Method for non-destructive ultrasonic inspection of a longitudinal metal bar

Abstract

The method comprises inspecting a longitudinal end (3) of a metal bar (1), with a step of acquiring data using at least one multi-element probe (5), and a step of post-processing the data; during the data acquisition step, moving the at least one multi-element probe (5) and the metal bar (1) relative to one another such that the at least one multi-element probe (5) sweeps a side surface (9) circumferentially and / or axially, the at least one multi-element probe (5) emitting, at regularly spaced defined positions, an incident ultrasonic signal towards a region of the longitudinal end (3) and detecting the returned signal, the returned signal being recorded with the corresponding circumferential and / or axial position of the multi-element probe (5) with respect to the side surface (9); the incident ultrasonic signal comprising multiple successively emitted incident signals that are distinct from one another, the returned signal comprising a matrix of elementary time-domain signals corresponding to the echoes of each of the incident signals detected by each receiving element (7); during the data post-processing step, the returned signal is post-processed by determining, in the elementary time-domain signals, the echoes returned by each point of the region of the longitudinal end (3) and by combining the echoes into a combined signal that characterizes the point.

Description

[0001] METHOD FOR NON-DESTRUCTIVE ULTRASONIC TESTING OF A LONGITUDINAL METAL BAR

[0002] The present invention relates in general to the non-destructive ultrasonic testing of a longitudinal metal bar, and more particularly to the testing of the longitudinal ends of the bar.

[0003] Some metal bars must be inspected over 100% of their volume, in order to detect possible indications such as cracks, porosities, decohesions on inclusions, shrinkage bottoms, large gains, segregations, or even macro-zones. Customers can request a map of the metal bar, showing the position of these possible indications within the metal bar.

[0004] The control of bars made of a titanium alloy such as Ti-6AI-4V is particularly delicate when the bars have large sections.

[0005] In fact, alloys of this type are not very permeable to ultrasound, so it is necessary to use a high-energy incident ultrasonic signal. This difficulty is further increased when the bars are of large sections, since it is necessary to be able to detect even the indications located in the center of each section.

[0006] A large section is understood here to mean a diameter greater than 100 mm.

[0007] Controlling the longitudinal ends of such a metal bar poses particular problems.

[0008] It is possible to check the longitudinal ends of the bar using conventional equipment equipped with single-element probes operating at 2 or 4 MHz. These probes are not associated with a displacement mechanism allowing recording of indication maps, i.e. locating the geometric coordinates of the indication within the bar.

[0009] Furthermore, the use of conventional probes generates cliff waves from the cutting faces of the bar, i.e. the end faces of the bar, so that the interpretation of the signals, and therefore the search for defects, is extremely delicate over a length of approximately 20 mm at the two opposite longitudinal ends of the bar.

[0010] It is also possible to use multi-element probes comprising several transmitter / receiver elements aligned longitudinally. These multi-element probes can be associated with a displacement device allowing the probe to be referenced, and therefore mapping to be carried out.

[0011] Such a multi-element probe can be used to generate an incident ultrasonic signal forming an angle of incidence of 0° relative to the lateral surface of the metal bar. The preferential elongation of the alloy grains, called fibering, is generally parallel to the lateral surface of the metal bar due to the shaping process applied to said bar. The incident ultrasonic signal is therefore substantially perpendicular to the fibering, which is optimal for detecting indications. At the longitudinal ends of the metal bar, the multi-element probe is positioned in the immediate vicinity of the end surface of the bar, or even straddling the peripheral edge of the bar. As a result, the amount of energy sent into the bar is reduced, such that the detection sensitivity is degraded and the probed depth is reduced.

[0012] It is also possible to use the multi-element probe by positioning it slightly away from the end face of the bar, and using a focal law to orient the incident ultrasonic signal towards the end face of the bar. This keeps all the transmitter / receiver elements in contact with the lateral surface of the metal bar, and therefore sends a greater quantity of energy into the material. However, due to the orientation of the incident ultrasonic signal, successive reflections occur on the end surface and on the lateral surface of the bar, which causes energy losses. The returned signal is thus attenuated. Furthermore, it is also disturbed due to edge effects.

[0013] In both cases, the energy losses introduced make these methods unsuitable for controlling the end of a large diameter bar made of a titanium alloy such as TÎ-6AI-4V.

[0014] Therefore, for titanium alloy bars with large cross-sections, in some cases it is necessary to systematically cut both longitudinal ends of the metal bar, for example to a length of 20 mm on each side. This allows us to provide the customer with a metal bar with reliable control over 100% of the bar length, and a mapping of indications over 100% of the bar length.

[0015] Such cutting generates costly material losses. The manufacturing process includes an additional sawing step, so the manufacturing process is longer and more expensive.

