JP2005351660A - Ultrasonic flaw detection method and ultrasonic flaw detection apparatus - Google Patents

Abstract

An object of the present invention is to efficiently and accurately measure a defect position and a defect depth in a welded portion of an austenitic stainless steel, a nickel-base superalloy, or the like so as to ensure the soundness of equipment and members.
An array-type ultrasonic probe comprising a plurality of vibrators is used for a welded portion made of a dissimilar metal including the same kind of austenitic stainless steel and nickel-base superalloy, or carbon steel and low alloy steel. Ultrasonic waves are transmitted and received, and signal processing such as fast Fourier transform, wavelet transform, and maximum entropy method processing is performed on the detection signal. These feature amounts are extracted as characteristic value data, and defect identification is performed by identifying the defect signal and the weld structure signal from the difference between the characteristic value data.
[Selection] Figure 1

Description

  The present invention is an ultrasonic flaw detection capable of efficiently and highly accurately performing defect inspection of welded parts such as austenitic stainless steel and nickel-base superalloy, which are difficult to distinguish between a defect signal and a reflected signal from a welded part structure. The present invention relates to a method and an ultrasonic flaw detector.

  For example, welded parts made of dissimilar metals such as austenitic stainless steel, the same kind of nickel-base superalloy, carbon steel, low alloy steel, etc., which are high heat resistant materials used in gas turbine equipment of power plants, Depending on the operating environment such as temperature and welding residual stress, deterioration and damage may occur over time and defects such as cracks may occur. Detecting these defects and measuring and evaluating the dimensions with high accuracy are extremely important matters for stable operation of high-temperature equipment and structures for gas turbines.

  Conventionally, as a defect inspection method for a welded part, generally, a welded part is detected using an oblique probe composed of a single vibrator, and a reflected wave from the defect is detected to obtain a defect position and a defect depth. The pulse reflection method is mainly applied (for example, ultrasonic inspection method for steel welds of Japanese Industrial Standard (JIS Z3060-1996)).

  However, in weld metal parts of welds made of the same or different metals such as austenitic stainless steel and nickel-base superalloy, the crystal grains become coarse due to welding and columnar crystal structures are generated. It is known that ultrasonic attenuation and distortion occur when ultrasonic waves are incident. When the above pulse reflection method is applied, the reflected wave from the defect is buried in the material noise waveform of the weld structure, and the defect waveform cannot be detected, or the pseudo waveform from the material such as the columnar crystal structure is mistaken. May be recognized as a defective waveform. Even if the defect waveform can be detected, the defect position and the defect depth may not be accurately measured.

On the other hand, a defect inspection method using an array type ultrasonic probe composed of a plurality of transducers is also known (for example, Patent Document 1). However, conventionally, this array-type ultrasonic probe is installed directly on the subject or via an acoustic medium such as acrylic, machine oil, water, etc., and defect inspection is performed using flaw detection by linear scanning or flaw detection by sector scanning. In this case, as in the case of the pulse reflection method, it may be impossible to identify whether the detection signal is a signal detected from a defect or a signal detected from material noise of a welded structure.
Japanese Patent Application Laid-Open No. 9-257763

  As described above, in the weld metal part of the weld part composed of the same or different metals such as austenitic stainless steel and nickel-base superalloy, the crystal grains are coarsened by welding to generate a columnar crystal structure, and the specimen When ultrasonic waves are incident, ultrasonic attenuation and distortion occur, so when applying the pulse reflection method using an oblique probe consisting of a single transducer, the reflected wave from the defect causes material noise in the weld tissue. There are cases where the waveform is buried and the defect waveform cannot be detected, or a pseudo waveform from a material such as a columnar crystal structure is erroneously recognized as a defect waveform. Even if the defect waveform can be detected, the defect position and the defect depth may not be accurately measured.

  On the other hand, in the case of normal ultrasonic flaw detection using an array type ultrasonic probe, the detection signal is a signal detected from a defect or detected from material noise in the welded structure, as in the pulse reflection method. It may not be possible to identify whether the signal is a bad signal.

  The present invention has been made in view of such problems, and it is possible to inspect defects in welded parts such as austenitic stainless steel and nickel-base superalloys, which are difficult to distinguish between defect signals and reflected signals from the welded part structure. An object of the present invention is to provide an ultrasonic flaw detection method and an ultrasonic flaw detection apparatus that can be performed efficiently and with high accuracy, and in particular, can measure a defect position and a defect depth with high accuracy.

