Device and method for non-destructive testing of a test specimen using ultrasound according to the reference body method

Abstract

Device (1) for non-destructive testing of a test specimen (100) using ultrasound, wherein the device (1) is set up for characterizing defects or imperfections (99) in the material of the test specimen (100) according to the reference specimen method, wherein the device (1) comprises the following: a. an ultrasonic probe (10) with an ultrasonic transducer (12) for generating and coupling an ultrasonic field into the test object (100) at an oblique angle of incidence Θ and for recording resulting echo signals from the test object (100), b. a control unit (20) for controlling the ultrasonic probe (10) so that it generates ultrasonic pulses with a specific bandwidth B, c. a receiving unit (30) for receiving echo signals using the ultrasonic probe (10), as well as d. an evaluation unit (40) connected to the receiving unit (30), which is configured to process the recorded echo signals, wherein the evaluation unit (40) is configured to determine an angle-of-incidence DAC curve (“distance-amplitude-correction”) for the test specimen (100) from at least one echo signal from a reference reflector from a comparator body, wherein the comparator body consists of a material that is identical or acoustically equivalent to the material of the test specimen (100), characterized in that e. the evaluation unit (40) is set up to determine the angle-of-incident DAC curve for the device under test (100) on the basis of a general DAC curve by adapting it to the echo signal of the reference reflector, f. wherein the evaluation unit (40) is set up to carry out a calibration procedure for determining the angle-of-insonance specific DAC curve for the test specimen (100) by adjusting the echo signals of a plurality of identical reference reflectors, which have different distances from the ultrasonic probe (10) and are measured for a fixed angle of insonance Θ, using a least-squares method by calculating the current frequency-dependent sound attenuation in the test specimen (100) and shifting the measured values ​​in such a way that they are distributed around an adapted DAC curve with the smallest sum of the squared error distances between the adapted DAC curve and the measured values.

Description

[0001] The present invention relates to a device and a method for non-destructive testing of a test specimen using ultrasound, wherein the device and method are configured to characterize defects or discontinuities in the material of a test specimen using the reference block method. In particular, the device and method are designed to record an angle- and probe-specific DAC curve for a test specimen to be examined in order to subsequently examine it for defects or discontinuities using the reference block method.

[0002] The so-called reference block method is a method, particularly common in the USA, for the quantitative assessment of defects or discontinuities in the material of a test object. It is based on a comparison of the echo indication resulting from a defect or discontinuity with the amplitude of a reference reflector subjected to a defined acoustic beam. In practice, cross holes of a defined diameter, drilled at different depths into a reference block, are used as reference reflectors. The reference block consists of a material which, in terms of its acoustic properties, is as similar as possible to that of the test object. These acoustic properties depend, for example, on the material (e.g. the type of steel used), the manufacturing process and the thermal treatment. Before carrying out the actual test task, the reference block method must always be carried out for each desired angle of incidence orFor each probe to be used, a so-called angle and probe-specific DAC curve (DAC = "distance amplitude correction") for the test object is recorded on a suitable reference block. This curve plots the gain as a function of the time of flight. This is the gain required to bring the echo signals from identical reference reflectors, usually cross-bores, located at different depths to a predetermined height, e.g. 80% of the maximum indication height. From the decrease in the echo height of a selected reference reflector as a function of its distance from the probe, a distance-dependent, i.e. time-of-flight-dependent, amplification factor TCG ("time corrected gain") can be determined. This amplification factor can be applied so that identical cross-bores always produce the same echo indication regardless of their respective distance from the probe.This can be automatically applied to the experimentally determined echo signals when performing the test task.

[0003] The reference block method has found its way into numerous testing standards, including US testing standards published by ASME, ASTM, and AWS, all of which require the recording of a DAC curve. Further information on the reference block method can be found, for example, in the standard work Krautkrämer J. and Krautkrämer H., Materials Testing with Ultrasound, Springer Verlag, 5th edition (1986), in the reference block entry listed in the subject index.

[0004] The advantage of the reference block method is that experimental effects that could fundamentally cause uncertainties, such as frequency-dependent sound attenuation in the test object, are not influenced due to the experimentally determined DAC curve. However, the disadvantage is the enormous effort required to record the required DAC curve(s), which can be on a similar scale to the actual test task.

[0005] This disadvantage is avoided by the so-called DGS method (AVG = distance, gain, size), which was developed in Europe in the late 1950s (see, for example, KRAUTKRÄMER, J.; KRAUTKRÄMER, H.: Material Testing with Ultrasound. Berlin: Springer, 1986, pp. 98-107. - ISBN 3-540-15754-9). This method, too, is a method for the quantitative assessment of defects or discontinuities in the material of a test specimen and is based on a comparison of the echo indication resulting from a defect or discontinuity with the amplitude of a reference reflector subjected to a defined ultrasound beam. In practice, the perpendicularly scanned circular disk reflector is used, whose distance-dependent echo amplitude can be theoretically determined, at least for the far field of the ultrasonic transducer used to generate the ultrasound. In practice, at least one calibration measurement is required on a reference reflector which is inserted into a reference body and, for example,can consist of a circular arc-shaped back wall of the reference block (so-called "reference block no. 1"), is required to determine the gain factor that must be applied to bring the echo amplitude of this reference reflector to a defined reading height, e.g. 80% of the maximum reading height. This is used to calibrate a previously determined DGS curve, often stored in a memory unit of the test device used, which is specific to the probe and the selected angle of incidence. A DGS curve specifies the distance-dependent course of the gain factor that must be set to bring the echo of a reference reflector with a certain size D to the height of the back wall echo, where the size D in the case of circular disk reflectors as reference reflectors corresponds to their diameter. A DGS diagram, in turn, comprises a plurality of DGS curves for reference reflectors of different sizes D.