[0016] In this context, the invention aims to propose a method for non-destructive ultrasonic testing of a longitudinal metal bar, which allows the testing of the longitudinal ends of the metal bar without exhibiting the above defects. To this end, the invention relates to a method for non-destructive ultrasonic testing of a longitudinal metal bar, the method comprising testing a longitudinal end of the metal bar delimited by a lateral surface with a closed contour and having a central axis, the testing method comprising a step of acquiring data using at least one multi-element probe comprising several emitting elements and several receiving elements, and a step of post-processing the data;during the data acquisition step, moving the at least one multi-element probe and the metal bar relative to each other such that the at least one multi-element probe circumferentially and / or axially scans said lateral surface, the at least one multi-element probe emitting at regularly spaced determined positions an incident ultrasonic signal towards a region of said longitudinal end and detecting the returned signal, the returned signal being recorded with the corresponding circumferential and / or axial position of the multi-element probe relative to the lateral surface; the incident ultrasonic signal comprising several incident signals emitted successively, distinct from each other; the returned signal comprising a matrix of elementary time signals corresponding to the echoes of each of the incident signals detected by each receiving element;during the data post-processing step, the returned signal is post-processed by determining in the elementary time signals the echoes returned by each point of said longitudinal end region and by combining said echoes into a combined signal characterizing said point.;

[0017] During the data acquisition step, the multi-element probe is placed in direct or indirect contact with the lateral surface of the metal bar, with all the transmitter / receiver elements in direct or indirect contact with this lateral surface. The amount of energy sent into the metal bar is sufficient to control large diameter bars, even if these bars are made of materials with low permeability to ultrasound, such as titanium alloys.

[0018] Collecting a matrix of elementary time signals and performing post-processing of these elementary time signals makes it possible to overcome edge effects and disturbing signals observed in particular when the multi-element probe is used with a non-zero angle of incidence, as described above. The joint recording of the circumferential and / or axial position of the multi-element probe with the returned signal makes it possible to establish maps and to record these maps.

[0019] The control method may further represent one or more of the following characteristics, considered individually or in all technically possible combinations:

[0020] - the data acquisition step is carried out according to the FMC method;

[0021] - the incident ultrasonic signal comprises several incident signals emitted successively, each by a different emitting element;

[0022] - the data acquisition step is carried out using the PWI method;

[0023] - the incident ultrasonic signal comprises several incident signals emitted successively, each incident signal being emitted by all the emitting elements jointly, each incident signal forming a plane wave, the incident signals having angles of incidence different from each other;

[0024] - the data post-processing step is carried out using the TFM method;

[0025] - the elementary time signals each have the equation a(t) = 4(t).sin 4>(t), the echoes returned by each point of the region of said longitudinal end being combined by adding the amplitudes A(t) to form the combined signal characterizing said point;

[0026] - the data post-processing step is carried out according to the PCI method;

[0027] - the elementary time signals each have the equation a(t) = 4(t).sin 4>(t), the echoes returned by each point of the region of said longitudinal end (3) being combined by adding the phases <t>(t) to form the combined signal characterizing said point;

[0028] - the metal bar is made of titanium or a titanium alloy, for example Ti-6AI-4V or Ti-5553 or Ti 10-2-3, or martensitic steel, or maraging steel, or austenitic steel, or a nickel or cobalt-based superalloy, or an aluminum-based alloy;

[0029] - during the data acquisition step, the at least one multi-element probe scans circumferentially and / or axially the lateral surface such that the entire volume of the longitudinal end of the metal bar is controlled;

[0030] - the method comprises a step of obtaining a map of the longitudinal end, showing the combined signal at each point of the longitudinal end.

[0031] According to a second aspect, the invention relates to an assembly for non-destructive ultrasonic testing of a longitudinal metal bar, the assembly being designed for testing a longitudinal end of the metal bar delimited by a lateral surface with a closed contour and having a central axis, the testing assembly comprising a data acquisition device and a data post-processing device;the data acquisition device comprising at least one multi-element probe comprising several emitting elements and several receiving elements, a mechanical device configured to move the at least one multi-element probe and the metal bar relative to each other such that the at least one multi-element probe scans circumferentially and / or axially the lateral surface, and a controller configured so that the at least one multi-element probe emits at regularly spaced determined positions an incident ultrasonic signal towards a region of said longitudinal end and detects the returned signal, the controller being configured so that the returned signal is recorded with the corresponding circumferential and / or axial position of the multi-element probe relative to the lateral surface; the incident ultrasonic signal comprising several incident signals emitted successively, distinct from each other;the returned signal comprising a matrix of elementary time signals corresponding to the echoes of each of the incident signals detected by each receiving element; the data post-processing device being configured so that the returned signal is post-processed by determining in the elementary time signals the echoes returned by each point of said region of the longitudinal end and by combining said echoes into a combined signal characterizing said point.;

[0032] The control set may further represent one or more of the following characteristics, considered individually or in any technically possible combination:

[0033] - the controller is configured so that data acquisition is performed using the FMC method or the PWI method;

[0034] - the elementary time signals each have the equation the data post-processing device being configured so that the echoes returned by each point of the region of said longitudinal end are combined by adding the amplitudes A(t) to form the combined signal characterizing said point;

[0035] - the elementary time signals each have the equation the data post-processing device being configured so that the echoes returned by each point of the region of said longitudinal end are combined by adding the phases <t>(t) to form the combined signal characterizing said point. Other characteristics and advantages of the invention will emerge from the detailed description given below, for information purposes only and in no way limiting, with reference to the appended figures, among which:

[0036] - Figure 1 is a simplified schematic representation of the control assembly of the invention, placed on a longitudinal metal bar;

[0037] - Figure 2 is a simplified schematic representation of a data acquisition step carried out according to the FMC method;

[0038] - Figure 3 is a simplified schematic representation of a data acquisition step carried out according to the PWI method;

[0039] - Figure 4 is a schematic representation of an elementary time signal collected by a transmitter / receiver element;

[0040] - Figure 5 is a side view of a standard used for testing;

[0041] - Figure 6 is a representation of the signals collected for a static test with the standard of Figure 5, for post-processing of the data according to the TFM method;

[0042] - Figure 7 is a representation of the signals collected for the same static test as in Figure 6, for post-processing of the data according to the PCI method;

[0043] - Figures 8 and 9 are representations of two other standards, used in particular for a calibration phase of the non-destructive ultrasonic testing process; and

[0044] - Figures 10 and 11 illustrate the signals collected during a dynamic test carried out with the standard of Figure 5.

[0045] As illustrated in Figure 1, the method aims to non-destructively control, by ultrasound, a longitudinal metal bar 1.

[0046] The longitudinal direction corresponds to the direction of extension of the metal bar 1.

[0047] The metal bar 1 is typically one of the materials in the following list: titanium; titanium-based alloy; martensitic steel; maraging steel; austenitic steel; nickel- or cobalt-based superalloy; aluminum, aluminum-based alloy.

[0048] For example, the bar is made of a titanium alloy of grade Ti-6AI-4V, or Ti-5553, or even Ti 10-2-3.

[0049] Here, a bar is understood to mean a piece of elongated shape in the longitudinal direction, and having, perpendicular to the longitudinal direction, sections smaller than the dimension of the bar longitudinally.

[0050] Typically, the metal bar 1 is rectilinear. Alternatively, it has a non-rectilinear central line, for example comprising portions oblique to the longitudinal or curved direction. The metal bar 1 is solid or hollow. When it is hollow, it therefore has a conduit shape.

[0051] The method provides in particular for checking a longitudinal end 3 of the metal bar 1.

[0052] The method typically involves checking both opposite longitudinal ends of the metal bar, carried out in the same way.

[0053] The method preferably also provides for the control of the central part of the bar 1, over its entire length between the two longitudinal ends.

[0054] The longitudinal end 3 of the metal bar 1 has, perpendicular to the longitudinal direction, a round, oval, elliptical, or axisymmetric cross-section, or a polygonal shape, for example square, hexagonal, octagonal, or any other shape.

[0055] Typically, the section orthogonal to the longitudinal direction is constant along the metal bar 1. Alternatively, it is not constant.

[0056] The largest dimension of the cross-section of the longitudinal end is greater than or equal to 100 mm, typically greater than or equal to 150 mm, or even greater than or equal to 200 mm.

[0057] It is typically less than 500 mm.

[0058] The control process aims to detect indications in the material constituting the metal bar.

[0059] The method aims to detect these indications throughout the volume of the longitudinal end 3, regardless of the depth at which this indication is located.

[0060] Preferably, the method aims to detect the indications throughout the volume of the metal bar 1.

[0061] Here, an indication is understood to mean the image of a potential defect highlighted by a non-destructive testing method. It is only if this image is greater than the authorized tolerances that the term defect is used. In other words, an indication may or may not conform to the acceptance criteria issued by the customer. Only in the event of non-conformity is the indication considered a defect.

[0062] The indications may be:

[0063] - well-detached peaks: these indications characterize cracks, porosities, decohesions on inclusions, shrinkage bases;

[0064] - indications from noise, and therefore coming from the very structure of the material: these indications characterize large gains, segregations, or even macro-zones.

[0065] For the control of the longitudinal end 3 of the metal bar, the control method comprises a step of acquiring data using at least one multi-element probe 5 comprising several active elements 7, and a step of post-processing the data.

[0066] The active elements 7 are transmitters and / or receivers.

[0067] The multi-element probe 5 comprises several active transmitting elements and several active receiving elements.

[0068] Typically, each active element 7 is both a transmitter and a receiver. Alternatively, the transmitter active elements are distinct from the receiver active elements.

[0069] The active elements 7 are called transmitter / receiver elements in the following description.

[0070] The longitudinal end 3 is delimited by a lateral surface 9 with a closed contour.

[0071] It has a central Z axis. Typically, the central Z axis is longitudinal.

[0072] The lateral surface 9 extends all around the central axis Z of the longitudinal end 3. It delimits the longitudinal end 3 outwards, in radial directions from the central axis Z.

[0073] The longitudinal end 3 is also delimited by an end surface 10. In the example shown in FIG. 1, this end surface 10 is perpendicular to the central line Z.

[0074] The transmitter / receiver elements 7 of the multi-element probe are axially aligned.

[0075] Alternatively, the multi-element probe 5 is matrix-based, the transmitter / receiver elements 7 being arranged in a matrix with several rows and several columns.

[0076] According to another variant, the transmitter / receiver elements 7 are arranged in a circle or in any other suitable arrangement.