  In order to achieve the above object, in the invention according to claim 1, a welded portion made of a dissimilar metal including austenitic stainless steel, the same kind of nickel-base superalloy, carbon steel, low alloy steel, or the like is an inspection object. An ultrasonic flaw detection inspection method comprising the steps of installing an array-type ultrasonic probe comprising a plurality of transducers on a subject including the welded portion, the shape and dimensions of the subject, and the material of the welded portion A step of setting flaw detection conditions by selecting any one or a plurality of transducers of the array-type ultrasonic probe as an ultrasonic transmission transducer group and a reception transducer group based on welding conditions, etc. And using the transmitting transducer group and the receiving transducer group to focus and deflect an ultrasonic beam on the inspection target region of the subject, and electronically connecting the transmitting transducer group and the receiving transducer group Scanning, and said sending and receiving A signal processing of the ultrasonic waveform detected by the transducer, and a characteristic value of the signal; a defect signal detected from a defect such as a crack based on the characteristic value obtained in the signal processing process; There is provided an ultrasonic flaw detection method comprising a step of identifying a reflected signal from material noise such as a welded portion structure, and a step of outputting the signal processing result and the signal identification result.

  In the invention according to claim 2, the transmitting transducer group and the receiving transducer group are configured when an ultrasonic beam is transmitted / received to / from the subject using the array-type ultrasonic probe. The number of transducers is set to the same, and the transmission angle and the reception angle of the ultrasonic beam to the subject are set to be the same, linear flaw detection or sector scanning flaw detection is performed, and the detected ultrasonic waveform is signal-processed The ultrasonic flaw detection method according to claim 1, wherein a signal detected from a defect such as a crack and a reflected signal from a welded structure are discriminated based on the obtained characteristic value. .

  In the invention according to claim 3, when an ultrasonic beam is transmitted to and received from the subject using the array type ultrasonic probe, an incident angle to the subject is set to a predetermined angle at the time of ultrasonic transmission. Set and receive ultrasonic waves, and at the time of reception, set a reception angle so that a predetermined position inside the subject can be received as a sound source, and cover the position of the sound source to cover the entire flaw detection range of the subject Flaw detection is performed sequentially, and the ultrasonic waveform detected for each flaw detection with each sound source is signal-processed. Based on the obtained characteristic values, signals detected from defects such as cracks and the weld structure The ultrasonic flaw detection method according to claim 1, wherein a reflected signal is discriminated.

  In the invention according to claim 5, as the signal processing of the detected ultrasonic waveform, fast Fourier transform processing is performed, and based on the obtained characteristic value, the signal detected from the defect such as a crack and the welded portion structure are used. The ultrasonic flaw detection method according to any one of claims 1 to 3, wherein the reflected signal is identified.

  The invention according to claim 6 uses at least one of a half-value width, a center frequency, a peak frequency, and a spectrum area of the spectrum distribution as the characteristic value of the fast Fourier transform process. Provide flaw detection methods.

  In the invention according to claim 7, as signal processing of the detected ultrasonic waveform, wavelet transform processing is performed, and based on the obtained characteristic value, a signal detected from a defect such as a crack and a weld structure The ultrasonic flaw detection method according to any one of claims 1 to 3, wherein a reflected signal is identified.

  In the invention according to claim 8, the signal processing by the maximum entropy method is performed as the signal processing of the detected ultrasonic waveform, and the signal detected from the defect such as a crack and the welded portion based on the obtained characteristic value The ultrasonic flaw detection method according to any one of claims 1 to 4, wherein a reflected signal from a tissue is identified.

  In the invention according to claim 9, the center frequency, peak frequency, and half width of the spectrum distribution are used as signal processing characteristic values by the maximum entropy method, and signals detected from defects such as cracks and from the weld structure The ultrasonic flaw detection method according to claim 8, wherein a reflected signal is discriminated.

