the operation of the control unit 10 is not described unless particularly necessary.
[0080]
Next, the operation of the ultrasonic measurement apparatus will be described.
FIG. 15 is a flowchart showing a schematic procedure of the signal processing method according to the present invention.
[0081]
When the inspector instructs the processing device 9 to start the operation of ultrasonic measurement, in step S1, the control unit 10 outputs a control signal to the moving device 13 so that the probes 2a and 2b are moved to predetermined positions. To do.
[0082]
When the movement completion signal is received from the moving device 13 together with the position information of the probes 2a and 2b in step S2, the control unit 10 instructs the transmission unit 6 to transmit an ultrasonic pulse in step S3. The receiving unit 7 receives and amplifies the signal from the probe 2 b, and outputs the amplified signal to the processing device 9 via the multiplexer 5 and the AD converter 8. In step S3, a transmission pulse may be always output to instruct to store the waveform.
[0083]
In step S4, the waveform storage unit 3 stores the input ultrasonic signal in the memory in association with the position information. In step S5, the control unit 10 checks whether or not a predetermined number of ultrasonic signals have been measured. If No in step S5, that is, if a predetermined number of ultrasonic signals have not yet been measured, the control unit 10 On the other hand, a control signal is output so that the probes 2a and 2b are moved by one pitch.
[0084]
If Yes in step S5, that is, if a predetermined number of ultrasonic signals are being measured, in step S6, an operation corresponding to the signal processing mode is performed.
[0085]
That is, when the signal processing mode is the minimum value processing, the minimum value calculation unit 4a is activated and performs the signal processing represented by the above formula (1). When the signal processing mode is polarity processing, the polarity calculation unit 4b is activated and performs signal processing represented by the above-described formula (2) or formula (3).
[0086]
Subsequently, in step S7, the noise removing unit 11 performs a pulse noise removing process using the algorithm described above. This noise removal may use a band pass filter. And after removing noise, the determination part 12 determines the presence or absence of a defect by step S8.
[0087]
In the present embodiment, the two-probe method using the probes 2a and 2b is adopted, but a single-probe method using only the probe 2a may be used.
[0088]
In addition, the probes 2a and 2b are moved using the moving device 13, but a plurality of probes 2a and 2b are installed at positions where the probes 2a and 2b are measured in advance, and each measuring point is switched by the multiplexer 5. It may be measured. Further, these may be combined, and the moving device 13 may be configured to move the plurality of probes 2a and 2b.
[0089]
Furthermore, each function of the processing device 9 may be configured by hardware, or the processing device 9 may be configured by using a CPU, and each function may be realized by using software.
[0090]
In the present embodiment, the signal processing method according to the present invention is applied to the flaw detection apparatus. However, the present invention is not limited to this form and can be applied to a thickness measuring apparatus.
[0091]
In the present invention, it is necessary to pay attention to the following points regarding the amount of movement of the probe. That is, when flaw detection is performed by a single detection method, since the defect echo cannot be obtained when the beam width is exceeded, the movement amount should be kept within the beam width. When flaw detection is performed by the two-probe method, the angle of the two probes viewed from the defect must not exceed the pointing angle of the probe. Therefore, the distance between the probes is determined according to the expected position of the defect. When the thickness is measured, the moving amount can be wide in both the one and two methods, but when the thickness is locally changed, it is the same as the case of the flaw detection.
[0092]
In addition, since the above-described embodiments include various stages of the invention, various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, the problem described in the column of the problem to be solved by the invention can be solved, and the effect described in the column of the effect of the invention Can be extracted as an invention.