[0006] Within the DGS method, a detected defect or discontinuity is quantitatively characterized by comparing it with the size of a circular disc reflector, which provides a defect indication with an equivalent echo amplitude. The size of such a circular disc reflector is referred to as the "equivalent reflector size" (ERS) of the defect or discontinuity to be characterized. Theoretically, other reference reflectors, such as cross-bores, could also be used. However, this is uncommon.

[0007] Again, the defect sizes / ERG values ​​determined using the DGS method are only comparable to the actual defect size or the size of a discontinuity, as can be determined, for example, using destructive testing, under strict conditions. Nevertheless, it has found its way into numerous testing standards, such as the European testing standard EN 583-2:2001.

[0008] DE 10 2008 037 173 A1 describes a method for non-destructive testing using ultrasound, which can use a phased array probe, whereby a general DGS diagram, which was determined for a specific angle of incidence, can be converted to other electronically adjusted angles of incidence. It is further disclosed that in order to determine an ERS value of a defect using a phased array probe, which allows electronic variation of the angle of incidence into the test object, an adaptation of a known general DGS diagram, which is based on theoretical considerations, may be necessary. Such a DGS diagram can be stored in the device and calibrated for a specific probe or material using a reference echo obtained from a test specimen, which may be recorded at different angles of incidence.

[0009] The publication YAN, X, [et al.]: DGS Sizing Diagram with Single Element and Phased Array Angle Beam Probe. In: NDT.net, The Open Access NDT Database, 2010, pp. 1-9 describes software that allows the calculation of a pulse-echo response from a reflector as a function of the pulse bandwidth within the framework of a DGS method.

[0010] DE 10 2011 051 546 A1 describes a method for operating a device for non-destructive testing of a test specimen using ultrasound, wherein an amplification curve is determined from echo amplitude values ​​experimentally determined on a test specimen. DE 10 2011 051 546 A1 discloses the features of the preambles of independent claims 1 and 11.

[0011] WO 2010 / 130 819 A2 describes a probe for testing by means of ultrasound, wherein the probe is designed to generate a rotationally symmetric sound field.

[0012] LERCH, TP [et al.]: Attenuation coefficient estimation using experimental diffraction corrections with multiple interface reflections. In: Ultrasonics, 44, 2006, pp. 83-92 describes a method for determining an ultrasonic attenuation coefficient, using ratios of reflection amplitudes originating from the front-side reflection, the first back-side reflection, and the second back-side reflection from a piece with a flat plate geometry. The object of the present invention is to provide a method that allows the calibration effort required for testing a test piece using ultrasound according to the reference body method to be significantly reduced. Furthermore, the object of the invention is to provide a device configured to carry out such a method.

[0013] This object is achieved by a method according to claim 11 and a device according to claim 1. The subclaims relate to advantageous developments whose features can be combined with one another as desired within the scope of what is technically reasonable.

[0014] The method according to the invention is used for the non-destructive testing of a test object using ultrasound. The method for characterizing defects or discontinuities in the material of the test object is set up according to the reference block method. Within the scope of the method, ultrasonic pulses with a specific bandwidth B are generated and coupled into a reference block using oblique sound at a fixed beam angle θ. The reference block consists of a material that is identical to the material of the test object, or at least acoustically equivalent, i.e. has comparable acoustic properties. Furthermore, the reference block comprises at least one reference reflector of known geometry, whereby the reference reflector in the simplest case can be a semicircular back wall with a known radius. However, the reference reflector can also be a transverse bore with a known diameter.Equivalent acoustic properties are generally achieved when the materials are identical, for example, the same steel grade, and the test specimen and the reference specimen have undergone comparable processing or heat treatment. In particular, this means that the longitudinal and transverse sound velocities, as well as the material-specific sound attenuation, are identical in all propagation directions in the test specimen and the reference specimen.

[0015] In a further process step, echo signals originating from the reference reflector are recorded from the reference body. The recorded echo signals are then processed to determine an angle-specific DAC curve for the reference body and thus for the test object based on a general DAC curve. This is done by adapting the general DAC curve to the echo signal from the reference reflector.

[0016] In the context of the present invention, a general DAC curve is understood to be a DAC curve that is specific to the ultrasonic probe used to generate the ultrasonic pulses, the acoustic properties of the material of the test object or the reference object, and generally the angle of incidence 0. The following describes a method that can be used to determine such a general DAC curve on the basis of theoretical considerations. The result is a function f that describes the general DAC curve and is available at least in numerical form. Depending on the calculation method used, the function f can also be in an analytical form, although it is generally dependent on various parameters such as the speed of sound in the material of the test object or the reference object.reference body, the angle of incidence Θ, the geometric dimensions of the reference reflector, a correction factor to take into account the transfer correction and further correction factors to take into account the geometric or material-specific sound attenuation in the material of the test object or the reference body.