  The invention according to claim 11 is an ultrasonic flaw detection apparatus for welds composed of dissimilar metals including austenitic stainless steel, the same kind of nickel-base superalloy, or carbon steel, low alloy steel, etc. A probe placement means for placing an array-type ultrasonic probe comprising a plurality of transducers on a subject, a moving means for moving the array-type ultrasound probe, a shape of the subject, and One or more transducers of the array-type ultrasonic probe are selected as an ultrasonic transmission transducer group and a reception transducer group based on dimensions, welded material, welding conditions, and the like. A flaw detection condition setting stage for setting flaw detection conditions including: control means for converging and deflecting an ultrasonic beam on an examination target region of the subject using the transmission transducer group and the reception transducer group; Credit oscillators and A scanning unit that electronically scans the trusted transducer group, a signal processing unit that performs signal processing on the ultrasonic waveform detected by the transducer group for transmission and reception, and obtains a characteristic value of the signal; Signal identifying means for identifying a defect signal detected from a defect such as a crack and a reflected signal from material noise such as a welded structure based on the measured characteristic value, and signal processing results and signal identification by these signal processing means There is provided an ultrasonic flaw detector provided with display means for outputting a signal identification result by the means.

  According to the present invention, in an ultrasonic flaw inspection of a welded portion made of a dissimilar metal such as austenitic stainless steel, the same kind of nickel-base superalloy, or carbon steel, low alloy steel, an array composed of a plurality of vibrators Type ultrasonic probe is used to transmit and receive ultrasonic waves, perform signal processing such as fast Fourier transform, wavelet transform, and maximum entropy method processing on the detection signal, and extract the feature values as characteristic value data. By identifying the defect signal and weld zone structure signal from the difference in data, it becomes possible to identify the defect, and furthermore, the defect position and the depth of the defect can be measured efficiently and with high accuracy by various calculation processing of the detected waveform data. Can do. That is, by performing such flaw detection on the welded portion, high-precision inspection can be performed, and the soundness of the equipment and members can be ensured.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In the following embodiment, for example, a case in which a defect inspection inside a weld metal is performed for a butt weld portion of metals using Inconel (IN738LC (trade name), etc.) which is a nickel-based superalloy as a base material will be described. To do.

  FIG. 1 is an explanatory diagram showing the configuration of an ultrasonic flaw detector. As shown in FIG. 1, a butt weld structure as a subject 1 is configured by joining a pair of parts having the above-mentioned material as a base material 1a with a weld metal 1b. It is assumed that a defect 1c opened on the bottom side, for example, is generated inside. Outside the subject 1, a plurality of transducers A (1), A (2),..., A (i),..., A (n) are provided via a contact medium C typified by water. An array type ultrasonic probe 2 is installed.

  The array-type ultrasonic probe 2 is held by a sensor holding mechanism 3 as probe installation means, and this sensor holding mechanism 3 is connected to the array-type ultrasonic probe 2 and the site 1 to be inspected. It is connected to a probe installation means for installing the distance at an appropriate position and a drive mechanism 4 as a drive means for moving the array type ultrasonic probe 2 to an appropriate position. The amount and method of movement of the drive mechanism 4 are controlled by the drive mechanism control device 5.

  The drive control device 5 is controlled by a control device 6 for overall control together with other means described later. The control device 6 outputs various commands based on information from the database DB1 storing various information such as flaw detection conditions.

  The array-type ultrasonic probe 2 includes one or a plurality of transducers as transmission transducer groups G (p1), G (p2), ..., G (pi), ..., G (pn). And G (r1), G (r2),..., G (ri),.

  An ultrasonic transmitter group 8 for generating ultrasonic waves is connected to each transducer of the transmitting transducer group G (p1)... G (pn), and a receiving transducer group G (r1). ) Is connected to an ultrasonic receiver group 9 for receiving an ultrasonic signal waveform. Then, by controlling each transducer of the ultrasonic transmitter group 8 and the ultrasonic receiver group 9 with a delay time controller 7 that operates based on a command from the control device 6, the predetermined position of the examination target region is obtained. The ultrasonic beam can be focused and deflected. Further, the set of transmitting transducer groups G (p1)... G (pn) and receiving transducer groups G (r1)... G (n) can detect a subject by predetermined electronic scanning.