[0077] During the data acquisition step, as illustrated in Figure 1, the at least one multi-element probe 5 and the metal bar 1 are moved relative to each other such that the multi-element probe 5 circumferentially and / or axially scans the lateral surface 9.

[0078] Typically, the at least one multi-element probe 5 is moved circumferentially around said lateral surface 9 and / or axially along said lateral surface 9, the metal bar 1 remaining fixed.

[0079] Alternatively, the at least one multi-element probe 5 remains fixed, the metal bar 1 being driven in rotation around its axis and / or being moved axially relative to the at least one multi-element probe 5.

[0080] It is also possible to move both the multi-element probe 5 and the metal bar 1.

[0081] Circumferentially is understood here to mean that the multi-element probe 5 scans a perimeter of the lateral surface 9, around the central axis Z. The path of the multi-element probe 5 relative to the lateral surface 9 depends on the shape of the cross-section of the longitudinal end 3. This movement is not necessarily in a circle. It can be in a polygon, an ellipse, etc.

[0082] The at least one multi-element probe 5 emits at regularly spaced determined positions an incident ultrasonic signal towards a region of said longitudinal end 3, and detects the returned signal.

[0083] The returned signal is recorded with the corresponding circumferential and / or axial position of the multi-element probe 5 relative to the lateral surface 9.

[0084] The recorded circumferential and / or axial position corresponds to the position of the multi-element probe 5 at the time of emission of the incident ultrasonic signal.

[0085] Typically, the multi-element probe 5 scans the entire periphery of the lateral surface 9. The multi-element probe 5 is for example moved all around the central axis Z.

[0086] Axially, the relative displacement of the multi-element probe 5 and the metal bar 1 is such that the control covers a predetermined axial length of the metal bar 1 from the end surface 10.

[0087] This length is typically chosen to ensure 100% control of the volume of the metal bar 1.

[0088] Typically, the multi-element probe 5 must be able to control a length of several tens of millimeters from the end surface 10, for example a length of between 10 and 100 mm, preferably between 15 and 50 mm, for example equal to 20 mm.

[0089] The axial displacement is a function of the axial size of the multi-element probe 5 and the axial length to be controlled.

[0090] In some cases, the relative movement between the multi-element probe 5 and the metal bar 1 is not axial, but only circumferential.

[0091] Typically, the central part of the metal bar 1 is also checked, by any suitable method. The method for checking the longitudinal end 3 of the metal bar 1 is different, if applicable, from the method used for the central part of the metal bar 1, due to the particular constraints existing at the longitudinal ends 3.

[0092] In the paragraphs below, only the control of the longitudinal end 3 of the metal bar 1 will be detailed.

[0093] The phased array probe 5 and the metal bar are moved relative to each other regularly, continuously, circumferentially and / or axially, the incident ultrasonic signal being emitted at a regular time interval during the movement of the phased array probe 5. The returned signal, i.e. the echo of the incident ultrasonic signal, is recorded in a database for analysis in the post-processing step. It is recorded together with the circumferential and / or axial positions of the phased array probe 5 relative to the lateral surface at the time the ultrasonic signal was emitted.

[0094] During the data acquisition step, the multi-element sensor 5, and more precisely the transmitter-receiver elements 7, are pressed against the lateral surface 9 of the longitudinal end 3, directly or via a coupling member.

[0095] The coupling member is a water-filled bag or a shoe made of a rigid material.

[0096] The incident ultrasonic signal emitted at each determined position comprises several incident signals emitted successively, distinct from each other.

[0097] Alternatively, the incident signals emitted may be identical.

[0098] The returned signal comprises a matrix of elementary time signals corresponding to the echoes of each of the incident signals detected by each receiving element 7.

[0099] In other words, the matrix of elementary time signals contains, for each incident signal, the echo of this incident signal detected by each of the receiving elements 7.

[0100] The matrix therefore contains several rows, each corresponding to an incident signal. It contains several columns, each corresponding to a receiving element 7. The columns contain, for each row, the echo detected by the receiving element in response to each incident signal.

[0101] Advantageously, the longitudinal end 3 of the metal bar is checked using a single multi-element probe 5. The choice of the multi-element probe 5 depends on the material and the dimensions of the metal bar to be checked.

[0102] Alternatively, the longitudinal end 3 of the metal bar is monitored using several different multi-element probes 5, emitting ultrasonic signals distinct from each other. For example, these ultrasonic signals are differentiated by their frequencies specific to each multi-element probe.

[0103] Again, the probes are chosen according to the material and geometry of the metal bar to be tested.

[0104] For example, for a bar of approximately 300 mm diameter made of a titanium alloy of the Ti-6AI-4V type, it is advantageous to use two probes, one configured to emit an ultrasonic signal in a frequency range of 2 to 7.5 MHz, and the other to emit an ultrasonic signal in a frequency range of 5 to 10 MHz. A 5 MHz signal, for example, is well suited to detect the deepest indications, and a 7.5 MHz signal is well suited to detect the shallowest indications. When several phased array probes are used, several successive data acquisition steps are carried out, each with one of the probes. Each data acquisition step is carried out as indicated above.