[0017] According to the invention, to determine the angle-specific DAC curve for the test object, a calibration method is carried out in which echo signals from a plurality of identical reference reflectors, which have different distances from the ultrasonic probe and are measured for a fixed angle of incidence, are adjusted using a least-square method by calculating a current frequency-dependent sound attenuation in the test object and shifting the measured values ​​such that they are distributed around an adjusted DAC curve with the smallest sum of the squared error distances between the adjusted DAC curve and the measured values.

[0018] The proposed procedure significantly reduces the workload for a tester performing a test task using the reference block method, as the considerable effort previously required to record a material-, beam angle-, and probe-specific DAC curve for the test object is essentially reduced to the recording of a single echo signal from a reference reflector in the reference block, in order to account for the individual sensitivity of the probe used. This is especially true when a probe is used to generate and transmit the ultrasound that allows adjustment of the beam angle 0. Electronic adjustment of the beam angle 0 is particularly possible with so-called "phased array" probes, which have a plurality of individually controllable transducer segments.Through targeted selection and phase-accurate control of the transducer segments, it is possible to electronically adjust essential ultrasound parameters such as near-field length, transmission aperture, focus position and size as well as insonification angle 0.

[0019] Prior to a specific test task, a general DAC curve is determined which takes into account the material properties of the test specimen or reference body, the angle of incidence 0, and the properties of the probe used. This general DAC curve is valid for any angle of incidence 0. It is recorded by measuring at least one, preferably at least two, and particularly preferably a plurality of identical reference reflectors at different distances from the probe at a fixed angle of incidence 0. A DAC curve specific to the selected angle of incidence is then fitted to these experimentally obtained echo signals. A general DAC curve can then be calculated from this angle-specific DAC curve. This process is also possible on personal computers with little time expenditure, given the computing power available at the time of registration.A general DAC curve theoretically determined in advance of a test task can then be transferred to the device used to perform the test task. Of course, the test device used can also be configured to determine a general DAC curve based on inputs from the tester relating to the probe used, the properties of the test object, and the angle of incidence used.

[0020] In the method according to the invention, to determine an angle-specific DAC curve for the test object at a fixed angle of incidence 0, a plurality, in particular at least two, reference reflectors are measured at different distances from the probe. The general DAC curve is then fitted to the echo signals thus obtained, which preferably originate from identically dimensioned reference reflectors. Measuring just two reference reflectors allows the material-specific sound attenuation to be taken into account. For this purpose, a factor F for the material-specific sound attenuation can be used as a fit parameter in addition to the individual sensitivity of the probe used. If more than two reference reflectors are measured, this further improves the accuracy of the fitting and thus the reliability of the DAC curve determined on the basis of the fitting. With trueDGS ®-probes, the tabulated values ​​for the beam angle-specific delay length can also be used.

[0021] If the multiple reference reflectors incorporated into the reference block are designed as transverse bores, this offers the particular advantage that various American test standards for ultrasonic testing prescribe the use of the reference block method, with transverse bores in the reference block being specified as reference reflectors. In this embodiment, the method according to the invention can therefore be advantageously used directly within the framework of the test procedures prescribed by the cited US test standards.

[0022] In principle, a particularly good agreement between the theoretically calculated DAC curves and experimental results is achieved when the frequency-dependent amplitude distribution of the ultrasonic test pulses used is used as a basis. In practice, for example, a Gaussian frequency distribution with a probe-specific bandwidth B and a probe-specific center frequency fo can be used as a basis. This means that the general DAC curve underlying the method according to the invention is generally specific to the bandwidth B of the coupled ultrasonic pulses. It follows directly from this that the determined angle-specific DAC curve is also generally specific to the bandwidth B of the coupled ultrasonic pulses.

[0023] An even higher degree of agreement with experimentally determined echo values ​​from reference reflectors can be achieved if the polarization of the ultrasonic pulses is also taken into account when determining the general DAC curve for the geometry of the selected reference reflector. This means that the general DAC curve underlying the method is specific to the polarization P of the coupled ultrasonic pulses, which directly implies that the angle-specific DAC curve is generally also dependent on the polarization of the coupled ultrasonic pulses.

[0024] In principle, the theoretical determination of the general DAC curve is simple if the calculation can be based on the assumption that the sound field of the ultrasonic pulses impinging on the reference reflector is rotationally symmetrical to the propagation direction of the ultrasonic pulses. In a particularly preferred embodiment of the device according to the invention, the device according to the invention, generally the entire assembly consisting of ultrasonic probe and control unit, is therefore designed to generate a rotationally symmetrical sound field in the test object. Corresponding ultrasonic probes, in particular those with an oblique beam, are known, for example, from WO 2010 / 130 819 A2 by the applicant, which is known under the designation trueDGS ®sold. This application reveals in particular how a one-piece transmitting transducer of an obliquely incident ultrasonic probe must be designed in order to generate a rotationally symmetrical sound field in the test object. For most testing tasks using pulse-echo technology, oblique beaming is desirable or even a prerequisite. Within the scope of the present invention, a rotationally symmetrical sound field in the test object allows a simplified calculation of the material- or transmitting transducer-specific DGS diagram. WO2010 / 130819 A1 further shows that it is also possible to generate a rotationally symmetrical sound field in the test object using phased array technology. For this purpose, for example, selected transducers of a two-dimensional, preferably but not necessarily planar array of individually controllable transducers are controlled with precise phase control. The control unit must be designed to suit this purpose.Both possibilities mentioned here for generating a rotationally symmetrical sound field in the test object are part of the subject matter of the present invention.