  That is, the control device 6, the delay time controller 7, the ultrasonic transmitter group 8, the ultrasonic receiver group 9 and the like are used to form an array type superstructure based on the shape and size of the subject 1, the material of the weld, the welding conditions, and the like. A flaw detection condition setting stage for setting flaw detection conditions such as selecting any one or a plurality of transducers of the acoustic probe as an ultrasonic transmission transducer group and a reception transducer group, and a transmission transducer group And a control means for converging and deflecting the ultrasonic beam on the examination target site of the subject 1 using the receiving vibrator group, and a scanning means for electronically scanning the transmitting vibrator group and the receiving vibrator group. Composed.

  Furthermore, the ultrasonic reception waveform obtained by the electronic scanning is subjected to signal processing by a signal processor 10 as signal processing means, and a signal discriminator 11 as signal discrimination means based on a characteristic value obtained by the signal processing. Thus, it can be identified whether the detection signal is a signal from a defect or a material noise signal of a welded portion structure. This identification is performed based on data stored in the database DB2.

  That is, when the detection signal is a defect, it is possible to calculate defect information such as a defect position and a defect depth, and the signal processing result and the identification processing result are output or displayed on the display device 12 as display means. can do.

  In this embodiment, the contact medium C typified by water is used for flaw detection from the array-type ultrasonic probe 2 to the subject. However, direct contact flaw detection or resin typified by acrylic is used. It is also possible to carry out a flaw detection using as a medium.

  FIG. 2 is a flowchart showing the procedure of an ultrasonic flaw detection method performed using the ultrasonic flaw detection apparatus described above.

  As shown in FIG. 2, first, a target part of the welded part structure is selected based on the past operation history or the like (S101). Next, flaw detection conditions stored in advance in the database DB1 are selected based on the material of the welded portion, the groove shape, welding conditions, and the like (S102). The selected flaw detection conditions are, for example, the number of transmission drive elements, the number of reception drive elements, the transmission incident angle, the reception incident angle, the convergence position, the distance between the probe and the subject 1, and the like. The transmission incident angle and the reception incident angle as the flaw detection conditions will be described in detail later.

  Next, flaw detection is performed on the subject 1 based on the selected flaw detection conditions, and ultrasonic waveform data such as a reflected wave from the defect 1C is recorded (S103).

  Based on the recorded waveform data, signal processing such as amplitude value distribution, fast Fourier transform, wavelet transform, maximum entropy method processing, split spectrum processing, which will be described later, is performed, and characteristic values obtained by this signal processing are obtained. It is calculated (S104).

  Further, the signal processing result is compared with data stored in the database DB2 in advance, and it is identified whether the detection signal is a defect signal or a material noise signal of the welded portion structure (S105).

  Then, based on the results of step S104 and step S105, when the received signal is a defect signal, defect information such as a defect position or a defect depth is calculated (S106), and these results are finally converted into a predetermined format. Is output and displayed on the display device 12 (S107).

  Next, transmission / reception of ultrasonic beams (UB), signal processing, and the like will be described in detail with reference to FIGS.

  FIG. 3A is a conceptual diagram relating to a method of performing ultrasonic transmission to the subject 1 using the array-type ultrasonic probe 2. As shown in FIG. 3 (a), a transmitting transducer group G (p1) selected from a large number of transducers A (1),... A (n) constituting the array-type ultrasonic probe 2. ), G (p2),..., G (pi),..., G (pn) are used to transmit the ultrasonic beam to the subject 1 with a constant incident angle θ, and in one direction (direction of arrow d). Linear scanning as electronic scanning along the line is performed.

  FIG. 3B is a conceptual diagram regarding a case where ultrasonic reception is performed using the array-type ultrasonic probe 2 when the ultrasonic transmission shown in FIG. 3A is performed. As shown in FIG. 3B, the receiving transducer group G (r1), G (r2),..., G (ri),. , G (p2), G (ri),..., G (pn), and the reception angle at the time of reception is set to be the same as the incident angle θ at the time of transmission. Then, flaw detection is performed by linear scanning as electronic scanning along one direction (arrow d direction).

  In this way, the same incident angle θ is set for both transmission and reception, and the receiving vibrator group G (r1), G (r2),..., G (ri),. By setting G (p1), G (p2), G (ri),..., G (pn) to the same number and transmitting / receiving ultrasonic beams (UB) in a form corresponding to each other, Identification can be easily performed.

  That is, flaw detection is performed under certain ultrasonic transmission / reception conditions, and various signal processing is performed using the detected waveform data, so that the defect 1c and material noise can be identified, and the position and defect depth of the defect 1c can be efficiently determined. And it can measure with high accuracy.

