Method for detecting discontinuities and system for implementing said method

Abstract

Method for detecting discontinuities in a medium and system for implementing said method, said method comprising the step of defining a transmission sequence, wherein a plurality of transmitting transducers selected among the transducers of a probe are uniformly and randomly located on the active surface of said probe, and a time offset defined for each transmitting transducer is uniformly and randomly located within a predetermined transmission duration. Subsequently, said transmission sequence is transmitted in said medium by said plurality of transmitting transducers, the received signals are received and recorded, and said received signals are processed with a focusing law adapted to said transmitting transducers, and using said time offsets so as to thus derive a detection level.

Description

TECHNICAL FIELD

[0001] The present invention relates to a method for detecting discontinuities in a product, in particular using ultrasonic waves. BACKGROUND

[0002] There exist many techniques for detecting discontinuities or defects in products, for example metallic products. These techniques fall within the field of non-destructive testing. Techniques using the propagation of ultrasonic waves efficiently detect discontinuities or defects in the material of a product, after its manufacture or during its lifetime, to ensure that it works properly.

[0003] Thus, patent document FR 2 830 328 exhibits a method for detecting discontinuities using several beams simultaneously in several directions, but this method creates interference on reception and the signal-to-noise ratio is degraded compared to a single-beam method.

[0004] Document "Post-processing of the full matrix of ultrasonic transmit-receive array data for non-destructive evaluation" (International NDT & E 38 (2005) 701-711) evaluates a technique for capturing a full matrix of time-domain signals between all pairs of transmitting and receiving transducers. These time-domain signals are then processed in post-processing, for example by a focusing algorithm at each point of the medium.

[0005] However, in this technique, each transducer is emitted one after the other, the received signals from all transducers being ideally stored between each emission. This method is therefore not satisfactory with regard to the execution speed. SUMMARY

[0006] The object of the present invention is to improve the method for detecting discontinuities of this type, in order to improve in particular the quality of detection at high speed.

[0007] To this end, the method for detecting discontinuities according to the present disclosure is implemented using a probe comprising a plurality of transducers forming an active surface and capable of emitting and receiving ultrasonic waves in a medium, the method comprising the following steps:

[0008] - defining an emission sequence, in which:

[0009] A plurality of transmitting transducers are selected from among the transducers of the probe, each transmitting transducer of the plurality of transmitting transducers having a determined spatial position, such that the spatial positions of the plurality of transmitting transducers are uniformly and randomly distributed over the active surface of the probe, and

[0010] defining a time offset for each transmit transducer of the plurality of transmit transducers such that the time offsets of the plurality of transmit transducers are uniformly and randomly distributed within a predetermined transmit duration,

[0011] - transmitting said transmit sequence in a medium by said plurality of transmit transducers,

[0012] - receiving and recording received signals by said plurality of transducers in response to a transmission sequence transmitted in said medium,

[0013] - Process the received signal according to the following procedure, wherein:

[0014] for each transmitting transducer, calculating the focusing delay corresponding to the determined focal law of said transmitting transducer for the desired target point detected in the medium and taking into account the time offset of the considered transmitting transducer,

[0015] The focus signal is calculated for each transmitting transducer as the sum of the received signals of the multiple transducers of the probe realigned by the focusing delay,

[0016] Calculate the composite signal, which is the sum of the focused signals from all transmitting transducers, and

[0017] The composite signal is analyzed in order to derive thus the detection level in said medium at the target point, and thus the detection of the discontinuity.

[0018] By means of these arrangements, the detection method results in little interference and the detection quality is improved especially at high speeds.

[0019] Specifically, this method requires only a single transmission sequence. Detection and / or imaging is done entirely in post-processing, and the specific sequence and possibly its processing make it possible to avoid interference between the recorded received signals as much as possible and / or reduce the effects of such interference, thus providing fast and accurate discontinuity detection (detection of level amplitudes).

[0020] In various embodiments of methods according to the present disclosure, one or more of the following arrangements may further be used.

[0021] According to one aspect, the method further comprises the step of reducing interference between the plurality of discontinuities before calculating the focus signal by a process consisting of:

[0022] - determining a curve of peaks in the received signals as a function of the transducers, said curve being determined by identifying the peaks in the received signals of the target points and according to a focusing law,

[0023] - calculating a model curve approximating the curve of peaks, and

[0024] - calculating corrected received signals from the recorded received signals and the model curve, then using said corrected received signals instead of the recorded received signals to process these signals to derive a detection level.

[0025] According to one aspect, said model curve is a polynomial curve.

[0026] According to one aspect, a focusing delay is determined as a function of the speed of movement of the medium relative to the probe (and vice versa).

[0027] According to one aspect, the processing of the received signals is iterated for a plurality of target points so as to build an image of the medium representative of different detection levels in said target points.

[0028] The disclosure also relates to a system for detecting discontinuities implementing the method described above. This system comprises a probe comprising a plurality of transducers capable of emitting and receiving ultrasound waves in a medium, and a processing unit connected to said probe, this processing unit comprising at least one memory for recording received signals, and a controller to implement said method. BRIEF DESCRIPTION OF DRAWINGS

[0029] Other features and advantages of the application will appear clearly to the person skilled in the art from the following description of at least one embodiment given as a non-restrictive example, with reference to the attached drawings.

