Classifications

• • G—PHYSICS
  • G01—MEASURING; TESTING
  • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
  • G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
  • G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
  • G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
  • G01N27/90—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws using eddy currents

Definitions

• An eddy current imaging method and device for the detection and characterization of defects embedded in complex structures An eddy current imaging method and device for the detection and characterization of defects embedded in complex structures.
• the present invention relates to a method and an eddy current imaging device for the detection and characterization of defects buried in complex structures, particularly in the aeronautical and nuclear fields.
• Early detection and fine characterization of defects, for example cracks, in structures such as aeronautics are a major issue for the safety and maintenance of aircraft.
• the eddy current detection using electromagnetic sensors is made difficult because of the complexity of the structure which generates parasitic signals preventing easy reading of the images, the low penetration of eddy currents due to the skin effect, the low spatial resolution of commonly used sensors, and the small possibilities of measurement configuration that conventional sensors allow.
• crack detection in riveted joints is done locally using point eddy current sensors or using 4 or 5-element structures that operate differentially over a limited observation area of the order of magnitude. 10 mm in diameter. With these devices, defects are sought at the foot of the rivets, at a second plate for example a riveted assembly of aircraft sheets, about 1.6 mm deep.
• the objective problem is the weak performance offered by the currently used imaging methods, for the detection and fine localization of defects deeply buried in a complex structure.
• An object of the invention is to propose an imaging method making it possible to determine the position and the dimensions of defects deeply buried in a complex structure, for example more than 1.5 mm deep in a riveted assembly of metal sheets. plane.
• the invention relates to an eddy current imaging method for the detection and characterization of defects buried in a complex structure comprising the steps of:
• the generating step comprises generating a set of at least two field waveforms exciter magnet in the target material, each distinct waveform being determined by a narrow-band frequency spectrum and an orientation angle in the target material
• the measuring step comprises measuring a set of configurations of the resulting magnetic field , in the form of a set of images, each image being associated with a magnitude and a waveform of the exciter magnetic field, and in that it comprises a step of processing all the images by combination, to detect a fault and determine its location and type.
• the method comprises one or more of the following characteristics taken in isolation or in any technically possible combination:
• each waveform is orientable according to a viewing angle determined by the location and geometry of a defect
• the frequency band of each waveform is chosen according to a desired depth of observation in the target material
• the set of waveforms of the exciter magnetic field is a time-division multiplex of the waveforms;
• the set of waveforms of the exciter magnetic field is a frequency multiplex of the waveforms;
• the waveforms have about the same energy
• each waveform is a sinusoid of different frequency
• an image of the resulting complex magnetic field is a plurality of values of a real or imaginary spatial component of the resulting magnetic field, picked up near several points of the inspection zone;
• the step of processing the raw images comprises: a learning step consisting in constructing a projection operator adapted to the detection of predetermined defects type from raw images acquired in an acquisition step on a structural standard known device comprising calibrated fault types according to a chosen measurement configuration, - a projection step of applying the projection operator determined in the previous step to the raw images acquired in the acquisition step on a test sample on which we are looking for defects according to the same predetermined configuration used in the previous step to obtain usable useful images;
• the step of constructing the projection operator comprises the steps of: choosing a row from a crossing line of the structure and buried defects of the standard,
• the step of applying the projection operator comprises the steps of:
• the method further comprises the step of:
• the filtering of the useful images is a filtering of the type belonging to the family consisting of Wiener integration, deconvolution and filtering methods; the method further comprises the step of:
• the classification and fault diagnosis step is a process of the family of the methods consisting of the threshold decision, maximum likelihood decision and neural network decision methods;
• the eddy current imaging method further comprises a step of constructing a database on the basis of a set of standards;
• the excitation magnetic field generation means are capable of generating a set of at least two excitatory magnetic field waveforms in the target material, each distinct waveform being determined by a band frequency spectrum narrow and an orientation angle in the target material,
• the measurement means are able to measure a set of configurations of the resulting magnetic field, in the form of a set of images, each image being associated with a magnitude and a waveform of the exciter magnetic field, and the image processing means are able to process all the images by combination in order to detect a defect and to determine its location and its type.
• the device comprises one or more of the following characteristics taken separately or in any technically possible combination:
• the measuring means comprise a network of type sensors belonging to the assembly consisting of coils, micro-coils, Hall effect probes, GMRs (magneto-resistance with a giant effect) and GMIs (magneto-impedance) giant effect); and
• the measuring means comprise a linear magneto-optical imager consisting of an optical device, a linear magneto-optical garnet and photo-detector means.
• FIG. 1 is a general view of an embodiment of the eddy current imaging device according to the invention
• FIG. 2 is a flowchart of one embodiment of the imaging method according to the invention.
• FIG. 3 is a thickness section of a test sample comprising a riveted structure with defects
• FIG. 4 is a section through the thickness of a standard of a riveted structure with a number of calibrated defects;
• FIG. 5 is a flowchart of an embodiment of the imaging treatment method according to FIG. 'invention,
• FIG. 6 is a view of the set of raw images obtained during the measurement carried out on the calibrated standard with the imaging device of FIG. 1;
• FIG. 7 is a view of all the raw images; obtained on the test sample during measurements with the imaging device of Figure 1, 7
• FIG. 8 is a view of all the image components obtained after the processing step of FIG. 5, and
• FIG. 9 is a flowchart of a variant of the processing method of FIG. 5 in which a database of several calibrated standards is constituted and used.
• FIG. 1 diagrammatically represents an embodiment of an eddy current imaging device 2 according to the invention.
• the imaging device 2 is disposed on the surface of a target material 4 and comprises an eddy current inductor 6, powered by a generator 8 of alternating currents, a field measuring device 10, here a magneto-optical imager operating over a wide area (12) of inspection and synchronized with the inductor 6 by a dedicated digital synchronization and control board.
• the eddy current imaging device 2 also comprises a computer 14 for controlling the different imaging and image processing tasks.
• the inductor 6 comprises a magnetic circuit consisting of a first and a second magnetic poles 15, 16 for circulating a uniform field H exc oriented along an axis parallel to the surface of the target material 4, the axis y being a component of a reference trihedron (x, y, z) with the axis z defining the normal to the surface of the inspection zone (12) of the target material.
• the inductor 6 circulates the uniform field H exc from one pole to the other thanks to the presence of a first and a second excitation coil 18, 20 traversed by alternating currents I delivered by a generator 8 .
• the excitation windings 18, 20 are powered by alternating currents I of sinusoidal waveforms and of adjustable frequency in a wide range (for example 10 Hz, 10 MHz), provided by the current generator 8, consisting of a DC power source 24, a DC inverter 26 and a cycloconverter 28.
• the current generator 8 is connected to the windings 18, 20 via two branches of a first power cable 32 connected to the side of the magnetic pole 15 and two branches of a second power cable 34 connected to the side of the second magnetic pole 16. 7,000791
• the frequency-adjustable alternating current generator 8 is provided with an input 35 for supplying the frequency setpoint of the waveform of the current signal.
• a second inductor not shown here, oriented perpendicular to the first inductor can be added.
• the measuring device 10 of the magnetic field is here of linear magneto-optical type. It comprises an optical device 36, a magneto-optical material 38 and photo-detector means 40.