[0025] In a particularly preferred development of the method according to the invention, the material-specific frequency-dependent sound attenuation of the ultrasonic pulses used for the test in the test object or in the reference body is taken into account when determining the angle-specific DAC curve for the test object. For this purpose, it can be provided, for example, that a factor F for the frequency-dependent sound attenuation in the material of the test object is determined from at least two echo signals received from the reference body from identical reference reflectors that are spaced at different distances from the ultrasonic probe.

[0026] The factor F determined in this way for the frequency-dependent sound attenuation in the material of the test object can, for example, be directly taken into account in the calculation of the general DAC curve.

[0027] Conversely, this means that the general DAC curve, in its general form, includes a factor F for the frequency-dependent sound attenuation in the test object. If a practical spectral distribution of the transmitted pulses used is assumed, for example, in the form of a Gaussian function, it is possible to determine this frequency-dependent factor F by fitting the DAC curve to the experimentally determined reference echoes by measuring a limited number of identical reference reflectors at different distances from the probe.

[0028] An alternative approach is to determine the factor F for the frequency-dependent sound attenuation in the test specimen using an independent method, whereby reference is again made to the applicant's application of the same date, already referred to above, entitled "Method and device for non-destructive testing of a test specimen by means of ultrasound taking into account the frequency-dependent sound attenuation".Once the factor F has been determined, which can be done, for example, by measuring two identical reference reflectors at different distances, the influence of the material-specific frequency-dependent sound attenuation in the material of the test object can be compensated for both when carrying out the method according to the invention and when carrying out the specific test task itself by calculating corrected spectra from the spectrum of the received reference echoes using the factor F, in which the influence of the material-specific frequency-dependent sound attenuation is precisely eliminated. The corrected spectra obtained in this way can then be transformed back into the time domain, resulting in corrected echo signals in which the influence of the material-specific frequency-dependent sound attenuation is no longer included.These corrected echo signals can then be used, on the one hand, in the implementation of the method according to the invention, so that consideration of the material-specific frequency-dependent attenuation factor F is no longer necessary when determining the general DAC curve. On the other hand, the flaw echoes detected during the execution of the test task can be corrected in the same way, so that the influence of the material-specific frequency-dependent sound attenuation is precisely compensated.

[0029] A device according to the invention is provided for the non-destructive testing of a test specimen using ultrasound, wherein the device is configured to characterize defects or discontinuities in the material of the test specimen using the reference block method. The device comprises an ultrasonic probe with an ultrasonic transducer for generating and coupling an ultrasonic field into the test specimen using oblique sound at an incidence angle Θ. Furthermore, the ultrasonic probe is configured to record resulting echo signals from the test specimen. If necessary, a receiving probe designed separately from the transmitting probe and preferably with identical ultrasonic properties can be used here. Furthermore, the device comprises a control unit for controlling the ultrasonic probe so that it generates ultrasonic pulses with a specific bandwidth B.

[0030] Finally, the device according to the invention comprises a receiving unit for recording echo signals using the ultrasonic probe. Furthermore, an evaluation unit connected to the receiving unit is provided, which is configured to process the recorded echo signals. In particular, the evaluation unit is configured to determine a DAC curve for the test object from at least one echo signal of a reference reflector from a reference body. The reference body used is made of a material that is identical or acoustically equivalent to the material of the test object.

[0031] According to the invention, the evaluation unit is now configured to determine an angle-specific DAC curve for the test object by adapting the general DAC curve to the echo signal of the reference reflector on the basis of a general DAC curve, which can be specific to the test head used (transducer geometry, angle of incidence Θ, center frequency fo, bandwidth B) and the material of the test object and which can be stored in particular in a memory unit of the evaluation unit.

[0032] The evaluation unit is set up to determine the angle-specific DAC curve for the test object by carrying out a calibration procedure by adapting the echo signals of a plurality of identical reference reflectors which have different distances from the ultrasonic probe and are measured for a fixed angle of incidence, using a least-square method by calculating the current frequency-dependent sound attenuation in the test object and shifting the measured values ​​in such a way that they are distributed around an adapted DAC curve with the smallest sum of the squared error distances between the adapted DAC curve and the measured values.

[0033] In advantageous developments, the device according to the invention, in particular the evaluation unit included therein, is configured to carry out the advantageous developments of the method according to the invention described above. Reference is made to the relevant explanations.

[0034] Initially, it was explained that DAC curves in the context of the reference block method mentioned above generally refer to transverse bores with a diameter of Dz. Regarding DGS curves, it was explained in the context of the DGS method mentioned above that these refer to circular disc reflectors with a diameter of D r as reference reflectors. It can be shown that the DGS curve of a circular disc reflector with diameter D r in the far field is equidistant from the backwall echo curve as the DAC curve of a cross hole with diameter D z if the following condition is met: Dr=Ds22λNnum22πzDz with: Ä: Wavelength ultrasound D s : Diameter of the circular ultrasonic transducer used to generate the ultrasound Nnum: numerical value of the near-field length of the circular ultrasonic transducer used to generate the ultrasound.