[0030] In the drawings:

[0031] - Figure 1 is a general block diagram of an example of a system implementing the method according to the application;

[0032] - Figure 2 is a timing diagram of the signals from a selected transmitting transducer in an example of a transmitting sequence of the method;

[0033] - Figure 3 is a timing diagram of the received signals from the transducers of the probe according to the method;

[0034] - Figure 4 is a timing diagram of the focused delay applied to the received signals of Figure 3 ;

[0035] - Figure 5 is a block diagram of the summation block of Figure 1 ;

[0036] - Figure 6 is an example of a focused signal obtained by summing phased signals;

[0037] - Figure 7 is a timing diagram of the calculated focus signal associated with the selected transmit transducer;

[0038] - Figure 8 is through Figure 7 An example of a time graph of a composite signal calculated by summing the focus signals;

[0039] - Figure 9 is a timing diagram of the received signal in the case of several discontinuities in the medium and a single excitation of a single transmitting transducer;

[0040] - Figure 10 is an example of a curve with the peak value plotted as a function of the transducer for a target point in the first discontinuity.

[0041] Throughout the drawings, the same reference numbers refer to the same or similar elements. DETAILED DESCRIPTION

[0042] An example embodiment of a system for detecting discontinuities or defects in a medium of a product, and an example of a method implemented by such a system, are described below in an illustrative and non-limiting manner.

[0043] According to the present disclosure and Figure 1 In this example of a system shown in , the system 100 is a system for detecting discontinuities in a medium M. The medium M is, for example, a product, such as a metal part, which may include one or more discontinuities or defects within its material, such as one or more air pockets or cracks. The goal of the system and method is therefore to detect these discontinuities or defects. Detection is to be understood as meaning, for example, obtaining information about the presence of a discontinuity, and / or obtaining distance information relative to the system and / or obtaining position information relative to the system, and / or obtaining information about the shape of the discontinuity. It is possible that the product corresponding to the medium M moves at a relative speed relative to the system 100, which requires very fast detection of the discontinuity. In order to achieve reliable product control, it is important not to miss this detection.

[0044] The system and method according to the present disclosure are suitable for non-destructive testing of metal products or parts such as pipes and rails of railways. Specifically, the system and method are optionally used during the movement of the product or while it is rolling along the rails.

[0045] Figure 1is an example of a functional block diagram of the system 100 having various processing blocks. The system 100 may optionally be broken down into different functional blocks, but these functional blocks will contain the basic functionality of the example of the system 100 described below.

[0046] exist Figure 1 In the example of , the system 100 includes a probe 10 that exchanges signals with a processing unit 20 connected to the probe via, for example, an electrical or optical wired link or a wireless link such as radio waves. Thus, the probe 10 can be located at a certain distance from the processing unit 20. Optionally, the probe 10 and the processing unit 20 are integrated into a single device, or some elements (functions) of the processing unit are located in the probe 10.

[0047] The probe 10 comprises a plurality of transducers Tn, for example with index n=1...N, which form the active surface of the probe. The N transducers Tn are arranged, for example, along Figure 1 In practice, the number N of transducers is, for example, between tens and hundreds.

[0048] In the case of a one-dimensional linear probe, the transducer Tn of the probe has a spatial position indicated by the abscissa xn of the transducer along the longitudinal direction X of the probe. The depth direction Z, perpendicular to the longitudinal direction X, corresponds to the depth position in the medium from an origin O placed on the outer surface of the medium M in the case of contact between the probe 10 and the medium M. The origin O, the longitudinal direction X, and the depth direction Z form a reference system in which elements can be positionally identified by spatial coordinates (x, z).

[0049] Optionally, the probe 10 can be a curved probe in the form of an arc of a circle, a planar array probe of NxP transducers, or an array probe in the form of a cross section of a cylinder, or any other probe shape. The systems and methods described herein will be readily adapted to such a probe.

[0050] Thus, the product may for example be stationary or moved in translation or rotation relative to the probe 10 of the system (or vice versa, ie the probe moves relative to the product).

[0051] The active surface of the probe 10 is brought into direct or indirect contact with the outer surface of the medium M. Each transducer Tn of the probe 10 is an element capable of transmitting a transmission wave in the medium M and / or receiving an echo wave in the medium in response to the properties of the medium. The wave is usually an ultrasonic wave. For example, Figure 1As shown in FIG, a transmitting transducer Te with index e and coordinates (xe, 0) transmits a transmitting wave Em toward a target point C inside a medium, and the medium transmits an echo Re from this target point C, for example, toward a receiving transducer Tr with index r and coordinates (xr, 0). A region of interest (ROI) is defined for detecting discontinuities or defects in the medium. This region of interest is, for example, rectangular with sides parallel to the reference system.

[0052] The transducer Tn of the probe 10 receives a signal for transmitting a transmission wave and generates a signal after receiving an echo. The transducer of the probe 10 is therefore connected to a transmit-receive module 110 (also labeled "E / R"). The transmit-receive module 110 transmits to the transducer a transmission signal prepared by the system in the transmitter control unit 113 (also labeled "Em" in the figure), or a received signal from the transducer to one or more analog / digital converters 120 (also labeled "A / D" in the figure), which digitize these analog received signals and convert them into digital data, which are then stored in one or more memories 130 (also labeled "Mem" in the figure) of the system.

