TOMOGRAPHIC ANALYSIS PROCEDURE

Description

Technical field of the invention

[0001] The invention relates to a method for tomographic analysis of a mechanical part. The part is, for example, part of an aircraft, and more specifically of a turbomachine, such as part of a fan casing, a fan blade, or part of a fixed blade structure. Prior art

[0002] It is known to inspect structural mechanical parts using tomographic methods, for example radiographic or acoustic, in order to detect the presence of possible defects on the surface and inside the part.

[0003] These methods are reliable, non-invasive and allow for inspection of the inside of the parts, which makes it possible to quickly determine the condition of the part and decide whether to potentially put it back into operation or replace it.

[0004] A tomographic measurement consists of scanning an observed object, here a mechanical part, using a beam of waves, and measuring the transmitted beam in all directions in order to reconstruct a three-dimensional image of the object.

[0005] There figure 1 illustrates a tomography device 1 during the inspection of a mechanical part 20.

[0006] The tomography device 1 includes at least one emitting device 3, configured to emit an incident beam 5 of wave pulses towards the part 20, for example radio frequency waves, X-rays or acoustic waves, and at least one receiver 7 capable of capturing a transmitted beam 9 of waves, arranged on either side of the part 20.

[0007] Part 20 is generally mounted on a support 11 rotating around an axis A, so that it can be observed from all directions during tomographic measurement.

[0008] Part 20, observed by tomography, may be made of a composite material, as shown in detail on the figure 2 , or a metallic material.

[0009] Such a composite material comprises weft fibers 21 and warp fibers 22 woven together on a weaving plane P, or several weaving planes P superimposed along a thickness direction Z.

[0010] The fibers 21 and 22 can be carbon fibers, glass fibers or a mixture of the two. They are embedded in a matrix 23 comprising for example one or more polymers and / or resins, which solidify to form the final part 20.

[0011] In the case of a metallic material, it can be defined at the mesoscopic scale as a set of matter grains 32 linked together by grain boundaries 33, as shown on the figure 3These grains 32 are three-dimensional domains that can have diverse and varied shapes, with a preferred direction or orientation over the extent of the grain 32, which can lead to an orthotropic metallic material.

[0012] The amplitudes of the transmitted waves from different observation angles are converted into grayscale levels in a processing unit 13, and digital analysis allows the volume of part 20 to be reconstructed. The processing unit 13 includes a computer. Tomographic analysis then provides a visual representation of the exterior and interior of the part. This type of inspection has the major advantage of allowing visualization through the thickness of the material and the part being studied, while being non-destructive and reliable.

[0013] A human operator can then assess the quality of the part, by searching in the three-dimensional image for the presence or absence of an anomaly or damage.

[0014] In what follows, the terms "non-conformity", "anomaly", "damage", "damage" or equivalents are used interchangeably to refer to a part of the part where the mechanical properties of the part are locally degraded compared to those of a part in good condition.

[0015] It could also be the presence of a foreign body within the material, such as a void (porosity) or material that was accidentally introduced into the part during its forming. Such an anomaly may justify removing and replacing the part when its mechanical strength is compromised, or, depending on the circumstances, may not impede the part's proper functioning. In the case of a composite material, it could also be an area where the fibers 21, 22 are stretched or broken.

[0016] In the case of a metallic material, it may again be an area with porosity or a foreign body, lower quality joints affecting mechanical characteristics, or atypical grain sizes.

[0017] It is important to note that the detection and characterization of the size and shape of the anomaly is currently still left to the human operator. This step generates a significant delay because it is necessary to accurately define the non-conforming area, and observation of the entire volume of the part is currently required to detect all potential anomalies.

[0018] Another tomographic analysis method is described in Dakak Abdel Rahman et al.: "Application of artificial intelligence algorithms to the exploitation of tomography data of aluminum alloy casting parts", The Forge and Foundry Review, vol. 26, June 2021, pages 18-25, XP055965777. Presentation of the invention

[0019] The invention aims to remedy these drawbacks by providing an analysis method that allows for the rapid detection of potential anomalies without requiring operator intervention and without risking overlooking a potential anomaly.

[0020] To this end, the invention relates to a method for tomographic analysis of a part to detect anomalies, the method comprising the following steps: acquisition of at least one three-dimensional image of the part by means of a tomography device, subdivision of the image into elementary sub-parts, analysis of a distribution of grey levels in each sub-part and obtaining at least one parameter representative of said distribution of grey levels for each sub-part, comparison of the parameter(s) obtained for each sub-part with standard value ranges characteristic of a healthy region and detection of abnormal sub-parts whose parameter(s) are outside the standard value ranges, determination of risk regions, including each abnormal sub-part and each sub-part adjacent to at least one abnormal sub-part, and analysis of the risk regions to detect anomalies in the part.