[0105] Several sets of returned signals are therefore recorded, and are analyzed at the data post-processing stage.

[0106] According to a first example of implementation, the data acquisition step is carried out according to the FMC method (Full Matrix Capture, also called Inter-element Matrix). This method is illustrated in Figure 2.

[0107] As indicated above, the incident ultrasonic signal emitted at each determined position comprises several incident signals emitted successively. According to the FMC method, these incident signals are emitted successively each by a different emitting element 7.

[0108] In other words, each incident signal is emitted by a single transmitting element 7. As illustrated in Figure 2, this signal propagates in the region located under the multi-element probe 5, the echoes being returned to the receiving elements 7.

[0109] The incident signal is typically emitted with an angle of incidence substantially equal to 0°, i.e. perpendicular to the lateral surface 9 and to the fiber.

[0110] The matrix of elementary time signals is therefore a rectangular matrix, the number of rows being equal to the number of transmitting elements and the number of columns being equal to the number of receiving elements 7. When the active elements 7 are all transmitting and receiving, the matrix is ​​a square matrix.

[0111] According to a second exemplary embodiment, the data acquisition step is carried out using the PWI (Plane Wave Imaging) method.

[0112] Again, the ultrasonic signal comprises several incident signals distinct from each other and emitted successively. As illustrated in Figure 3, each incident signal is emitted by all the emitting elements 7 jointly. Each incident signal forms a plane wave.

[0113] This wave is in a plane perpendicular to the direction of propagation of the wave in the longitudinal end 3.

[0114] The different incident signals emitted have different angles of incidence ai from each other. The angle of incidence ai corresponds to the angle formed by the direction of propagation with the normal to the lateral surface 9, as shown in Figure 3.

[0115] In other words, to constitute a given incident signal, each transmitter-receiver element 7 emits an incident wave. These incident waves together form a plane wave propagating in a direction forming an angle of incidence ai relative to the normal to the lateral surface 9 of the bar. The angles of incidence ai are regularly distributed between -90° and +90°, or between -45° and +45°, or between -10° and +10°, or between -30° and 0°. The angles of incidence ai are separated for example by increments of 1°.

[0116] As illustrated in Figure 3, all receiving elements 7 record the echoes of each incident signal emitted.

[0117] The matrix of elementary time signals is therefore a rectangular matrix, the number of rows being equal to the number of incident signals emitted with different angles of incidence ai, and the number of columns being the number of receivers.

[0118] During the data post-processing step, the signal returned for each determined position of the multi-element probe 5 is post-processed.

[0119] This signal is the signal returned by the region towards which the multi-element probe 5 emitted the incident ultrasonic signal.

[0120] The returned signal is post-processed by determining in the elementary time signals the echoes returned by each point of said region of the longitudinal end, and by combining said echoes into a combined signal characterizing said point.

[0121] An example of an elementary time signal is illustrated in Figure 4. This figure shows the evolution of the elementary time signal as a function of time.

[0122] The origin of the time axis corresponds to the instant of emission of the incident signal. The position along the time axis corresponds to the time of flight of the ultrasonic waves from the transmitter to the receiving element 7 having recorded the elementary time signal.

[0123] This time of flight can be associated with a material point. The corresponding material point is the one for which the time of flight of the incident signal to this point and then, after reflection, to the receiving element, corresponds to the position along the time axis.

[0124] Thus, for each material point, it is possible to determine in each elementary time signal the echo possibly returned by said point.

[0125] These echoes can be combined in different ways.

[0126] According to a first variant, the data post-processing step is carried out according to the TFM method (Total Focusing Method, i.e. Method by focusing at all points).

[0127] The elementary time signals each have the equation: a(t) = 4(t).sin 4>(t) where a(t) is the value of the elementary time signal as a function of time, A(t) is the amplitude of the elementary time signal as a function of time, and <t>(t) is the phase of the elementary time signal as a function of time. According to the TFM method, the echoes returned by a given point in the region are combined by adding the amplitudes A(t) to form the combined signal characterizing said point.

[0128] In other words, according to the TFM method, we consider the matrix of elementary time signals.

[0129] For each point in the region to which the ultrasonic signal was emitted, the time of flight t is identified in each elementary time signal v corresponding to said point. This flight time corresponds to the flight time of the incident signal to said point, plus the flight time of the reflected signal, i.e. the echo, from said point to the receiving element having recorded the elementary time signal. This flight time may be different, for the same point, in the different elementary time signals.

[0130] The values ​​of the amplitudes A of the elementary time signals for the flight times are then determined.

[0131] For each elementary signal, the values ​​of A and to t v are determined by analyzing the shape of said elementary time signal in the vicinity of t v This analysis can be done simply and will not be detailed here.

[0132] The combined signal, for the point considered, is equal to the simple sum of the amplitudes A(t v ) determined for all elementary time signals.

[0133] Alternatively, the data post-processing step is performed using the PCI (Phase Coherence Imaging) method.

[0134] Again, each elementary time signal has the equation: a(t) = ï4(t).sin

[0135] The echoes returned by each point in the region of said longitudinal end are combined by adding the phases <t>(t) to form the combined signal characterizing said point.