• the optical device 36 comprises a light source 42, a polarizer 44 and an analyzer 46.
• the polarizer 44 and the analyzer 46 are conventional and known to those skilled in the art.
• the light source 42 is constituted by a matrix of light-emitting diodes.
• the optically active material here a linear magneto-optic garnet, is interposed between the polarizer 44 and the analyzer 46 on the optical path and disposed near the surface of the inspection zone of the target material 4.
• the polarizer assembly 44, magneto-optical garnet 38, analyzer 46, constitutes a magneto-optical light modulator.
• the photo-detector means 40 are here an analog CCD camera associated with a video acquisition card.
• the measuring device 10 is a network of type sensors belonging to the assembly consisting of coils, micro-coils, Hall probes, GMRs (magneto-resistance with a giant effect) and GMIs ( magneto-impedance with giant effect), and other types of magnetic sensors.
• the computer 14 comprises an interface 50, for example of the USB type, for receiving the video data provided by the acquisition card into a video input 51.
• the computer 14 also comprises a visual display 52, here a liquid crystal screen, a processor 54 for image processing and coordination of imaging tasks, the processor 54 being coupled to a database 56 in the form of memories of the type classic.
• a visual display 52 here a liquid crystal screen
• a processor 54 for image processing and coordination of imaging tasks, the processor 54 being coupled to a database 56 in the form of memories of the type classic.
• the computer 14 also has an output 58 for controlling the dedicated digital synchronization and control board.
• two waveforms 62, 64 of sinusoidal currents are produced by the current generator 8, each waveform 62, 64 having an associated frequency f 1, f 2 and form a time multiplex 60 of two sinusoids of different frequency. and almost equal energy.
• f1 is equal to 100 Hz while f2 is equal to 700 Hz.
• waveforms are sinusoids so as to each have a narrow band spectrum.
• the number of waveforms is between 3 and 25.
• the two current waveforms are emitted simultaneously and form a frequency multiplex with two sinusoids of distinct frequency.
• FIG. 2 represents an embodiment of the imaging method used and in accordance with the invention.
• the imaging method 70 comprises a set of successively executed steps.
• a first step 72 the imaging device 2 is positioned near the zone 12 of the target material 4 containing a sample to be tested.
• the alternating current generator 8 generates the first sinusoidal current waveform 62 at a frequency f1 of 100 Hz, which makes it possible to generate the excitation field H exc at the level of the inductor 6 oriented along the y-axis of Figure 1.
• a current ply referenced JF in FIG. 1 is locally induced uniformly and oriented along the x-axis in FIG. 1 in a wide zone (12) of inspection, here greater than 75 mm in diameter.
• the JF currents when they encounter a buried structure of electromagnetic impedance distinct from that of the homogeneous plate material deviate their trajectory and their distribution, creating a disturbance magnetic field modulating the resulting magnetic field, here oriented along the z axis , as illustrated in Figure 1 which shows a circular rivet structure.
• each waveform is orientable and can be adapted to an optimal angle of view of the tested structure.
• a step 76 the measurement of the resulting complex magnetic field at the surface of the inspection area (12) is performed.
• B z (x, y) is measured. This measurement must be obtained with a sufficient Shannon spatial sampling step in both x and y directions and can be obtained in module or in phase, or in part real or imaginary.
• the alternating current generator 8 generates a second sinusoidal waveform of current at the frequency f2 of 700 hertz for generating the excitation field H ⁇ ⁇ C -
• the field resulting complex B z (x, y) is then measured.
• the set of raw images obtained for the different frequencies is representative of phenomena appearing at depths different f1 and f2 (skin effect) and is therefore different views of the same situation.
• a frequency band of each waveform is selected according to a desired depth of observation in the target material (4).
• the set of digitized data, forming the four raw images Image 1, Image 2, Image 3, Image 4 is stored in the computer 14, then processed in a step 90 for processing the raw images.
• the complex field components are digitized with an amplitude of at least 2 bits, ie a dimension vector. greater than or equal to 2.
• the amplitude of a measured complex magnetic field component B z (x, y) is coded on 12 bits.