[0053] The transmit-receive module 110 is known and enables the sequential use of the transducers of the probe 10 during wave transmission or wave reception. It also comprises, for example (but not limited to):

[0054] - a transmit pulse generator in the transmit chain, which generates the transmit signal, typically a pulse with programmable amplitude and time width, and

[0055] -Programmable gain amplifier and anti-aliasing filter in the receive chain.

[0056] Transmitted signals and waves are typically pulses of relatively short duration. These pulses are, for example, single or multiple rectangular signals with variable amplitudes, or signals modulated according to one or more frequencies, or a combination of these signals. Returned waves and signals are echoes corresponding to these transmitted pulses, deformed by their transmission through the medium. For simplicity of explanation, the term "pulse" will be used more generally to refer to both.

[0057] like Figure 1 As shown in the figure, the system 100 further includes a synchronization module 114 (also labeled "Synch" in the figure) connected to the analog / digital converter 120 and the memory 130, which allows triggering the digitization of the received signal into digital data (analog / digital conversion) and saving these digital data in the memory 130.

[0058] Furthermore, the system 100 may optionally include an interference reduction block 150 (also labeled "IRB" in the figure) connected to the memory 130, whose function is to modify the digital data of the received signal by correcting the digital data so as to eliminate or reduce interference in the received signal due to the presence of certain discontinuities or defects in the medium. This interference reduction block will be described in more detail below.

[0059] The system 100 then includes a summing block 200 (also labeled "Accu" in the figure) connected to the aforementioned memory block 130 and / or interference reduction block 150, and which performs a calculation of the response of the medium at one or more target points C based directly on the data from the memory 130, or based on the corrected data from the interference reduction block 150, or based on a combination of the two. The summing block 200 then provides the response to a monitor, screen, or any display device 117 (also labeled "Mon" in the figure) to inform the system user of one or more items of information (presence, location, shape, image, etc.) about the discontinuity or defect in the medium. This summing block 200 will be described in more detail below.

[0060] System 100 includes a delay calculator 115 (also labeled “R” in the figure) that determines a time delay of a portion of the digital data (i.e., the received signal) or an index to a location in memory 130 that can be used in subsequent blocks (i.e., interference reduction block 150 and / or summation block 200).

[0061] The system 100 also optionally includes a dynamic velocity corrector 116 (also labeled “Corr” in the figure) that uses product velocity measurements measured by one or more sensors (not shown) and is connected to the delay calculator 115 to compensate for movement of the product (i.e., the medium) during the propagation of the transmitted wave Em from the probe 10 toward the target point C and during the propagation of the echo wave Re from the target point C toward the probe 10.

[0062] Finally, the controller 300 (also labeled "Contr" in the figure) is coupled to each of the aforementioned blocks to manage the general operation. More specifically, the controller 300 is coupled to the transmitter control unit 113 to transmit the transmit wave Em according to a transmission program predetermined by the user, to the synchronization module 114 to ensure the correct acquisition of the echo Re, to the delay calculator 115 to ensure the combination of the received signals suitable for the desired focusing, to the dynamic velocity corrector 116, to the interference reduction block 150 to provide it with user-operated parameters, to the beamforming block to control its calculation and correction parameters, and to the screen 117 to shape the various displays and control elements of the system 100.

[0063] The operation of system 100 will now be described.

[0064] Specifically, the controller 300 and the control unit 113 of the system 100 according to the present disclosure construct a transmission sequence for a specific transmission wave Em.

[0065] According to the first prior art, several transmitting transducers Te each generate pulses with a predetermined time offset, so that the generated transmitted waves are physically focused toward a target point in the medium. Consequently, several transmitted wave transmissions are performed sequentially, each followed by waiting for an echo and storing the received signal to scan multiple points in the medium and potentially construct an image. This technique therefore involves numerous transmissions of the transmitted wave. This method is particularly slow for detecting large regions of interest or for detecting with spatially precise resolution. Furthermore, it makes it nearly impossible to detect products moving at considerable speeds in front of the probe.

[0066] According to the second prior art, a region of interest in a medium is scanned using one or more transmissions of an unfocused wave, such as a plane wave. Processing the echoed signals allows information about the region of interest to be obtained and a rapid image of the region of interest to be generated. However, the quality of the generated image is poor because the energy of the transmitted wave is widely distributed in space, degrading the signal-to-noise ratio.

[0067] According to the third prior art, a single transmission is performed for each transducer of the probe, and the echo signal is recorded. This technique of recording the full matrix of signals, as used in "Post-processing of the full matrix of ultrasonic transmit-receive array data for non-destructive evaluation" (International Journal of Nondestructive Testing and Evaluation 38 (2005) 701-711), results in a slow testing method due to the need for multiple transmissions.

[0068] Therefore, the present disclosure proposes a specific transmission sequence in which a specific number of transmitting transducers Te are quasi-randomly selected in the probe 10, and / or the pulses of the transmission signals of these transmitting transducers are shifted by a quasi-random time offset te. As a result, the received signal will have minimal temporal coherence and minimal symmetry, and the temporal interference of the received signal is reduced, especially in the case of multiple discontinuities in the product medium. This therefore makes it possible to improve the quality (information or image) of the detection of the discontinuities. In addition, this transmission sequence makes it possible to detect substantially the entire region of interest ROI of the medium using a single transmission and the echoes from this transmission. Therefore, this technique is relatively fast and suitable for detecting products that are moving, for example, at a speed V relative to the probe 10 of the system. It is assumed that the speed V is substantially constant during the detection process.