[0093]
Moreover, if the ultrasonic signal processing method of this invention is utilized, the technique relevant to the measuring object can be developed. For example, the blast furnace repair method can be obtained by measuring the remaining thickness of the refractory in the blast furnace according to the present invention. That is, “the step of measuring the thickness of the blast furnace refractory by the ultrasonic signal processing method according to any one of claims 1 to 9 and the step of determining the timing of blast furnace repair based on the measurement result of the step”. The repair method of the blast furnace which has. "
[0094]
Further, in claim 2, “in other cases, the signal value having the maximum absolute value among the signal values at the same time of the plurality of time-series signals is used as the signal value of the time-series signal at that time. In claim 3, the signal value having the maximum absolute value among the signal values belonging to many polarities is defined as the signal of the time-series signal at that time. A new time-series signal is obtained as a value, "and the new time-series signal is not necessarily" a signal having the maximum value ".
[0095]
For example, (1) a value obtained by averaging a predetermined number of these signal values may be averaged, or (2) the second value may be replaced with the maximum value as a new time series signal (the maximum value is (If it is adopted, noise may be adopted and an abnormal value may be obtained.) (3) Further target signal values (excluding signal values with opposite polarity or signal values with less polarity) All of the average values may be used as new time series signals.
[0096]
The processing method may be modified without departing from the gist of the present invention, and these processing methods may be used in combination as appropriate.
[0097]
【The invention's effect】
As described above, according to the signal processing method according to the present invention, a high S / N can be obtained with a small number of data without using a probe with a large area in ultrasonic measurement of a highly scattering material. .
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing a first embodiment of the present invention.
FIG. 2 is a diagram illustrating the principle of the present invention.
FIG. 3 is a diagram showing experimental results according to the first embodiment of the present invention.
FIG. 4 is a diagram showing the relationship between the probe movement pitch, the number of signals, and the S / N reduction effect.
FIG. 5 is a diagram schematically showing the arrangement of a defect and a probe and a flaw detection signal at each position.
FIG. 6 is a diagram showing a relationship between a moving distance and S / N.
FIG. 7 is a diagram showing the relationship between the probe movement pitch, the number of signals, and the S / N reduction effect.
FIG. 8 is a diagram schematically showing the arrangement of a defect and a probe and a flaw detection signal at each position.
FIG. 9 is a configuration diagram showing another embodiment of the present invention.
FIG. 10 is a diagram showing the results of measuring cast copper, which is a refractory cooling plate.
FIG. 11 is a diagram showing an example in which filter processing is performed using a bandpass filter.
FIG. 12 is a configuration diagram showing another embodiment of the present invention.
FIG. 13 is a diagram showing a method for removing pulse noise.
FIG. 14 is a diagram showing a configuration of an ultrasonic measurement apparatus to which the signal processing method according to the present invention is applied.
FIG. 15 is a flowchart showing a schematic procedure of a signal processing method according to the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Test | inspection material, 2 ... Probe, 3 ... Waveform memory | storage part, 4 ... Calculation part,
4a ... minimum value calculation unit, 4b ... polarity calculation unit, 5 ... multiplexer,
6 ... Transmitter, 7 ... Receiver, 8 ... AD converter, 9 ... Processing device,
DESCRIPTION OF SYMBOLS 10 ... Control part, 11 ... Noise removal part, 12 ... Determination part, 13 ... Moving apparatus

Claims (11)

An ultrasonic signal processing method of an ultrasonic measurement device that performs flaw detection or thickness measurement of a subject by ultrasonic waves,
Obtain a plurality of time series signals for each position measured by changing the position of the probe relative to the subject,
When the polarity of at least one signal value among the signal values at the same time of the plurality of time series signals is different, the signal value at that time is set to 0, otherwise, at the same time of the plurality of time series signals For the plurality of signal values , the signal value having the maximum absolute value , the average value of the signal values from the largest absolute value to the predetermined number, the signal value having the second largest absolute value, the average value of all the signal values, A new time-series signal with any one of the values as the signal value of the time-series signal at that time,
An ultrasonic signal processing method, wherein flaw detection or thickness measurement is performed on the subject based on the new time series signal. An ultrasonic signal processing method of an ultrasonic measurement device that performs flaw detection or thickness measurement of a subject by ultrasonic waves,
Obtain a plurality of time series signals for each position measured by changing the position of the probe relative to the subject,
The signal values at the same time of these multiple time-series signals are classified into two positive and negative polarities. If the number of signals belonging to a small number of polarities is greater than or equal to a predetermined number, the signal value at that time is set to 0, otherwise in the case of larger plurality of signal values belonging to the polarity, the absolute value of the signal value is the maximum, average value of the signal value from the larger absolute value to a predetermined number, the greater the signal value the absolute value of the second, Find a new time-series signal with one of the average values of all signal values as the signal value of the time-series signal at that time,
An ultrasonic signal processing method, wherein flaw detection or thickness measurement is performed on the subject based on the new time series signal. The number of a plurality of time-series signals for each measured position is n, the distance to be changed for each measurement of the probe is L, and the beam width of the ultrasonic wave is B,
The ultrasonic signal processing method according to claim 1 , wherein the subject is measured such that n and L satisfy the following formula.