[0136] For each point in the region towards which the incident ultrasonic signal was emitted, the time of flight t is identified in each elementary time signal v corresponding to said point. This flight time t v corresponds to the time of flight of the incident signal to said point, plus the time of flight of the reflected signal, i.e. the echo, from said point to the receiving element having recorded the elementary time signal. This time of flight t v may be different, for the same point, in the different elementary time signals.

[0137] The phase values ​​of the elementary time signals for the flight times t v are then determined.

[0138] For each elementary signal, the values ​​of A and to t v are determined by analyzing the shape of said elementary time signal in the vicinity of t v . This analysis can be carried out simply and will not be detailed here. The combined signal, for the point considered, is equal to the simple sum of the phases <t>(t v ) determined for all elementary time signals.

[0139] According to the TFM method, it is necessary to correct the amplitude A(t) to take into account the attenuation of the ultrasonic signal along its path in the metal bar. To do this, the method uses calibration curves, indicating the attenuation of the signal as a function of the depth at which the singularity that returned the emitted signal is located. These curves are determined experimentally.

[0140] The use of calibration curves is not necessary with the PCI method.

[0141] The method described above was tested on Ti-6AI-4V titanium alloy standards. These tests were carried out on three standards.

[0142] The first standard is a cylindrical bar shown in Figure 5, with a longitudinal length of 1500 mm and a diameter of 330 mm. A counterbore 11 is hollowed out in the end surface 13. A radial flat-bottomed hole 15 with a diameter of 1.2 mm is hollowed out in the circumferential wall of the counterbore 11. The flat-bottomed hole 15 is located at a depth of 5 mm below the end surface 13. The flat bottom of the flat-bottomed hole 15 is located at a depth of 40 mm below the lateral surface 17 of the standard.

[0143] A static test was performed, using the FMC method for the data acquisition step. The data post-processing step was performed in two different ways: according to the TFM method (Figure 6) and according to the PCI method (Figure 7). The right part of Figure 6 is a B-scan representation, and the left part is an A-scan representation taken along the L line of the B-scan.

[0144] The B-scan is a representation corresponding to a sectional view of the standard. The abscissa corresponds to the axial length of the multi-element probe, the ordinate corresponds to the depth taken from the external surface 17 of the standard. The B-scan represents the combined signal characterizing each point of this section.

[0145] The B band clearly visible on the B-scan corresponds to the bottom echo, i.e. the echo returned by the circumferential wall of the counterbore 11. The indication D corresponds to the flat bottom of the hole 15.

[0146] The A-scan, constituting the left part of Figure 6, represents the combined signal along line L. The abscissa corresponds to the value of this combined signal, and the ordinate corresponds to the depth. This A-scan could also be interpreted as the value of the combined signal as a function of time. The ordinate corresponds to the time of flight of the ultrasonic signal reflected by each point along line L. The indication D corresponds to the peak visible on the A-scan.

[0147] Figure 6 therefore shows that the flat-bottomed hole 15 can be detected using the method of the invention. Figure 7 is similar to Figure 6, but the post-processing of the data is carried out according to the PCI method. The black trace B corresponds to the circumferential wall of the counterbore 11. The indication D is visible on the B-scan. It corresponds to the peak visible on the A-scan.

[0148] These static tests were carried out with a multi-element probe operating at 7.5 MHz, and comprising sixty 7 transmitter-receiver elements. The 7 transmitter-receiver elements are arranged with a pitch of 1 mm. The pitch corresponds to the sum of the width of a transmitter-receiver element and the inter-element width.

[0149] Further static tests were carried out with the same probe for the standard shown in Figure 8. This standard is also made of Ti-6AI-4V. It has a length of 1500 mm and a diameter of 330 mm. It is typically used for the calibration of probes.

[0150] Several flat-bottomed holes 18 with a diameter of 1.2 mm are made in this standard. The flat-bottomed holes 18 are located at different depths below the lateral surface 19 of the standard. The depths range between 3 and 170 mm.

[0151] These tests were carried out with the same multi-element probe as for the standard in Figure 5. For the detection of holes 18 close to the lateral surface 19, the multi-element probe was used with a wedge.

[0152] The FMC method was used for the data acquisition stage. Data postprocessing was performed using both the TFM and PCI methods.

[0153] These tests were carried out on the standards without prior creation of a CAD curve (amplitude / distance curve) or TCG (time correction gain).

[0154] Tests show that the combined signal for flat-bottomed holes varies between 60 and 87% of the screen height, for an average background noise of 5 to 10% of the screen height.

[0155] The amplitude of the combined signal and background noise is expressed as a percentage of the screen height. On the left part (A-scan) of Figures 6 and 7, the screen height corresponds to the length of the abscissa axis.

[0156] The TFM and PCI methods give results of the same order.

[0157] Further static tests were carried out on the standard shown in Figure 9. This standard being Ti-6AI-4V.

[0158] It has a length of 1500 mm, and a diameter of 10 inches, that is to say approximately 254 mm. Several flat-bottomed holes 21 with a diameter of 0.8 mm are provided along the standard. These holes are located at different depths under the lateral surface 23 of the standard. The depths range between 15 and 140 mm.