[0069] This transmission sequence is generated, for example, according to the following process:

[0070] - selecting a plurality (Ne) of transmitting transducers Te among the N transducers Tn of the probe 10, each transmitting transducer of the plurality of transmitting transducers having a spatial position xed determined so that the spatial positions of the plurality of transmitting transducers are uniformly and randomly distributed on the active surface of the probe 10, and

[0071] - defining a time offset te for each transmitting transducer Te among the plurality (Ne) of transmitting transducers, such that the time offsets te of the plurality (Ne) of transmitting transducers are uniformly and randomly distributed within a predetermined transmission duration DTe corresponding to a maximum duration of transmission in the medium M.

[0072] This transmit sequence thus defines a transmit transducer Te and a time offset te, each of which is associated with a transmit transducer from the plurality of transmit transducers. The transmit sequence thus defines the wave emission (ultrasound) of Ne transmit transducers at a time te relative to an initial reference time t0 of the transmit sequence. Each emission of the sequence from transmit transducer Te is a pulse of very short duration. These emissions of ultrasound pulses are thus spatially distributed according to the transmit transducer and temporally distributed over the transmit duration DTe of the transmit sequence.

[0073] The selected number Ne of transmitting transducers Te is two or more. Preferably, the selected number Ne of transmitting transducers Te is greater than or equal to five or ten, and this number depends on the number N of transducers of the probe. This number Ne of transmitting transducers is less than the number N of transducers of the probe.

[0074] For example, the selected number Ne of transmitting transducers Te is between 0.05xN and 0.25xN, ie between 5% and 25% of the transducers of the probe 10. Thus, within this range of transducer usage, it is possible to obtain a better detection quality of discontinuities.

[0075] In practice, the skilled person will adjust the density of the transmitting transducers and the density of the signals of the transmitting sequence in a compromise manner for improving the detection quality and depending on the application.

[0076] These spatial and temporal distributions are referred to as being "uniformly and randomly," meaning that the emissions are generally well-spaced from one another in both the spatial and temporal dimensions, but with random variations or deviations. In other words, the emission distributions are not regularly spaced or periodic in either dimension. They are also not completely random, as they are preferably spaced from one another.

[0077] Figure 2An example of this transmit sequence is shown for a probe 10 having 64 transducers, of which four (4) transmit transducers Te are selected. Figure 2 The graph shows four transmit signals from four selected transmit transducers, here the transducers with indices 5, 26, 40 and 61. The transmit signals of the other transducers (transducers that are not the transmit transducers) are therefore zero, and therefore they are not in the Figure 2 Therefore, we have the transmission signals se5, se 26 、se 40 and 61 These transmission signals are at the corresponding time te5, te 26 、te 40 and te 61 64 of the transducers of the probe 10, these times are quasi-random, or indeed distributed in a uniform and random manner within the transmission duration DTe. Similarly, the indices of the transmitting transducers Te are distributed in a quasi-random manner among the possible indices 1 ... 64 of the transducers of the probe 10, or indeed distributed in a uniform and random manner among the possible indices of the transducers.

[0078] A first technique for obtaining a uniform and random distribution in the space or time dimension is simply to define random values ​​within a size interval of the space or time dimension; in other words:

[0079] - for the spatial dimension of the transducers, a random value of the transducer index between 1 and N, N being the number of transducers of the probe 10; and

[0080] - For the time dimension, a random value of the time between the initial time t0 of the transmission sequence and the maximum final time (t0+Dte) of the transmission sequence, where DTe is the transmission duration.

[0081] A second technique for obtaining a uniform and random distribution is to divide the space or time dimension into a number NI of contiguous intervals of constant and equal size, and to define a random value within each of the intervals.

[0082] This makes it possible to obtain values ​​of the spatial or temporal dimension that are more evenly distributed over this dimension.In fact, each of the NI intervals of the spatial or temporal dimension contains only one element.

[0083] A third technique for obtaining a uniform and random distribution is to divide the spatial or temporal dimension into a constant number of contiguous intervals of equal size, NI, and to define a value in each of these intervals relative to the median MI of each interval. Thus, the value is equal to the median plus a random value corresponding to the deviation from the median. The random value can take positive or negative values, and its amplitude can be limited to, for example, half the size of the interval.

[0084] This makes it possible to obtain values ​​of the spatial or temporal dimension that are more evenly distributed over this dimension. In fact, not only does each of the NI intervals of the spatial or temporal dimension contain only one element, but said elements are placed around a central value (which is the median of the interval) with a deviation that can be controlled by the defined amplitude.

[0085] All of these techniques will be implemented by a person skilled in the art who has mathematical knowledge, but he or she may also use other techniques for uniform and random distribution of values ​​in space and time.

[0086] Thus, by at least one of these techniques, a skilled person will be able to determine a predetermined number of transmitting transducers Te among the number N of transducers of the probe, and by at least one of these techniques, will be able to determine the time offsets te associated with these transmitting transducers, each of these time offsets being distributed between zero and Dte (transmit duration). Specifically, the number of intervals NI listed above is equal to the predetermined (selected) number Ne of transmitting transducers Te.

[0087] The defined transmission sequence according to the present disclosure makes it possible to reduce the amount of interference between pulses in the received signal and to improve the detection of discontinuities, as will become more apparent from the following explanation.