[0021] Such a process makes it possible to detect anomalies in a part in a reliable and non-invasive way, with reduced intervention time for the operator during the analysis stage of the risk areas, without increasing the risk of missing an anomaly.

[0022] The subdivision step may include determining at least one standard dimension of the anomalies sought in the part, with each sub-part having dimensions ranging from half the standard dimension to double the standard dimension.

[0023] The standard dimension is obtained, for example, from the results of previous tests stored in a database.

[0024] The dimensions of the sub-parts may include a height, a width, a length of the sub-part.

[0025] Alternatively, each sub-part can have dimensions greater than twice the standard dimension.

[0026] Such a characteristic makes it possible to optimize the size of the risk region studied according to the anomalies to be considered.

[0027] At least one of the representative parameters of the grey level distribution can be chosen from the mean, maximum and minimum of the grey level distribution over the sub-part.

[0028] At least one of the representative parameters of the grey level distribution can be chosen from among the gradient, curl and divergence of the grey level distribution on the sub-part.

[0029] The comparison step can implement at least one numerical processing tool chosen from nearest neighbor analysis, classification tree analysis, support vector machine analysis and neural network analysis.

[0030] Alternatively, the comparison step may consist of comparing the measured value with a range of standard values ​​characteristic of a healthy region, to determine whether the measured value is within said range or not.

[0031] The range of standard values ​​characteristic of a healthy value can be a range centered on a standard average value and with an amplitude, for example, equal to 10% of this standard average value.

[0032] The part may include a woven composite material.

[0033] The composite material can be woven into a two-dimensional pattern. Such a material comprises warp and weft fibers woven together to form at least one weave plane, said fibers being embedded in a matrix.

[0034] The fibers can be in the form of independent weaving planes pre-impregnated with the uncured matrix, and deposited one on top of the other.

[0035] Alternatively, all weaving plans can be set up, for example in a mold, before the injection of the uncured matrix.

[0036] Alternatively, the composite material can be woven in a three-dimensional pattern, with fibers extending in at least three non-coplanar directions, embedded in a matrix as described above. Such a material is also known as " interlock » .

[0037] The fibers of the woven composite material may include glass fibers and / or carbon fibers.

[0038] The matrix may comprise at least one polymer, in particular a thermosetting polymer, and / or at least one resin.

[0039] The part may include a metallic material.

[0040] The part could be a turbomachine casing component, or a compressor rotor blade, stator or fan blade from a turbomachine. Brief description of the figures

[0041] [ Fig. 1 ] there figure 1 is a schematic side view of a tomography device during the implementation of a method according to the invention, [ Fig. 2 ] there figure 2 is a schematic detail view of a part made of woven composite material, [ Fig. 3 ] there figure 3 is a schematic detail view of a part made of metallic material, [ Fig. 4 ] there figure 4 is a schematic view of a three-dimensional image of a part containing an anomaly, [ Fig. 5 ] there figure 5 is a schematic view of a three-dimensional image subdivision step of the figure 4 , And [ Fig. 6 ] there figure 6 is a schematic view of a step in the analysis of the subdivisions of the figure 5 . Detailed description of the invention

[0042] A tomographic analysis method according to the invention will now be described. This method uses the analysis device 1 described previously, and aims to detect the presence of anomalies in a part 20, made for example of composite material or metallic material.

[0043] The process includes a first step of acquiring at least one three-dimensional image of part 20 using the tomography device 1.

[0044] An incident beam 5 of wave pulses is emitted by the emitting device 3, towards the part 20. The incident beam 5 is, for example, a beam of radio frequency waves.

[0045] The waves pass through part 20, and a transmitted beam 9 from part 20 is captured by the receiver 7. The intensity distribution of the transmitted beam 9 obtained is converted into a two-dimensional greyscale image of part 20 by the processing device 13.

[0046] Part 20 is rotated by means of support 11, and two-dimensional images of part 20 are acquired in all directions.

[0047] A three-dimensional image of part 20 in greyscale, or tomographic image, is then reconstructed by image processing using the processing device 13.

[0048] In the example described, part 20 is a part made of composite material, as shown on the figure 2 .

[0049] The composite material comprises weft fibers 21 extending along a weft direction X and warp fibers 22 extending along a warp direction Y, embedded in a matrix 23. The average gap between two adjacent weft fibers 21 or warp fibers 22 is on the order of 2 mm.

[0050] A thickness of part 20, measured along a thickness direction Z, is for example between 5 and 25 mm, notably close to 10 mm for a part 20 comprising four to eight superimposed weaving planes.

[0051] The weft fibers 21 and warp fibers 22 can be of the same or different materials (glass or carbon).

[0052] Matrix 23 comprises at least one organic polymer and / or at least one resin.

[0053] Woven composite materials can be considered as orthotropic materials, that is, materials possessing three planes of symmetry at the level of their internal microstructure.