2L ≤ nL ≤ B Ultrasonic signal processing method according to one of claims 1 to 3, further comprising a noise removal step removing noise contained in the time-series signal the new. The ultrasonic signal processing method according to claim 4 , wherein the noise removing step removes noise using a band-pass filter. In the noise removal step, an integer K is obtained by the following equation using the noise frequency f and the sampling frequency fs for measuring the time series signal, and the signal value is not zero in the new time series signal. 5. The ultrasonic signal processing method according to claim 4 , wherein when the number of measurement points is equal to or less than K, the signal value of the continuous measurement points is set to 0.
K = [fs / (2f)] The subject is, the ultrasonic signal processing method according to one of claims 1 to 6, characterized in that a cast material or refractory brick or concrete.   An ultrasonic measurement device that performs flaw detection or thickness measurement of an object using ultrasonic waves,
  Acquisition means for acquiring a plurality of time series signals for each position measured by changing the position of the probe with respect to the subject;
  The polarity of at least one of the signal values at the same time of the plurality of time series signals is different. The signal value at that time is set to 0, and in other cases, the signal value having the maximum absolute value and the value with the larger absolute value are determined for the plurality of signal values at the same time of the plurality of time-series signals. A new value that uses any one of an average value of signal values up to a predetermined number, a signal value having the second largest absolute value, and an average value of all signal values as the signal value of the time-series signal at that time A computing means for obtaining a time-series signal;
  Measuring means for performing flaw detection or thickness measurement of the subject based on the new time series signal;
An ultrasonic measurement apparatus comprising: An ultrasonic measurement device that performs flaw detection or thickness measurement of an object using ultrasonic waves,
Acquisition means for acquiring a plurality of time series signals for each position measured by changing the position of the probe with respect to the subject;
The signal values at the same time of these multiple time-series signals are classified into two positive and negative polarities. If the number of signals belonging to a small number of polarities is greater than or equal to a predetermined number, the signal value at that time is set to 0, otherwise In the case of a plurality of signal values belonging to many polarities, the signal value having the maximum absolute value, the average value of the signal values from the largest absolute value to the predetermined number, the signal value having the second largest absolute value, An arithmetic means for obtaining a new time series signal having any one of the average values of all the signal values as the signal value of the time series signal at that time point;
An ultrasonic measurement apparatus comprising: a measurement unit that performs flaw detection or thickness measurement of the subject based on the new time series signal.   A noise removing means for removing noise included in the new time series signal;
  The noise removing means obtains an integer K by the following equation using the frequency f of the noise and the sampling frequency fs for measuring the time series signal, and the signal value is not zero in the new time series signal. The ultrasonic measurement apparatus according to claim 8 or 9, wherein when the number of measurement points is equal to or less than K, the signal value of the continuous measurement points is set to 0.
  K = [Fs / (2f)] The ultrasonic measurement apparatus according to claim 8 , wherein the object is a cast material, a refractory brick, or concrete.

Images