• the measurement of the complex field is done in the context of a synchronous detection successively given to the different excitation frequencies of the inductor 6 and coordinated by means of the dedicated digital synchronization card 13 and ordered.
• each measurement is given to one of the frequency components constituting the alternating excitation signal supplying the inductor 6 (this corresponds to the case of frequency multiplexing).
• the processing 90 of the raw images will be described here in detail based on raw images of a test sample, and based on the characterization of a standard whose structure and defects are known a priori.
• FIG. 3 represents an example of a test sample 100 comprising a riveted structure composed of three plates or plates 102, 104, 106 of aluminum alloy (conductivity of 20 MS / m, relative permeability equal to
• the number of rivets shown here is 10 and the rivets 108 are numbered # 1 to # 10 from left to right in FIG.
• a first and second cracks 114, 116 of type 2, respectively 7 mm and 12 mm long for 100 ⁇ m of opening are placed on the third plate 106 and buried at 6 mm at the rivets numbered respectively # 8 and # 9.
• Figure 4 shows in section the structure of a standard 120, having a set of five rivets 128 spaced respectively 25 mm numbered from # 1 to # 5 from left to right in the figure, and assigned two types of defect.
• the structure 120 is composed of a stack of three metal plates 3 mm thick with a first plate 122, a second and third plates 124 and 126.
• the three plates are made of aluminum and have a conductivity of 20 MS / m and a relative permeability equal to 1.
• a first crack 130 of 12 mm length said type 1 is buried in the second plate 124 at the foot of the rivet # 3 and a second crack 132 of length 12 mm said type 2 is buried at the bottom of the rivet # 4 in the third plate 126.
• Figure 5 depicts the method of processing images completely.
• a first step 142 the etalon 120 as described in FIG. 4 is manufactured and has characteristics identical to the structure to be controlled, that is to say a stack of three alloy plates. riveted by rivets regularly spaced 25 mm.
• step 144 calibrated defects are placed, of the same type as those to be detected on the structure of FIG. 3 and described in FIG.
• a learning step 146 follows, in which one builds according to the construction step 152 a projection operator, adapted to the implementation evidence of the desired defects of type 1 and 2, from the raw images 150 acquired according to the acquisition step 148 on the standard 120 with riveted joints including the calibrated defects, known a priori.
• the set of raw images 150 of the standard is described schematically in FIG.
• step 154 the sample 100 of riveted seals to be tested is positioned.
• step 156 according to the application step 162, the projection operator determined in step 152 is applied to the raw images 160 acquired according to the acquisition step 158 under the same experimental conditions as the Step 148.
• this step 156 which is a deconvolution step, a new series of raw images is output of 162, some of which contain information on the defects to be detected.
• a filtering 166 is performed making it possible to better visualize the fault information provided at the output of step 162 in the form of a set of components 170 as well as a diagnostic step 172 for taking a decision. on the classification of detected faults.
• the diagnostic results are displayed.
• the measurement channels are chosen in sufficient number to place the three interesting sources of perturbations of the resulting field: rivets
• the second frequency f2 makes it possible to see the type 1 defect together with the rivet
• the first frequency f1 makes it possible to see the type 2 defect with the rivet.
• the convolution operator here projection, is constructed from the set 150 of four raw standard images schematically described in Figure 6.
• the first raw image 176 is an image of the real part of the complex result field at the frequency of 100 Hz while the second raw image 178 is the imaginary part of the complex result field at the frequency of 100 Hz.
• the third raw image 180 is the real part of the measured field when the induction has an excitation frequency of 700 Hz while the fourth raw image 182 is the imaginary part of the field for an excitation frequency at 700 Hz.
• first raw image 176 On the first raw image 176 are represented the positions of two rivets 184, 188 near which are the defects namely the rivet # 3 and the rivet # 4, the first 184 being assigned a crack 186 type 1 and the fourth rivet 188 being affected by a crack 190 of type 2.
• These defect structures are not visible clearly on the rough images generally. They are given here as a reference, on the basis of a priori knowledge, to allow the construction of the projection operator.
• M ⁇ 7 (O denotes the transpose of the matrix of the reference measure of line rank i, M ref (i).