[0159] The probe used is the same as for the standards in Figures 5 and 8.

[0160] The FMC method was used for the data acquisition step. Data post-processing is performed using the TFM method. Tests show that the combined signal for flat-bottomed holes varies between 56 and 86% of the screen height, for an average background noise lower than 20% of the screen height.

[0161] Dynamic tests were also carried out for the three standards described above.

[0162] These dynamic tests are performed by circumferentially moving the multi-element probe around the standards.

[0163] For the standard in Figure 5, the test was carried out with the same probe as previously.

[0164] The data acquisition step is performed using the FMC method. The data post-processing step is performed using the TFM method.

[0165] The results are shown in Figures 10 and 11. Figure 10 is similar to Figure 6. The right part of Figure 10 is a B-scan, and the left part is an A-scan taken along the L line of the B-scan.

[0166] Again, the horizontal line B corresponds to the circumferential wall of the counterbore 11, and the indication D is clearly visible.

[0167] The lower right part of Figure 11 is a B-scan, but taken in a different plane than the B-scan in Figure 10. The indication is not visible on this B-scan.

[0168] The upper part of Figure 11 is a C-scan. In other words, it is a developed view of the lateral surface 17 of the standard. The ordinate corresponds to the axial direction, i.e., the length of the phased array probe. The abscissa axis corresponds to the perimeter of the standard.

[0169] The D indication is clearly visible on the C-scan.

[0170] The lower left part of Figure 11 is a circumferential map. It is a section taken along a circumference of the standard. The ordinate corresponds to the depth below the lateral surface 17, and the abscissa to the position along the circumference. Line B corresponds to the circumferential wall of the counterbore 11, and the indication D is clearly visible.

[0171] Dynamic tests were also carried out on the standard in Figure 8.

[0172] For the holes 18 located closest under the lateral surface 19, the multi-element probe used is that described previously, mounted on a shoe. For the other holes 18, a multi-element probe operating at 5 MHz was used. It has 64 transmitter-receiver elements, arranged with a pitch of 1 mm.

[0173] The data acquisition step is performed using the FMC method. The data post-processing step is performed using the TFM method. The tests show that the combined signal for flat-bottomed holes varies between 55 and 126% of the screen height, for an average background noise lower than 15% of the screen height.

[0174] Dynamic tests were also carried out for the standard in Figure 9. The probe used is the 7.5 MHz probe described above. The data acquisition step is carried out using the FMC method. The data post-processing step is carried out using the TFM method.

[0175] Tests show that the combined signal for flat-bottomed holes varies between 62 and 188% of the screen height, with an average background noise of less than 20% of the screen height. For the deepest hole, the combined signal is only 29% of the screen height, compared with an average background noise of less than 10% of the screen height.

[0176] Thus, the above tests show a capability of the invention to detect in static or dynamic mode indications simulated by flat-bottomed holes (a) both at the end of the bar (5 mm from the end surface) and (b) also in the central part of the bar, and this (c) at different depths in the diameter (3 to 170 mm).

[0177] The invention also relates to an assembly for non-destructive ultrasonic testing of the longitudinal metal bar 1.

[0178] This set is specially designed for implementing the control method described above. Conversely, the control method is specially adapted to be executed with the control set which will now be described.

[0179] The assembly is designed to carry out a check of the longitudinal end 3 of the metal bar 1.

[0180] The control assembly comprises a data acquisition device 25 and a data post-processing device 27.

[0181] The data acquisition device 25 comprises at least one multi-element probe 5 comprising several emitting elements and several receiving elements 7, a mechanical device 29 configured to move the at least one multi-element probe 5 and the metal bar 1 relative to each other such that the multi-element probe 5 scans circumferentially and / or axially said lateral surface 9, and a controller 31 configured so that the at least one multi-element probe 5 emits at regularly spaced determined positions an incident ultrasonic signal towards a region of said longitudinal end 3 and detects the returned signal.

[0182] The controller 31 is configured so that the returned signal is recorded with the corresponding circumferential and / or axial position of the multi-element probe 5 relative to the lateral surface 9. To do this, the data acquisition device 25 comprises a mechanism for encoding the position of the multi-element probe 5, adapted according to the shape and size of the cross-section of the longitudinal end 3 of the metal bar.

[0183] The incident ultrasonic signal comprises several incident signals emitted successively, distinct from each other.

[0184] The returned signal comprising a matrix of elementary time signals corresponding to the echoes of each of the incident signals detected by each receiving element 7.

[0185] The data post-processing device 27 is configured so that the returned signal is post-processed by determining in the elementary time signals the echoes returned by each point of said region of the longitudinal end and by combining said echoes into a combined signal characterizing said point.

[0186] The phased array probe 5 is as described above.

[0187] The mechanical device 29 is advantageously a carrier configured to move the at least one multi-element probe 5 circumferentially around said lateral surface 9 and / or axially along said lateral surface 9, the metal bar 1 remaining fixed. It is chosen according to the geometry of the longitudinal end 3. For example, it comprises a frame 33 rigidly fixed to the metal bar 1, a carrier structure 35, a connection 37 of the carrier structure to the frame configured to allow the movement of the carrier structure 35 circumferentially around the lateral surface 9, and a connection 39 of the multi-element probe 5 to the carrier structure 35 configured to allow the movement of the probe 5 longitudinally along the carrier structure 35.