[0035] Of course, all features of the above-described embodiments or further developments of the invention can be combined with one another as desired, as far as this is technically possible. This also applies to the features of the following exemplary embodiment, from which further features and advantages of the device according to the invention and the method according to the invention emerge. The exemplary embodiment serves to illustrate the invention to the person skilled in the art and is therefore to be understood as exemplary and not restrictive. It refers to the figures, which show the following: Fig. 1: an embodiment of a device according to the invention, Fig. 2: the sound pressure of the sound field used perpendicular to the acoustic axis, Fig. 3: a calculated DGS curve for a transverse bore taking into account the transverse distribution, Fig. 4: a screen curve determined from this, taking into account the cross-distribution, Fig. 5: an AVG curve without taking the cross distribution into account, Fig. 6: a screen curve determined from this without taking the cross-distribution into account, Fig. 7: an exemplary evaluation of experimentally determined echo values, Fig. 8: Comparison of two DGS curves without / with sound attenuation correction, Fig. 9: a screen curve with sound attenuation correction, Fig. 10: a theoretical sound pressure curve for a cross hole, Fig. 11: a first experimental example of fitting a theoretical DAC curve to experimental echo values ​​using the “Least Square” method, Fig. 12: a second experimental example of fitting a theoretical DAC curve to experimental echo values ​​using the least square method, Fig. 13: a first experimental example of screen curves for different angles of incidence determined from reference measurements at a single angle of incidence Θ = 53°, Fig. 14: a second experimental example of screen curves for different angles of incidence determined from reference measurements at a single angle of incidence Θ = 53°, Fig. 15: Screen curves for all angles at the same gain setting for the experimental example according to Fig. 14, and Fig. 16: Screen curves for all angles at the same gain setting for the experimental example according to Fig. 15.

[0036] It should be noted that in the context of the embodiments, the terms AVG curve / diagram and DAC curve / diagram are used equivalently. Fig. 1 shows an embodiment of a testing device 1 according to the invention for non-destructive testing of a test specimen 100 using ultrasound. The testing device 1 is configured to characterize defects or discontinuities 99 in the material of the test specimen 100 using the reference block method, which may be located in particular in the near field of an ultrasonic transducer 14 used to generate the ultrasound. The testing device 1 comprises an ultrasonic probe 10 with a one-piece ultrasonic transducer 12 for generating and coupling an ultrasonic field into the test specimen 100 and for recording resulting echo signals from the test specimen 100. The probe 10 is configured to project an ultrasound beam into the test specimen 100 at an angle. For this purpose, the ultrasonic transducer 12 is arranged on a wedge-shaped guide body 14.

[0037] Furthermore, the test device 1 comprises a control unit 20 for controlling the ultrasonic probe 10, so that it generates a sequence of ultrasonic pulses with a typical center frequency f0 between 1 and 5 MHz and a bandwidth B that is typically between 20 and 40%. The pulse repetition frequency is typically in the range of a few kHz. The control unit 20 is connected to the probe 10 and, in particular, to its ultrasonic transducer 12.

[0038] Furthermore, a receiving unit 30 is provided for receiving echo signals using the ultrasonic probe 10. The receiving unit 30 is also connected to the probe 10 and in particular to its ultrasonic transducer 12.

[0039] Finally, an evaluation unit 40 is provided, which is connected to both the control unit 20 and the receiving unit 30 and is configured to process the echo signals from the material of the test object 100, which are recorded by the ultrasonic transducer 12 of the test head 10. The evaluation unit 40 is connected to a display device 42 in the form of an LCD or OLED, on which, for example, the amplitude of the received echo signals can be displayed in a time-resolved manner (A-scan), in particular using the DAC curve determined according to the invention, ie the runtime-dependent amplification factor determined according to the invention.

[0040] The evaluation unit 40 is configured to carry out the method according to the invention in various forms, which will be discussed in more detail below. The method can be implemented in the control unit 20, the evaluation unit 40, or a higher-level control unit that is part of the device 1.

[0041] Control unit 20, receiving unit 30 and evaluation unit 40 including display device 42 are housed in a common ultrasonic control unit 60, which is connected to the test head 10 via a communication line 50.

[0042] In an alternative embodiment, the control unit 20, the receiving unit 30 and the evaluation unit 40 can be integrated individually or jointly and partially or completely in the test head 10.

[0043] The ultrasonic probe 10 is constructed according to WO 2010 / 130 819 A2, so that it generates a sound field in the test object 100 that is rotationally symmetrical to the acoustic axis. This generally requires an approximately circular and non-planar ultrasonic transducer, whose "diameter" is referred to below as D. In such a case, for example, in the context of the present invention, the diameter D of the transmitting transducer can be assumed to be the diameter of a circular transmitting transducer of the same area.

[0044] The evaluation unit 40 is configured to normalize the amplitude of echo signals received from the test object relative to the echo amplitude of at least one selected reference reflector based on a DAC curve determined using a suitable calibration procedure. The echo amplitude of this reference reflector is typically determined using a reference body separate from the test object.