  FIG. 4A is another conceptual diagram regarding ultrasonic transmission to the subject 1 using the array-type ultrasonic probe 2. Transmitting transducer groups G (p1), G (p2),..., G (pi),..., G (pn) selected from a large number of transducers 2 constituting the array-type ultrasonic probe 2 ) Is used to detect flaws in linear scanning at an incident angle θp to the subject 1 in the ultrasonic phase.

  FIG. 4B is a conceptual diagram regarding ultrasonic reception using the array-type ultrasonic probe 2 when ultrasonic transmission is performed in FIG. During ultrasonic reception, flaw detection is performed by setting the receiving angle of each transducer Ai to θr1, θr2,..., Θrn so that a predetermined position inside the subject 1 can be received as a sound source.

  By performing such flaw detection, it is possible to capture the change in the detection signal with respect to the refraction angle θ at the target position. Note that information on the entire flaw detection range can be obtained by sequentially changing the sound source position so as to cover the entire flaw detection range of the subject.

  According to the present embodiment, the array type ultrasonic probe is installed at an appropriate position, the flaw detection is performed under an appropriate flaw detection condition without moving the array type ultrasonic probe, and signal processing of detected waveform data is performed. Thus, the defect can be identified, and the defect position and the defect depth can be measured efficiently and with high accuracy. In addition, flaws can be detected under appropriate flaw detection conditions, and signal processing of the detected waveform data can be used to identify defects and material noise in the weld zone structure, and the defect position and defect depth can be measured efficiently and accurately. Can do. Furthermore, by setting the incident angle θ to various settings for the ultrasonic transmission / reception conditions, it becomes possible to distinguish between defects and material noise, and the defect position and defect depth can be measured efficiently and with high accuracy.

  5 to 9 are explanatory diagrams showing specific examples of various signal processing relating to the ultrasonic flaw detection method.

  FIG. 5A is a conceptual diagram regarding ultrasonic reception using the array-type ultrasonic probe 2 when a crack is present inside the weld metal 1 b of the subject 1.

  In this example, the flaw detection is performed with the reception angles (refractive angles) set to θr1, θr2,..., Θrn so that the tip position of the crack 1c inside the subject 1 can be received as a sound source during ultrasonic reception. The detected waveform is schematically shown in FIG. 5B, and the amplitude value Hn of the detected waveform is obtained.

  When the amplitude value of the waveform detected under the flaw detection conditions of the refraction angles θr1, θr2,..., Θrn is obtained, and the amplitude value distribution with respect to the refraction angle at the time of reception is obtained, an amplitude value distribution as shown in FIG.

  As the characteristic value at this time, for example, the half-value width Hw can be obtained as shown in FIG.

  On the other hand, FIG. 6A is a conceptual diagram regarding ultrasonic reception using the array-type ultrasonic probe 2 when no crack exists in the weld metal 1b of the subject 1. FIG.

  At the time of ultrasonic reception, similarly to the case of FIG. 5A, the reception angle (refraction angle) is set to θr1, so that the position of the columnar crystal structure interface 1d, which is a welded portion structure inside the subject 1, can be received as a sound source. The flaw detection is performed by setting θr2,..., θrn. The detected waveform is schematically shown in FIG. 6B, and is reflected from a relatively wide range of the interface, so that the wave number is higher than the signal waveform from the tip of the crack 1c (FIG. 5B). There are many.

  Here, when the amplitude value Hn of the waveform detected under the flaw detection conditions of the refraction angles θr1, θr2,..., Θrn is obtained in the same manner as described above, the amplitude value distribution with respect to the refraction angle at the time of reception is obtained as shown in FIG. An amplitude value distribution is obtained. When the half-value width Hw is obtained as the characteristic value at this time as shown in FIG. 6C, the half-value width Hw is wider than that in the case of FIG.

  Thus, the measured half width Hw is collated with data stored in advance as the database DB2, and it is identified whether the detection signal is a defect signal or a material noise signal of the welded portion structure. be able to.

  Therefore, according to the present embodiment, an appropriate defect can be detected by obtaining a characteristic value from the amplitude value distribution data of the detection signal and performing defect identification from the characteristic value data. Furthermore, by performing various arithmetic processes based on the detected waveform data, the defect position and the defect depth can be measured efficiently and with high accuracy.