[0088] The memory 130 of the system 100 contains N received signals in the form of digital data constituting a complete capture matrix of the medium M of Ne transmissions from the transmitting transducer Te during a transmission sequence, these transmissions each being shifted by a time offset te among the transmissions of this transmission sequence.

[0089] Figure 3 In the case of a single discontinuity or defect in the medium M of the product, for example, Figure 2 N received signals sr are obtained and stored in the memory 130 after transmission by the four transmitting transducers Te represented in FIG. Figure 3 Display 64 received signals sr1 to sr 64 In order to help readers understand the diagram, this figure and subsequent figures present the situation of a linear probe that simplifies the distribution of the signal.

[0090] In the present case of N transducers, the peaks in the pulses of these N received signals form curves that do not intersect with and correspond to the echoes from the Ne transmitted pulses. These curves are parallel zigzag lines, meaning they are spaced apart from each other in time by a constant value that depends on the time offset at the time of transmission and the location of the discontinuity in the medium.

[0091] Each of these N received signals consists of Ne pulses, corresponding to the Ne echoes from a single discontinuity in the corresponding receiving transducer. The time intervals between these pulses depend on the temporal distribution of the Ne emissions and also on the spatial location of the single discontinuity in the medium M. However, the time offset between the two received signals of the two receiving transducers Tr depends only on the spatial location of the single discontinuity in the medium M.

[0092] The controller 300 extracts the N portions of the received signal from the memory 130 by programming the delay calculator 115 with a focused delay calculation denoted as Delay for the target point C, the transmitting transducer Te and the receiving transducer Tr. This focused delay calculation is, for example, the following calculation:

[0093] Delay=te+tem+tre+tadj

[0094] in:

[0095] te is the time offset of the emission from the transmitting transducer,

[0096] tem is the outward travel time of the transmitted wave from the transmitting transducer Te to the target point C in the region of interest ROI,

[0097] tre is the return travel time of the echo from the target point C to any transducer of the probe 10, and

[0098] tadj is the shift time to be adjusted according to the maximum value of the received signal.

[0099] Specifically, the simplified beamforming arrangement for a linear transducer is given by the following formula via geometric calculation:

[0100] tem=1 / Va.sqrt((xe-xc) 2 +zc 2 )

[0101] tre=1 / Vr.sqrt((xr-xc) 2 +zc 2 )

[0102] (xe,ze), (xr,zr) and (xc,zc) are the coordinates of the transmitting transducer Te, the receiving transducer and the target point C in the coordinate system, and ze and zr are zero for the linear probe.

[0103] Va is the velocity of the transmitted wave in the medium for the outward path between the transmitting transducer Te and the target point C,

[0104] Vr is the velocity of the echo in the medium for the return path between the target point C and the receiving transducer Tr, and

[0105] sqrt is the square root mathematical function.

[0106] The focus delay of the present disclosure is a calculation of receive beamforming, but it is different from the common receive beamforming delay because a time offset te used in the transmit sequence of each transmit transducer Te is added.

[0107] The above delay calculation can be expressed as a memory index in the memory 130 by multiplying the outgoing travel time tem, the return travel time tre and the time offset te by the sampling frequency Fs in the case of a system sampled at this constant sampling frequency.

[0108] Other variations of the delay calculation are possible and accessible to those skilled in the art. Specifically, these variations depend on the probe's geometry, which changes the distance of the outgoing path of the transmitted wave and the return path of the echo. Similarly, these variations depend on the consideration of the intermediate medium between the probe and the product's medium, which also modifies the distance calculations in the path.

[0109] It is possible to perform the delay calculation for focusing in a predetermined direction. In this case, the formula for calculating the delay is different. This specification details and clarifies the operation for focusing toward a target point, but it is possible for those skilled in the art to establish other formulas for directional focusing and construct methods for detecting discontinuities and systems suitable for other types of focusing.

[0110] The controller 300 thus causes the delay calculator 115 to calculate the focus delays of the N received signals in the memory 130 for the target point C and the transmitting transducer Te and the receiving transducer Tr, and to extract the focused signal excerpts corresponding to the target point C of the region of interest ROI from the memory 130. We call these signal excerpts time-shifted by the focus delays received phased signals, usually denoted as sp.

[0111] If the target point C is located at the position of the discontinuity, the above extraction of the received phased signal (time-shifted by the focus delay) will have the effect of returning to Figure 4 The pulse peaks of the N receiving transducers Tr with temporal coherence represented in , which will allow to obtain a focused signal with large amplitude by summing (corresponding to the beamforming A-scan signal of the prior art), which means that the discontinuity at the spatial position of the target point C is detected.

[0112] If the target point C is not located at the location of the discontinuity, the phased signal will not be as Figure 4 The pulse peaks in are vertically aligned, and the summation of these phased signals will not form a focus signal with a large amplitude, which indicates that the discontinuity at the spatial position of the target point C was not detected.

[0113] For each transmitting transducer Te can be calculated Figure 4 The received phased signal has a time offset te.

[0114] Figure 4 Therefore, the 64 received phased signals sp1 to sp 64 , again to avoid overcrowding the graph in this figure.

[0115] These received phased signals sp are then supplied directly to the summing block 200 , or are supplied to the interference reduction block 150 (optional), which corrects these signals before supplying them to the summing block 200 .

[0116] We will therefore first describe the summing block 200, which performs the processing of the received phased signal in order to detect discontinuities.