[0045] One possible method for the theoretical determination of the sound pressure on the acoustic axis for impulse sound is based on considerations of the sound pressure of a circular transmitting transducer with a diameter D for continuous sound. For the sake of simplicity, reference is made here to the considerations contained in the German patent application filed on the same day by the same applicant entitled "Device and method for non-destructive testing of a test specimen using ultrasound according to the DGS method," in particular to equations 1 to 25 therein, which form the basis for calculating the DGS diagram for both continuous sound and impulse sound of a circular disk reflector. By this reference, the content of the referenced patent application is incorporated in its entirety into the disclosure of the present application. It is necessary to consider the different geometry of the reflector (transverse bore instead of circular disk).

[0046] To calculate an AVG curve for transverse bores, the sound pressure distribution perpendicular to the acoustic axis must first be calculated for each sound path. Fig. Figure 2 shows, as an example, the sound pressure distribution of a 2 MHz probe with a sound path of 50 mm. Let p(z, y) be this sound pressure distribution at a sound path of z and a distance of y from the acoustic axis. To calculate the DGS curve, equation (25) of the referenced patent application "Device and method for non-destructive testing of a test specimen using ultrasound according to the DGS method" must then be modified according to the transverse distribution and the distance law for transverse bores.

[0047] The decrease in sound pressure from a circular disk as a function of the sound path z follows the following law: 1z

[0048] For cross holes, however, the following applies [see Krautkrämer J., Krautkrämer H., Material Testing with Ultrasound, Springer Verlag, 5th edition (1986), page 50] 1z

[0049] With these conditions, the calculation of an AVG curve for a transverse bore is as follows: pquer(z,t)=∫x=0D2∫y=0D2p¯(z,y)eA(2z−4z2+x2)2x4z2+x24cos(ωt−k4z2+x2)dx dy

[0050] In Fig. Figure 3 shows the result of this first calculation approach, which takes into account the transverse distribution of the sound pressure along the longitudinal axis of the transverse bore. Accordingly, a screen curve for the ultrasound device can be calculated, which is shown in Fig. 4 is shown.

[0051] Alternatively, based on the calculation for circular disc reflectors, the DGS curve is calculated without taking into account the transverse distribution of the sound pressure: pquer(zt)=p¯A(z)∫x=0D2∫y=0D2eA(2z−4z2+x2)x4z2+x24cos(ωt−k4z2+x2)dx dy

[0052] If necessary, Deff = 0.97 D is used. This DV curve is shown in Fig. 5. Here again, a screen curve for the ultrasound device can be calculated, which is shown in Fig. 6. It can be seen that even the simplified latter calculation leads to sufficient accuracy. Therefore, the following evaluation is based on the calculation without cross-distribution.

[0053] The evaluation unit 40 is configured to perform a first calibration procedure, which is applicable if only one DGS curve (e.g., for a specific angle of incidence) is required. This first calibration procedure comprises the following steps: • Acoustic sounding of two identical cross-bores with as different sound paths as possible • Determine the sound paths and the dB value to place these echoes at 80% screen height • Calculation of the difference ΔG of the two dB values

[0054] First, the corresponding point on the DGS curve is entered for the sound path s1 of the first measurement, see the circle marker in the Fig. 7.

[0055] In a further step, a point is entered into the diagram for the second sound path s2 at a distance of 1XG from the first measured value, which is indicated by the plus marker in the Fig. 7. For the sound path s2, the distance of this point to the curve is determined, from which the gain difference ΔG 12 (distance between the X marker and the plus marker). In this example, ΔG 12 a value of 2 dB.

[0056] The concrete evaluation of this example is carried out with the following values, which are also Fig. 7 can be taken: • First sound path s1 = 10.9 mm • Second sound path s2 = 79 mm • Gain difference Difference ΔG 12 = 2 dB

[0057] This now allows the sound attenuation s a calculate: sa=ΔG1210002(s2−s1)sa=2dB10002(79mm−10.9mm)=14.7dBm with s1 and s2 in mm and ΔG 12 in dB. With the sound attenuation determined in this way, both the DGS curve (cf. Fig. 8) as well as the screen curve (cf. Fig. 9) can be adapted accordingly. With this procedure or its implementation in the evaluation unit 40 of the test device 1, the recording of a DAC curve using a plurality of transverse bores, which is common in the prior art and prescribed by relevant test standards, is reduced to the recording of the echo signals from only two identical reference reflectors, which are spaced at different distances from the test head 10 and are designed here, for example, as transverse bores in a reference body.

[0058] The evaluation unit 40 is further configured to perform a second calibration method based on recording only one reference echo from the circular arc of a reference body of type No. 1. Using this second calibration method, it is particularly possible to determine angle-specific DAC curves for a phased array with only one calibration measurement if its angle-dependent amplitude correction values ​​are known for all relevant angles.

[0059] In this case, the distance v quer from the DGS curve for the cross-hole used for calibration to the calculated backwall echo curve. The following considerations in this section all refer to the far field (very large z).

[0060] According to Krautkrämer J., Krautkrämer H., Material Testing with Ultrasound, Springer Verlag, 5th edition (1986), p. 75, the sound pressure p on the acoustic axis of a circular oscillator is calculated according to the formula: p≈p0πDs24λz with D s : Diameter of the circular oscillator.

[0061] The sound pressure of a transverse bore in the far field follows the distance law without proportionality factors: Dzz with D z : Diameter of the cross hole.