  FIG. 7 is an explanatory diagram of another specific example related to the ultrasonic flaw detection method.

  7A1 and 7B are explanatory diagrams showing detection waveforms and analysis results when there is a defect, and FIGS. 7A2 and 7B show detection waveforms and analysis results when there is no defect. It is explanatory drawing shown. Here, for the defect signals shown in FIGS. 7A1 and 7A2, the diffracted wave from the tip of the defect is shown as a detected waveform, and FIGS. 7B1 and 7B2 use the detected waveform. It is a conceptual diagram of a fast Fourier transform process (FFT process) result. For example, a half-value width Hw is obtained as a spectrum characteristic value from the spectrum distribution obtained by the FFT processing.

  The half width Hw measured in this way is collated with data stored in advance as a database, and it can be identified whether the detection signal is a defect signal or a material noise signal of a welded portion structure. .

  Here, as the characteristic values of the FFT processing, the characteristic values obtained by the FFT processing such as the center frequency and the peak frequency of the spectrum distribution are obtained, and this characteristic value data is collated with data stored in advance as a database. It is possible to identify whether the detection signal is a defect signal or a material noise signal of the weld structure.

  Thus, in the present embodiment, a proper defect can be detected by obtaining a characteristic value from the FFT processing of the detection signal and performing defect identification from this characteristic value data. Further, by performing various arithmetic processes based on the detected waveform data, the defect position and the defect depth can be measured efficiently and with high accuracy.

  As a result, as shown in FIG. 7 (B1), when there is a defect, the half width Hw appears smaller than that in FIG. 7 (B2).

  In the above example, the half-value width Hw is applied to the analysis. However, as a modification, the defect may be identified based on the amplitude value of 25% Hw. In other words, it is possible to perform the calculation similarly with Hw of 25% by calculating the energy based on the area X peak frequency with respect to the center frequency.

  FIG. 8 is an explanatory diagram of another specific example relating to the ultrasonic flaw detection method.

  FIGS. 8A1 and 8B1 are explanatory diagrams showing detection waveforms and analysis results when there is a defect, and FIGS. 8A2 and 8B2 show detection waveforms and analysis results when there is no defect. It is explanatory drawing shown. Here, the analysis is a result of wavelet transform processing using the detected waveform.

  In the defect signal, a diffracted wave from the tip of the defect is shown as a detected waveform, and, for example, the number of peak distributions is obtained as a characteristic value from the processing result obtained by the wavelet transform process.

  In the welded structure signal, the detected waveform indicates a received waveform from the welded structure interface, and the number of peak distributions, for example, is obtained as a characteristic value from the result obtained by the wavelet transform process as described above.

  The number of peak distributions measured in this way can be compared with data stored in advance as a database to identify whether the detection signal is a defect signal or a material noise signal of a welded part structure.

  Here, as the characteristic values of the wavelet transformation process, the characteristic values obtained by the wavelet transformation process such as the time-frequency function of each peak distribution, the center frequency of each peak distribution, the peak frequency, etc. are obtained. By comparing with data stored in advance as a database, it is possible to identify whether the detection signal is a defect signal or a material noise signal of the weld structure.

  As a result, as shown in FIG. 8 (B1), when there is a defect, it appears in a smaller range of time-frequency feature amounts than in FIG. 8 (B2).

  Thus, in this embodiment, a proper defect can be detected by obtaining the characteristic value from the wavelet transform process of the detection signal and identifying the defect from the characteristic value data. Further, by performing various arithmetic processes based on the detected waveform data, the defect position and the defect depth can be measured efficiently and with high accuracy.

  FIG. 9 is an explanatory diagram of another specific example relating to the ultrasonic flaw detection method.

  FIGS. 9A1 and 9B1 are explanatory diagrams showing detection waveforms and analysis results when there is a defect. FIGS. 9A2 and 9B2 show detection waveforms and analysis results when there is no defect. It is explanatory drawing shown. Here, as an analysis, it is a conceptual diagram of a signal processing result by a maximum entropy method using a detected waveform.

  In the defect signal, a diffracted wave from the tip of the defect is shown as a detected waveform, and for example, the center frequency f1 of the peak distribution is obtained as a characteristic value from the spectrum distribution (b) obtained by signal processing by the maximum entropy method.