[0117] The general principle is a double summation: the phased receive signal sp is first summed according to the index of the N receive transducers Tr in order to obtain a focused signal sf for each excitation of the transmit transducer Te (i.e. for each time offset te), and then the focused signal sf is secondly summed according to the index of the Ne transmit transducers Te.

[0118] Figure 5 is an example of a functional block diagram of a summation block 200 including various processing blocks that we will further explain.

[0119] In the example shown, summation block 200 includes:

[0120] a first adder 210 which sums the N received phased signals sp of a particular transmitting transducer Te in order to provide a focus signal sf for the transmitting transducer Te,

[0121] a second summer 220 which sums the Ne focus signals sf of each transmitting transducer Te in order to provide a composite signal, and

[0122] - A filter and envelope detection block 240 for determining the level of the composite signal denoted ss.

[0123] Optionally, the summation block 200 further includes an artifact reduction block 230 (also in Figure 5 denoted as “ARB” in FIG. 1 ), which is a block whose function is to reduce background noise from the processing of system 100 .

[0124] According to a first embodiment, the first adder 210 and the second adder 220 simply sum the signals received as inputs. Figure 4The N received phased signals sp are summed to provide a focused signal sf (for all transmitting transducers Te). A second summer 220 sums the focused received signals sf of all Ne transmitting transducers to provide a composite signal ss.

[0125] Figure 6 An example of the focus signal sf for all transmitting transducers Te is shown.

[0126] Similar to the received signal, the focused received signal sf comprises Ne return pulses transmitted Ne times in the case of a single discontinuity. The time intervals between these pulses depend on the time distribution of the Ne transmitted pulses and the spatial position of the target point C.

[0127] However, if the delay calculation does not correspond to focusing on the target point corresponding to the discontinuity, the focused signal sf will have a different appearance with a mixture of pulses of low amplitude spread out in time.

[0128] Therefore we Figure 6 It is assumed in φ that the phased signal sp is correctly focused on the discontinuity in the medium M, and we will maintain this assumption for subsequent figures to facilitate understanding and make the representation simpler.

[0129] Figure 7 An example of four focus signals sf for each of the transmit transducers Te selected for the transmit sequence and, for example, the transmit transducers with indices 5, 26, 40 and 61 as previously selected, i.e., focus signals sf5, sf 26 、sf 40 and sf 61 These focus signals are transmitted from memory to summing block 200 in parallel or serially, depending on the hardware implementation chosen (eg, the amount of memory available in the hardware), with general operation controlled by controller 300.

[0130] The second summer 220 thus takes all Ne focus signals sf from each transmission of the transmitting transducer and adds them together to obtain Figure 8 The synthetic signal ss.

[0131] In the case of focusing at the target point C corresponding to the discontinuity, the sum of the focused signals obtained for the individual transmitting transducers Te will be combined with good coherence at the combining instant ts to produce e.g. Figure 8 The synthetic signal ss in .

[0132] This composite signal comprises a peak at this combining instant ts, the value of said peak representing the level of the echo Re of the target point in question C. However, even in this case, the composite signal ss is a signal comprising the sum of pulses forming background noise b in the signal.

[0133] According to Figure 5 In a second, more desirable embodiment presented in , the signals are also summed as explained in the first embodiment, but a phase calculation is also performed.

[0134] Thus, the first adder 210 sums the received phased signals to provide a focus signal, and also sums the phases of these signals to provide a focus phase signal intended to be used to estimate the extent to which the received phased signals are "in phase" (correlated) or not "in phase" (decorrelated) with each other. Thus, the first adder 210 supplies a first signal as the focus signal (A-scan signal), and a second signal as the focus phase signal.

[0135] One way to calculate this phase sum is to sum the signs of the received phased signals. If the signals are in phase, they have the same sign and the sum of the signs has a larger absolute value. If the signals are not correctly in phase, the sum of the signs has a lower absolute value. Other ways to calculate the phase sum can be developed by those skilled in the art.

[0136] A second summer 220 first sums the focused receive signals of all Ne transmitting transducers to provide a composite signal.

[0137] Figure 5 The second summer 220 also sums the focused phase signals of all Ne transmitting transducers to provide a composite phase signal.

[0138] As explained above for the phase signal, the composite phase signal has a larger absolute value when the signals are in phase and therefore correspond to a discontinuity.

[0139] Therefore, the artifact reduction block 230 placed at the output of the second adder 220 uses the synthesized phase signal to correct the synthesized signal and improve its signal-to-noise ratio.

[0140] One basic approach is to multiply the composite signal by the absolute value of the composite phase signal normalized between zero and one. Thus, if the phased signal is "in phase," the composite signal's amplitude is not modified, and if it is not "in phase," its amplitude is reduced. With this arrangement, a composite signal is obtained with extremely accurate amplitude and reduced background noise.

[0141] Those skilled in the art may use other adjustments, normalizations, and combinations of the composite signal and the composite phase signal to obtain the corrected composite signal css.

[0142] Finally, the filter and envelope detection block 240 collects the synthesized signal ss directly from the second adder 220 (in the case of the first embodiment) or the corrected synthesized signal css from the artifact reduction block 230 (in the case of the second embodiment). The filter and envelope detection block 240 extracts the peak of this signal from the input signal in order to determine the level of the echo Re of the target point C in question. This level allows the presence of a discontinuity at the location of the target point C to be estimated.