[0062] The following proportionality factors are added: 2Dz2λz

[0063] This gives the resulting sound pressure p quer a transverse bore in the sound field of a circular oscillator from the product of equations (6) and (8): pquer=P0π2Ds2Dz8λ2z32

[0064] Fig. 10 shows the approximation of the cross-bore curve by p quer .

[0065] For the flat rear wall, the sound pressure p in the far field is R : pR=p0πNnum2z

[0066] The ratio of p quer to p R corresponds to the respective echo amplitudes H: HquerHR=2Ds2Dz4λ2Nnumz

[0067] This gives us the displacement v quer between backwall echo curve and cross hole curve in the DGS diagram in the far field: vbar=20 lg(2Ds2Dz4λ2Nnumz)

[0068] If the frequency-dependent sound attenuation is also determined using a suitable method, the measurement of a plurality of reference reflectors, e.g. in the form of cross-bores, can be completely dispensed with in order to create a DAC curve.

[0069] The evaluation unit 40 is further configured to calculate DGS curves for all desired angles of incidence Θ on the basis of the thus determined angle-specific DGS curve on the basis of the known angle dependence of the phased array probe 10 used.

[0070] Finally, the evaluation unit 40 is configured to validate the DAC curve(s) thus determined using experimentally determined echo amplitudes of a plurality of identical reference reflectors that are spaced differently from the test head.

[0071] Furthermore, the evaluation unit 40 is configured to perform a third calibration procedure, which allows the recording of a reference echo to be completely dispensed with. Within the scope of this third calibration procedure, an echo from a known reflector is randomly selected and mathematically shifted onto the curve calculated for the selected insonification angle.

[0072] Finally, the evaluation unit 40 is configured to perform a fourth calibration method, which is referred to below as the "Least Square" method. Within the scope of this method, a plurality of identical transverse bores, which are spaced differently from the test head 10, are measured for a fixed angle of incidence Θ. These measured values ​​can be adjusted to the calculated transverse bore curve using the Least Square method. The current frequency-dependent sound attenuation in the test body is calculated, and the measured values ​​are shifted so that they are distributed around the curve with the smallest square error distances. The result of this method is shown in the Fig. 11 and Fig. 12 shown as an example.

[0073] The following values ​​are used for the calculation according to the least square method: • n: Number of measured values • si: Measured sound paths • ri: the dB values ​​for the measured sound paths on the calculated cross-bore curve • mi: measured amplitude values • a: be the sound attenuation • v: be the necessary displacement in the y-axis

[0074] The sum f of the squared error distances is then calculated as: f=∑i=1n[ri−(mi+2asi)+v]2

[0075] The two partial derivatives are calculated and set equal to zero: ∂f∂v=0∂f∂a=0

[0076] This results in the following system of equations for calculating the necessary displacement v and the sound attenuation a: (2∑i−1nsin2∑i=1nsi2∑i=1nsi)(av)=(∑i−1n(ri−mi)∑i=1nsi(ri−mi))

[0077] The following example shows the result of the "least square" method. Seven cross-hole echoes were initially recorded at an angle of 53° (close to the nominal angle) using both a 2 MHz and a 4 MHz phased array probe. In the next step, the cross-hole curves for the other angles were calculated and shifted in the DGS diagram by the respective angle-dependent sensitivity differences, which are known for the trueDGS® phased array probes used (cf. WO 2010 / 130 819 A2). Fig. 13 and Fig. 14 shows the results of this calculation including the measured values ​​for all angles for validation.

[0078] The Fig. 15 and Fig. 16 shows, with the same gain setting of the ultrasound device, both the screen curve for the base measurement at 53° and the resulting screen curves for the other insonification angles for the 2 MHz and the 4 MHz calculation.

[0079] A particularly preferred testing method based on the method according to the invention and, if applicable, the device according to the invention using a phased array probe 10 with adjustable angle of incidence Θ comprises the following method steps: The tester measures a plurality of identical transverse bores in a reference body, which are spaced differently from the probe, as when using “conventional” single-element probes for a fixed angle of incidence Θ. As soon as two echoes have been recorded, the calculated screen curve can be displayed. The further measurements serve to improve the accuracy. The accuracy of the

[0080] Measurements can be directly assessed by the position of the measured values ​​relative to the calculated screen curve displayed. The other screen curves are then calculated using the least squares method as described above. The advantage of this method lies in its high accuracy and in the fact that inspectors can proceed with a conventional angle-beam probe with a fixed beam angle, as is usual with the reference block method. Retraining to use phased array probes is therefore unnecessary. Reference symbol 1 test device 10 test head 12 ultrasonic transducers 14 flow bodies 20 control unit 30 receiving unit 40 Evaluation unit 42 Display device 44 Information display field 50 communication lines 60 Ultrasonic control unit 99 defects, imperfections 100 candidates