  In the weld structure signal, the detected waveform indicates a received waveform from the weld structure interface, and the center frequency f2 of the peak distribution, for example, is obtained as a characteristic value from the spectrum distribution (b) obtained by the maximum entropy method processing as described above.

  The center frequency thus measured can be compared with data stored in advance as a database to identify whether the detection signal is a defect signal or a material noise signal of the welded portion structure.

  Here, the characteristic values obtained by the signal processing by the maximum entropy method, such as the peak frequency of the spectrum distribution, the half width, etc. are obtained as the characteristic values of the signal processing by the maximum entropy method. By comparing with the stored data, it is possible to identify whether the detection signal is a defect signal or a material noise signal of the weld structure.

  As a result, as shown in FIG. 9 (B1), when there is a defect, the peak distribution at the center frequency f1 appears in a smaller range than in FIG. 9 (B2).

  Thus, in this embodiment, a proper defect can be detected by obtaining a characteristic value from signal processing by the maximum entropy method of a detection signal and performing defect identification from this characteristic value data. Further, by performing various arithmetic processes based on the detected waveform data, the defect position and the defect depth can be measured efficiently and with high accuracy.

  Further, by performing signal processing by the split spectrum method on the detection signal, obtaining by signal processing, and performing defect identification from the obtained characteristic value data, proper defect detection can be performed. Furthermore, by performing various arithmetic processes based on the detected waveform data, the defect position and the defect depth can be measured efficiently and with high accuracy.

1 is a schematic configuration diagram showing an ultrasonic flaw detector according to an embodiment of the present invention. The flowchart which shows the ultrasonic flaw detection method which concerns on one Embodiment of this invention. (A), (B) is explanatory drawing which shows the ultrasonic transmission / reception method in the ultrasonic flaw detection method of the said one Embodiment. (A), (B) is explanatory drawing which shows the ultrasonic transmission / reception method in the ultrasonic flaw detection method of the said one Embodiment. (A), (B), (C) is explanatory drawing which shows the example of the signal processing of the detection signal from the defect in the ultrasonic flaw detection method of the said one Embodiment. (A), (B), (C) is explanatory drawing which shows the example of the signal processing of the detection signal from the welding part structure | tissue in the ultrasonic flaw detection method of the said one Embodiment. (A1), (B1), (A2), (B2) is explanatory drawing which shows the method by a fast Fourier-transform process as signal processing of the detection signal in the ultrasonic flaw detection method of the said one Embodiment. (A1), (B1), (A2), (B2) is explanatory drawing which shows the method by the wavelet transformation process as signal processing of the detection signal in the ultrasonic flaw detection method of the said one Embodiment. (A1), (B1), (A2), (B2) is explanatory drawing which shows the method by the maximum entropy method as signal processing of the detection signal in the ultrasonic flaw detection method of the said one Embodiment.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Subject, 2 ... Array type ultrasonic probe, 1a ... Base material, 1b ... Weld metal, 1c ... Defect, C ... Contact medium, A (1), A (2), A (i), A (N) ... vibrator, 3 ... sensor holding mechanism, 4 ... drive mechanism, 5 ... drive mechanism control device.

Claims (11)