[0054] Alternatively, the part 20 is made of a metallic material consisting of metallic grains 32 and including joints 33 between these grains 32. A thickness of the part measured along the Z direction is, for example, between 2 and 25 mm, particularly close to 7.5 mm. The metallic material can be isotropic or orthotropic depending on its internal microstructure.

[0055] The analysis process includes a step of determining the standard dimensions of the anomalies sought in part 20.

[0056] Standard dimensions depend on the nature of the material, as well as the acceptable degradation threshold in part 20.

[0057] For example, in a part 20 intended for the aeronautical field, anomalies that are too small do not cause sufficient degradation of the mechanical properties of part 20 to be taken into account.

[0058] An example of standard anomaly dimensions is 1 mm along the three directions X, Y, Z.

[0059] The process then includes a step of subdividing the three-dimensional image into elementary sub-parts.

[0060] Such a three-dimensional image 25 subdivided into elementary sub-parts 27 is schematically represented on the figure 4 , in which part 20 includes an anomaly 30.

[0061] The elementary sub-parts 27 are for example cubic, with a side size D equal to the standard dimension of the anomalies sought.

[0062] Alternatively, the side size D of subparts 27 is greater than the standard dimension of the anomalies sought, specifically twice greater.

[0063] The process includes a step of analyzing the distribution of grey levels in each sub-part 27 and determining a value of at least one parameter representative of this distribution for each sub-part 27.

[0064] Examples of representative parameters are: the average of the grey levels, the maximum level, the minimum level, but also more complete mathematical operators such as the average gradient of grey level on sub-part 27, or the curl and the divergence of the grey level on sub-part 27. These parameters allow a quick description of the grey level distributions in each sub-part 27.

[0065] The process then includes a comparison step, during which the values ​​of each representative parameter obtained for each subpart 27 are compared with the known values ​​corresponding to subparts known to be without anomalies.

[0066] This comparison step can consist of a simple check of the values ​​of the parameter studied against a range of values ​​identified on the healthy areas of the material, or a more in-depth classification analysis.

[0067] In the case of classification analysis, dedicated numerical processing tools such as nearest neighbor analysis and classification trees can be used. If the data volume is sufficient, machine analysis using vector support or neural networks can be implemented.

[0068] The comparison step then allows the determination of abnormal sub-areas 31, for which the values ​​of the representative parameters differ from the values ​​identified in the healthy areas. The abnormal sub-areas 31 are represented on the figure 6 with cross-hatching.

[0069] The process then includes a step of determining risk regions, including all abnormal sub-parts 31, as well as each sub-part 33 directly adjacent to at least one abnormal sub-part 31.

[0070] Subparts 33 directly adjacent to at least one abnormal subpart 31 are marked with simple hatching on the figure 6 .

[0071] In this way, the region at risk is identified with a margin around the identified abnormal sub-parts, so as not to miss any anomalies.

[0072] Sub-sections 27 not being part of a risk area are excluded and classified as a healthy area.

[0073] Risk areas are then manually inspected by an operator to detect anomalies in the room, while healthy areas are not inspected, thus reducing the intervention time of a qualified operator.

Claims

1. Method for the tomographic analysis of a part (20) in order to detect anomalies (30), the method comprising the following steps: - acquiring at least one three-dimensional image (25) of the part (20) by means of a tomography device (1), - subdividing the image (25) into elementary subparts (27), and being characterized by the following steps: - analyzing a grayscale distribution in each subpart (27) and obtaining at least one parameter representative of said grayscale distribution for each subpart (27), - comparing the one or more parameters obtained for each subpart (27) with standard values characteristic of a defect-free region and detecting abnormal subparts (31) for which the one or more parameters differ from the standard values, - determining risk regions, which comprise each abnormal subpart (31) and each subpart (33) adjacent to at least one abnormal subpart, and - analyzing the risk regions in order to detect the anomalies (30) in the part (20).

2. Method according to claim 1, wherein the subdividing step comprises determining at least one standard dimension of the anomalies (30) being looked for in the part (20), each subpart (27) having dimensions between half the standard dimension and double the standard dimension.

3. Method according to one of the preceding claims, wherein at least one of the parameters representative of the grayscale distribution is chosen among the average, the maximum, and the minimum in the grayscale distribution for the subpart (27).

4. Method according to one of the preceding claims, wherein at least one of the parameters representative of the grayscale distribution is chosen among the gradient, the divergence, and the curl of the grayscale distribution for the subpart (27).

5. Method according to one of the preceding claims, wherein the comparison step makes use of at least one digital processing tool chosen among nearest neighbor analysis, classification tree analysis, support vector machine analysis, and neural network analysis.

6. Method according to one of the preceding claims, wherein the part (20) comprises a woven composite material.

7. Method according to one of the preceding claims, wherein the part (20) comprises a metal material.

8. Method according to one of the preceding claims, wherein the part (20) is a turbomachine casing part, or a blade of a compressor rotor or stator or of a fan of a turbomachine.