[0188] Alternatively, the at least one multi-element probe 5 remains fixed, the mechanical device 29 being configured to drive the metal bar 1 in rotation around its axis and / or axially relative to the at least one multi-element probe 5.

[0189] The controller 31 is configured so that data acquisition is performed using the FMC method or the PWI method.

[0190] These methods have been described above.

[0191] The elementary time signals each have the equation a(t) = ^(t).sin(t).

[0192] According to a first variant, the data post-processing device 27 is configured so that the echoes returned by each point of the region of the longitudinal end 3 are combined by adding the amplitudes A(t) to form the combined signal characterizing said point.

[0193] Advantageously, in this case, the data post-processing device 27 is configured so that the data post-processing is carried out according to the TFM method.

[0194] This post-processing method is described above. According to a second variant, the data post-processing device 27 is configured so that the echoes returned by each point of the region of the longitudinal end 3 are combined by adding the phases <t>(t) to form the combined signal characterizing said point. Advantageously, in this case, the data post-processing device 27 is configured so that the data post-processing is carried out according to the PCI method.

[0195] This post-processing method is described above.

[0196] According to a variant of the method not shown, the metal bar is hollow, and therefore has an internal lateral surface and an external lateral surface. The multi-element probe is placed directly or indirectly against the internal lateral surface of the metal bar or on the external lateral surface. It is moved circumferentially and / or axially on the internal lateral surface or on the external lateral surface. The method is otherwise identical to the method described above.< / t> < / t> < / t> < / t> < / t> < / t>

Claims

CLAIMS 1. A method for non-destructive ultrasonic testing of a longitudinal metal bar (1), the method comprising testing a longitudinal end (3) of the metal bar (1) having a central axis, the longitudinal end (3) being delimited by a lateral surface (9) with a closed contour and by an end surface (10), the longitudinal end (3) having a cross-section whose largest dimension is between 100 and 500 mm; the testing method comprising a step of acquiring data using at least one multi-element probe (5) comprising several emitting elements and several receiving elements (7), and a step of post-processing the data;during the data acquisition step, moving the at least one multi-element probe (5) and the metal bar (1) relative to each other such that the at least one multi-element probe (5) scans circumferentially and / or axially said lateral surface (9), the at least one multi-element probe (5) emitting at regularly spaced determined positions an incident ultrasonic signal towards a region of said longitudinal end (3) and detecting the returned signal, the returned signal being recorded with the corresponding circumferential and / or axial position of the multi-element probe (5) relative to the lateral surface (9); the incident ultrasonic signal comprising several incident signals emitted successively, distinct from each other; the returned signal comprising a matrix of elementary time signals corresponding to the echoes of each of the incident signals detected by each receiving element (7);during the data post-processing step, the returned signal is post-processed by determining in the elementary time signals the echoes returned by each point of said region of the longitudinal end (3) and by combining said echoes into a combined signal characterizing said point; during the data acquisition step, the at least one multi-element probe (5) scans circumferentially and / or axially the lateral surface (9) so that the entire volume of the longitudinal end (3) of the metal bar (1) is monitored over a longitudinal length of between 10 and 100 mm from the end surface (10).; 2. Control method according to claim 1, in which the data acquisition step is carried out according to the FMC method.

3. Control method according to claim 1 or 2, in which the incident ultrasonic signal comprises several incident signals emitted successively, each by a different emitting element.

4. Control method according to claim 1, in which the data acquisition step is carried out according to the PWI method.

5. A control method according to claim 1 or 4, wherein the incident ultrasonic signal comprises several incident signals emitted successively, each incident signal being emitted by all the emitting elements (7) jointly, each incident signal forming a plane wave, the incident signals having angles of incidence different from each other.

6. Control method according to any one of the preceding claims, in which the data post-processing step is carried out according to the TFM method.

7. Control method according to any one of the preceding claims, in which the elementary time signals each have the equation a(t) = ï4(t).sin the echoes returned by each point of the region of said longitudinal end (3) being combined by adding the amplitudes A(t) to form the combined signal characterizing said point.

8. Control method according to any one of claims 1 to 5, in which the data post-processing step is carried out according to the PCI method.

9. Control method according to any one of claims 1 to 5 or 8, in which the elementary time signals each have the equation a(t) = ï4(t).sin the echoes returned by each point of the region of said longitudinal end (3) being combined by adding the phases <t>(t) to form the combined signal characterizing said point.

10. A testing method according to any one of the preceding claims, wherein the metal bar (1) is made of titanium or a titanium alloy, for example Ti-6AI-4V or Ti-5553 or Ti 10-2-3, or martensitic steel, or maraging steel, or austenitic steel, or a nickel or cobalt-based superalloy, or an aluminum-based alloy.

11. A control method according to any one of the preceding claims, wherein the method comprises a step of obtaining a map of the longitudinal end (3), showing the combined signal at each point of the longitudinal end (3).< / t>

Images