[0143] Therefore and in summary, the system 100 according to the present disclosure implements a method comprising the following steps:

[0144] - Define a transmission sequence where:

[0145] A plurality of transmitting transducers Te are selected from among the transducers of the probe, each transmitting transducer of the plurality of transmitting transducers having a determined spatial position, such that the spatial positions of the plurality of transmitting transducers are uniformly and randomly distributed over the active surface of the probe, and

[0146] defining a time offset te for each transmitting transducer Te of the plurality of transmitting transducers such that the time offsets of the plurality of transmitting transducers are uniformly and randomly distributed within a predetermined transmission duration,

[0147] - transmitting said transmit sequence in a medium by said plurality of transmit transducers,

[0148] - receiving and storing reception signals sr by said plurality of transducers in response to a transmission sequence transmitted in said medium,

[0149] - Process the received signal according to the following procedure, wherein:

[0150] for each transmitting transducer, calculating the focusing delay corresponding to the determined focal law of said transmitting transducer for a target point in the desired zone of interest detected in the medium and taking into account the time offset of the considered transmitting transducer,

[0151] The focus signal sf is calculated for each transmitting transducer as the sum of the received signals of the multiple transducers of the probe realigned by the focusing delay,

[0152] Calculate the composite signal ss, which is the sum of the focused signals from all transmitting transducers, and

[0153] The composite signal is analyzed in order to derive thus the detection level in said medium at the target point, and thus the detection of the discontinuity.

[0154] The transmission sequence of the above method is sufficient for processing the received signal to detect the product medium in order to detect discontinuities in the region of interest ROI at any target point or in any direction. Therefore, this transmission sequence is very effective in quickly detecting the medium and avoiding or reducing interference in the received signal sr.

[0155] The overall process of processing the received signal sp in the processing unit 20 is described here by dividing it into functional blocks, which are composed of at least the delay calculator 115, the summation block 200, and the controller 300, surrounded by other functional blocks, some of which are optional, but this process can be implemented in one or more computing units depending on the architecture defined by a person skilled in the art. Specifically, this process can be implemented in dedicated hardware such as an FPGA processor, in computing hardware such as a DSP processor, or in a standard processor. The capabilities of today's FPGAs, DSPs, and microprocessors allow for extremely fast calculations.

[0156] By repeating the processing procedure for a plurality of target points C based on the same received signal sr recorded after a single transmission sequence as described above, it is possible to construct an image of the medium.

[0157] Therefore, in a method for detecting discontinuities, the processing steps for the received signals are iterated for a plurality of target points in order to create an image of the medium representing different levels of detection.

[0158] Because this processing of the received signal is purely computational, it can be extremely fast. Scanning the product medium M to detect discontinuities can be very fast, which makes it possible to quickly detect products. For example, it is now possible to detect products that are moving relative to the probe (or vice versa).

[0159] We will now explain Figure 1 The operation of the interference reduction block 150 (referred to as "IRB") of the system 100 is described in detail. The specific goal of the interference reduction block 150 is to reduce the interference that may occur when several discontinuities are located within a region of interest ROI of a medium M. We consider the number of discontinuities in the medium to be Nd.

[0160] To simplify the explanation, we assume in this section that the excitation sequence consists of only one pulse emission from a single transmitting transducer Te (Ne=1), but there are three discontinuities inside the medium (Nd=3).

[0161] Therefore, in this case, Figure 9 (Similar to Figure 3 ) represents the N received signals sr obtained after transmission from a single transmitting transducer Te and stored in the memory 130 in the presence of three discontinuities in the medium M of the product. Figure 9Display 64 received signals sr1 to sr 64 Simplified representation of some of the received signals in order to avoid too high a density of curves in this figure.

[0162] This figure shows, for each receiving transducer Tr, a received signal sr containing three pulses corresponding to three echoes from three discontinuities. Each discontinuity has a different spatial location in the medium, which is Figure 9 The three curves representing the pulse peaks in the direction of N transducers (one for each discontinuity) are represented: curves C1, C2 and C3. Figure 3 curves of multiple pulses emitted in the Figure 9 The interference at the intersection points Pa and Pb in .

[0163] Of course, it will be appreciated that when there are several transmit pulses (Ne), this doubles the number of intersection points and the interference between the various curves.

[0164] Therefore, if the portion of the signal around the curve C1 is extracted by the delay calculation method explained above, it is possible to draw a curve with a peak M1 that passes through all the peaks of the curve C1 on average for the target point C focused on the discontinuity corresponding to the curve C1, or in fact use the value of the received signal sr obtained at the time of the reception delay law (delay calculation) corresponding to the target point C on the first discontinuity.

[0165] In particular, the curve of this peak M1 shows disturbances around the indices of the transducers corresponding to the intersection points Pa and Pb.

[0166] The curve of this peak M1 is related to the shape and size of the discontinuity.

[0167] It is then possible to calculate a model curve CM which approximates the curve of the peak M1 by means of a mathematical equation, for example a polynomial equation.Techniques for approximating experimental curves by polynomial regression are well known.

[0168] Therefore, a model curve CM of the curve closest to the peak M1 is determined, the model curve being a curve calculated by a polynomial equation, and the corrected received signal csr is calculated according to the model curve CM, which allows for significant reduction of interference.

[0169] The polynomial equation is, for example, a first-order (straight line), second-order (parabola) or third-order polynomial equation.

[0170] The determination of the closest curve may be performed by minimizing the distance between the model curve CM and the curve of the peak M1 .