Claims

[1] Device (1) for non-destructive testing of a test piece (100) by means of ultrasound, wherein the device (1) is designed to characterise defects or discontinuities (99) in the material of the test piece (100) according to the reference body method, wherein the device (1) comprises the following: an ultrasonic probe (10) with an ultrasonic transducer (12) for generating and coupling an ultrasonic field into the test object (100) in oblique sound at an angle of incidence Θ and for recording resulting echo signals from the test object (100), b. a control unit (20) for controlling the ultrasonic probe (10) so that it generates ultrasonic pulses with a specific bandwidth B, c. a receiving unit (30) for receiving echo signals by means of the ultrasonic probe (10), and d. an evaluation unit (40) connected to the receiving unit (30) and configured to process the recorded echo signals, wherein the evaluation unit (40) is configured to determine an angle-specific DAC curve (“distance-amplitude-correction”) for the test object (100) from at least one echo signal of a reference reflector from a reference body, wherein the reference body consists of a material that is identical or acoustically equivalent to the material of the test object (100), characterized by , that e. the evaluation unit (40) is designed to determine the angle-specific DAC curve for the test object (100) on the basis of a general DAC curve by adaptation to the echo signal of the reference reflector, f. wherein the evaluation unit (40) is configured to carry out a calibration method for determining the insonification angle-specific DAC curve for the test object (100) by adapting the echo signals of a plurality of identical reference reflectors which are at different distances from the ultrasonic test head (10) and are measured for a fixed insonification angle Θ, using a least-square method by calculating the current frequency-dependent sound attenuation in the test object (100) and shifting the measured values ​​in such a way that they are distributed around an adapted DAC curve with the smallest sum of the squared error distances between the adapted DAC curve and the measured values. [2] Device (1) according to claim 1, characterized by that the evaluation unit (40) is designed to determine a curve for a different angle of incidence Θ from the angle-specific DAC curve for the test object (100). [3] Device (1) according to one of the preceding claims, characterized by that the device (1) is designed to vary the angle of incidence Θ, preferably electronically. [4] Device according to claim 3, characterized by that the evaluation unit (40) is further configured to take into account the influence of any change in a delay line in the test head (10) upon a change in the insonification angle Θ on the insonification angle-specific DAC curve. [5] Device (1) according to claim 1, characterized by that the evaluation unit (40) is designed to base the adaptation on the echo signals of a reference reflector designed as a transverse bore. [6] Device (1) according to claim 1, characterized by that the DAC curve is specific for bandwidth B. [7] Device (1) according to claim 1, characterized bythat the ultrasonic pulses have a certain polarization P and the DAC curve is specific for this polarization P. [8] Device (1) according to claim 1, characterized by that the test head (10) is designed to generate a rotationally symmetrical sound field in the test object (100), if necessary in cooperation with the control unit (20). [9] Device (1) according to claim 1, characterized by in that the evaluation unit (40) is designed to determine a factor F for the frequency-dependent sound attenuation in the material of the test object (100) from at least two echo signals received from the comparison body from identical reference reflectors which are spaced differently from the ultrasonic test head (10). [10] Device (1) according to claim 1, characterized bythat the evaluation unit (40) is designed to determine a corrected echo signal from the received echo signal of the reference reflector, in which the influence of the frequency-dependent sound attenuation in the material of the test object (100) is compensated. [11] Method for the non-destructive testing of a test specimen (100) by means of ultrasound, wherein the method is designed to characterise defects or discontinuities (99) in the material of the test specimen (100) according to the reference body method, wherein the method comprises the following method steps: a. generating and coupling ultrasonic pulses with a specific bandwidth B into a reference body in oblique sound at an angle of incidence Θ, wherein the reference body consists of a material which is identical or acoustically equivalent to the material of the test object (100) and comprises at least one reference reflector of known geometry, b. Recording echo signals from the reference body originating from the reference reflector, c. processing the recorded echo signals in such a way that an angle-specific DAC curve is determined for the test object (100), characterized by , that d. the angle-specific DAC curve for the test object (100) is determined on the basis of a general DAC curve by adaptation to the echo signal of the reference reflector, e. wherein, to determine the angle-specific DAC curve for the test object (100), a calibration method is carried out in which echo signals from a plurality of identical reference reflectors, which have different distances from the ultrasonic probe (10) and are measured for a fixed angle of incidence Θ, are adjusted using a least-square method by calculating a current frequency-dependent sound attenuation in the test object (100) and shifting the measured values ​​such that they are distributed around an adjusted DAC curve with the smallest sum of the squared error distances between the adjusted DAC curve and the measured values. [12] Method according to claim 11, characterized by that a DAC curve for a different angle of incidence Θ is determined from the angle-specific DAC curve. [13] Method according to claim 12, characterized bythat when determining an insonification angle-specific DAC curve, the influence of any change in a delay line in a probe (10) used to generate and couple the ultrasonic pulses is taken into account when the insonification angle Θ changes. [14] Method according to claim 12, characterized by that the reference reflector is designed as a transverse bore. [15] Method according to claim 11, characterized by that the DAC curve is specific for bandwidth B. [16] Method according to claim 11, characterized by that the DAC curve is specific for the polarization P of the coupled ultrasound pulses. [17] Method according to claim 11, characterized by that the sound field of the coupled ultrasonic pulses in the test object (100) is rotationally symmetrical. [18] Method according to claim 11, characterized bythat a factor F for the frequency-dependent sound attenuation in the material of the test object (100) is determined from at least two echo signals received from the reference body of identical reference reflectors which are spaced differently from the ultrasonic test head (10). [19] Method according to claim 11, characterized by that a corrected echo signal is determined from the received echo signal of the reference reflector, in which the influence of the frequency-dependent sound attenuation in the material of the test object (100) is compensated.

Images