An ultrasonic flaw inspection method for inspecting a weld comprising a dissimilar metal such as austenitic stainless steel, the same kind of nickel-base superalloy, or carbon steel, low alloy steel, etc. Based on the step of installing an array-type ultrasonic probe comprising a plurality of transducers on the specimen, the shape and dimensions of the subject, the material of the weld, the welding conditions, etc. Selecting any one or a plurality of transducers as an ultrasonic transmission transducer group and a reception transducer group and setting flaw detection conditions; and using the transmission transducer group and the reception transducer group A step of focusing and deflecting an ultrasonic beam on a region to be examined of the subject and electronically scanning the transmitting transducer group and the receiving transducer group, and an ultrasonic wave detected by the transmitting / receiving transducer Process the waveform, The step of obtaining the characteristic value of the signal of the signal and the defect signal detected from the defect such as a crack and the reflected signal from the material noise such as the welded structure based on the characteristic value obtained in the signal processing step An ultrasonic flaw detection method comprising: a step; and a step of outputting the signal processing result and the signal identification result. When transmitting and receiving an ultrasonic beam to and from the subject using the array-type ultrasonic probe, the number of transducers constituting the transmission transducer group and the reception transducer group is set to be the same. In addition, the transmission angle and reception angle of the ultrasonic beam to the subject are set to be the same, flaw detection of linear scanning or sector scanning is performed, the detected ultrasonic waveform is signal-processed, and the obtained characteristic value is obtained. The ultrasonic flaw detection method according to claim 1, wherein a signal detected from a defect such as a crack and a reflected signal from a welded portion structure are identified based on the crack. When transmitting and receiving an ultrasonic beam to and from the subject using the array-type ultrasonic probe, an ultrasonic wave is incident upon setting the incident angle to the subject to a predetermined angle during ultrasonic transmission. The reception angle is set so that a predetermined position inside the subject can be received as a sound source at the time of reception, and the position of the sound source is sequentially changed so as to cover the entire flaw detection range of the subject, and flaw detection is performed. The ultrasonic waveform detected for each flaw detection with each sound source is signal-processed, and based on the obtained characteristic value, the signal detected from a defect such as a crack and the reflected signal from the weld structure are identified. The ultrasonic flaw detection method according to claim 1. As signal processing of the detected ultrasonic waveform, echo amplitude value distribution processing for the flaw detection angle at the time of reception is performed, and based on this distribution, signals detected from defects such as cracks and reflected signals from the welded part structure The ultrasonic flaw detection method according to claim 1, wherein the flaw detection method is identified. As signal processing of the detected ultrasonic waveform, fast Fourier transform processing is performed, and based on the obtained characteristic value, a signal detected from a defect such as a crack and a reflected signal from the welded structure are identified. The ultrasonic flaw detection method according to any one of claims 1 to 3. 6. The ultrasonic flaw detection method according to claim 5, wherein at least one of a half width of a spectrum distribution, a center frequency, a peak frequency, and a spectrum area is used as the characteristic value of the fast Fourier transform process. As signal processing of the detected ultrasonic waveform, wavelet transform processing is performed, and based on the obtained characteristic value, a signal detected from a defect such as a crack and a reflected signal from a welded structure are identified. The ultrasonic flaw detection method according to any one of claims 1 to 3. As signal processing of the detected ultrasonic waveform, signal processing by the maximum entropy method is performed, and based on the obtained characteristic values, signals detected from defects such as cracks and reflected signals from the weld structure are distinguished The ultrasonic flaw detection method according to claim 1, wherein the ultrasonic flaw detection method is performed. Using the center frequency, peak frequency, and half-value width of the spectral distribution as characteristic values of signal processing by the maximum entropy method, it is possible to distinguish between signals detected from defects such as cracks and reflected signals from welded structures. The ultrasonic flaw detection method according to claim 8. As signal processing of the detected ultrasonic waveform, signal processing by split spectrum method is performed, and based on the obtained characteristic value, the signal detected from defects such as cracks and the reflected signal from the welded part structure are distinguished The ultrasonic flaw detection method according to claim 1, wherein the ultrasonic flaw detection method is performed. An ultrasonic flaw detector for welds composed of dissimilar metals including the same kind of austenitic stainless steel, nickel-base superalloy, carbon steel, low alloy steel, etc. Probe installation means for installing an array-type ultrasonic probe comprising transducers, moving means for moving the array-type ultrasonic probe, shape and dimensions of the subject, material of the welded part, welding Flaw detection that sets flaw detection conditions including selecting any one or a plurality of transducers of the array-type ultrasonic probe as an ultrasonic transmission transducer group and a reception transducer group based on conditions and the like A condition setting stage, control means for converging and deflecting an ultrasonic beam on a region to be examined of the subject using the transmitting transducer group and the receiving transducer group, the transmitting transducer group and the receiving oscillation Run the child group electronically Scanning means, signal processing means for processing the ultrasonic waveform detected by the transmitting / receiving transducer group, and obtaining a characteristic value of the signal, and based on the characteristic value obtained by the signal processing means, A signal identifying means for identifying a defect signal detected from a defect such as a defect and a reflected signal from a material noise such as a welded structure, and a display for outputting a signal processing result by the signal processing means and a signal identification result by the signal identifying means And an ultrasonic flaw detector.

Images