[0171] Thus, in summary, the method for detecting discontinuities is done before calculating the focus signal, which is a step implemented, for example, by the interference reduction block, and during which the interference between several discontinuities is reduced by a process comprising the following operations:

[0172] - determining a curve of peaks in the received signal as a function of the transducer, said curve being determined by identifying the peaks in the received signal at the target point and according to the focal law,

[0173] - calculating a model curve of the curve approximating the peak, and

[0174] - Calculating a corrected received signal from the recorded received signal and the model curve, and then processing the signals to derive the detection level using the corrected received signal instead of the recorded received signal.

[0175] We will now explain the dynamic speed corrector 116 ( Figure 1 The purpose of the dynamic speed corrector 116 is to correct the delay calculation in the case where the probe 10 moves relative to the product (medium M) (or vice versa). This involves the case of movement at a constant speed in the longitudinal direction X. This is a very common case for continuous inspection of metal sections or, for example, railway tracks. The speed can therefore be extremely high (tens of kilometers per hour). Without correcting the delay calculation, the estimation of the detection level is completely wrong and, in the case of image calculation, the resulting image is completely blurred.

[0176] For example, in the case of the delay calculation detailed above, the corrected focus delay calculation is thus still:

[0177] Delay=te+tem+tre+tadj

[0178] However, the outbound travel time tem and return travel time tre are now:

[0179] tem=1 / Va.sqrt(((xe+dxm)-xc) 2 +zc 2 )

[0180] tre=1 / Vr.sqrt(((xr+drm+drA+drR)-xc) 2 +zc 2 )

[0181] in

[0182] dxm is the movement of the transmitting transducer Te during the time offset te of the transmitting transducer,

[0183] drm is the movement of the receiving transducer Tr during the time offset te of the transmitting transducer Te,

[0184] drA is the movement of the receiving transducer Tr during the outward travel of the wave between the transmitting transducer Te and the target point C,

[0185] drR is the movement of the receiving transducer Tr during the return travel of the wave between the target point C and the receiving transducer Tr.

[0186] The aforementioned focusing delay is therefore now a function of the movements dxm, drm, drA, drR, which can be easily calculated as a function of the movement speed of the product relative to the probe 10 .

[0187] The above delay calculation can be expressed as a memory index in the memory 130 by multiplying the outgoing travel time tem, the return travel time tre and the time offset te by the sampling frequency Fs in the case of a system sampled at this constant sampling frequency.

[0188] Thus, the system 100 retrieves a velocity measurement of the medium relative to the probe from another system, or includes a sensor for measuring this velocity. This velocity is supplied to a dynamic velocity corrector 116 which corrects the delay calculator 115, for example using the formula explained above.

[0189] The method for detecting discontinuities is thus improved by the fact that the focusing delay is determined as a function of the speed of movement of the medium relative to the probe (or vice versa).

Claims

1. A method for detecting a discontinuity in a medium, said method being implemented using a probe comprising a plurality of transducers forming an active surface and capable of transmitting and receiving ultrasonic waves in said medium, said method comprising the following steps: - Define a transmission sequence where: A plurality of transmitting transducers are selected from among the transducers of the probe, each transmitting transducer of the plurality of transmitting transducers having a determined spatial position, such that the spatial positions of the plurality of transmitting transducers are uniformly and randomly distributed over the active surface of the probe, and defining a time offset for each transmit transducer of the plurality of transmit transducers such that the time offsets of the plurality of transmit transducers are uniformly and randomly distributed within a predetermined transmit duration, - transmitting said transmission sequence in said medium by said plurality of transmitting transducers, - receiving and recording received signals by said plurality of transducers in response to a transmission sequence transmitted in said medium, - processing the received signal according to the following procedure, wherein: for each transmitting transducer, calculating, for a desired target point detected in the medium and taking into account the time offset of the considered transmitting transducer, a focusing delay corresponding to the determined focal law of the transmitting transducer, Calculate for each transmitting transducer a focus signal which is the sum of the receive signals of the multiple transducers of the probe realigned by the focus delay, Calculate the composite signal, which is the sum of the focused signals from all transmitting transducers, and The composite signal is analyzed to derive the detection level in the medium at the target point, and hence the detection of the discontinuity.

2. The method according to claim 1 , further comprising the step of reducing interference between a plurality of discontinuities before calculating the focus signal by a process consisting of: - determining a curve of peaks in the received signal as a function of said transducer, said curve being determined by identifying peaks in the received signal at target points and according to said focal law, - calculating a model curve of the curve approximating the peak, and - calculating a corrected received signal from the recorded received signal and the model curve, and then processing the signals using the corrected received signal instead of the recorded received signal to derive the detection level.

3. The method according to claim 2, characterized in that The model curve is a polynomial curve.

4. The method according to any one of claims 1 to 3, characterized in that The focusing delay is determined as a function of the speed of movement of the medium relative to the probe or vice versa.

5. The method according to claim 1, wherein The processing of the received signals is iterated for a plurality of target points to create an image of the medium representing different levels of detection in the target points.

6. A system for detecting discontinuities in a medium, comprising a probe (10), a processing unit (20) connected to the probe, and a controller (300) for implementing the method according to any one of claims 1 to 5, wherein the probe comprises a plurality of transducers capable of transmitting and receiving ultrasonic waves in the medium, and the processing unit (20) comprises at least one memory (130) for storing received signals.

Images