• the four eigenvectors of this variance-covariance matrix are then determined in a 4 ⁇ 4 projection matrix V pr0 j.
• V pr0 j By projection of IVU f (O on each projection vector according to the operation V P ro j .M re f (i), a new matrix is obtained whose four lines or components respectively correspond to the most important physical sources of disturbance. contained in the four lines of M rf (i)
• M rf (i) Here, only three sources are relevant: rivets, crack in sheet 2, crack in sheet 3. Knowing the position a priori of the reference cracks on standard 120 used to build V pr0j , it is allowed to identify which components correspond to the eigenvectors.
• the eigenvalues correspond to the energies emitted by the different types of defect. If we choose to order the components in the order: component 1-sharet, component 2-crack sheet 2, component 3-crack sheet 3, component 4-nothing, the learning is completed.
• the acquisition step 158 consists of making measurements on the test sample 100 under the same experimental conditions of the step 148 for the standard 120.
• the set of raw images 160 comprises a first raw image 192 of the tested sample 100 when the exciter field H eX c is at a frequency of 100 Hz and the real part of the magnetic field according to the component z is measured.
• the second raw image 194 represents the imaginary part of the same complex field for an exciter field at a frequency f1 of 100 Hz.
• the third raw image 196 represents the real part of the magnetic field measured for an excitation frequency f2 of 700 Hz.
• fourth image 198 finally represents the imaginary part of the magnetic field measured for an excitation field of frequency f2 of 700 Hertz.
• the application operation is repeated for all lines of rank i of each image (from rank 1 to rank 350 in this example).
• step 166 the filtering of the data obtained is performed after the deconvolution operation 162, that is to say the projection of the measurement matrix. This filtering 166 is performed in order to bring out this deconvolution in a more usable form for the direct realization or to implement one of the diagnostic methods of step 172.
• the filtering 166 may be a deconvolution, a Wiener filter or well adapted filtering.
• an integration is performed consisting of the sum of the accumulated samples from left to right of the image, line by line.
• the set 170 of the four usable useful images or components is obtained.
• the set 170 usable useful images or components is shown in Figure 8 which corresponds to actual measurements.
• These usable useful images obtained after filtering 166 represent a first component 202 on which the rivets alone are visible, a second component 204 on which are visible the cracks of the second plate or type 1 defects, a third component 206 in which buried cracks appear. at the level of the third plate or type 2 defects, a fourth component 208 showing a priori no particular structure.
• the diagnostic aid step 172 uses a thresholding, maximum likelihood or neural network type technique. Thus, an automatic decision on the classification or location of defects can be made more easily.
• the example described below has only one calibration standard serving as a reference base. 007/000791
• a step 212 set between 146 and 162, of forming a multi-standard database having different calibrated defects, constructs a set of relevant projectors associated with the corresponding measurement conditions.
• step 212 The data obtained from these additional test configurations and stored in step 212 makes it possible to enrich the measurement system performed in step 162 on the test sample and the reference signature database that can be used for the test. performing the diagnostic assistance step 172.

Landscapes

• Chemical & Material Sciences (AREA)
• Chemical Kinetics & Catalysis (AREA)
• Electrochemistry (AREA)
• Physics & Mathematics (AREA)
• Health & Medical Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Analytical Chemistry (AREA)
• Biochemistry (AREA)
• General Health & Medical Sciences (AREA)
• General Physics & Mathematics (AREA)
• Immunology (AREA)
• Pathology (AREA)
• Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=37547461

Family Applications (1)

Country Status (5)

Cited By (1)

Families Citing this family (11)

Family Cites Families (8)

• 2006
  • 2006-05-12 FR FR0604255A patent/FR2901025B1/fr active Active
• 2007
  • 2007-05-10 US US12/300,514 patent/US8159216B2/en not_active Expired - Fee Related
  • 2007-05-10 CA CA2651718A patent/CA2651718C/fr not_active Expired - Fee Related
  • 2007-05-10 EP EP07731432A patent/EP2024738A1/de not_active Withdrawn
  • 2007-05-10 WO PCT/FR2007/000791 patent/WO2007135265A1/fr active Application Filing

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events