Method for analyzing the weave topology of a part made of a woven composite material and associated control method
An automated method for analyzing weaving topology in woven composite materials through three-dimensional imaging reduces analysis time to 1 hour 15 minutes by automating anomaly detection and characterization, addressing the inefficiencies of manual methods.
Patent Information
- Application Number
- FR2021009330
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-09-06
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2041-09-06
AI Technical Summary
Current methods for detecting weaving anomalies in woven composite materials are time-consuming, often taking up to 10 hours, due to manual analysis of three-dimensional images, which leads to inefficiencies in identifying discrepancies between actual and theoretical weaving topologies.
An automated method for analyzing weaving topology using a three-dimensional image, involving automatic scrolling through image sets, marking potential anomalies, and subsequent verification and characterization, reducing analysis time to approximately 1 hour 15 minutes.
The method significantly reduces analysis time by automating the detection and characterization of weaving anomalies, enhancing efficiency and precision in identifying deviations from theoretical topologies.
Smart Images

Figure 00000014_0000 
Figure 00000014_0001 
Figure 00000015_0000
Abstract
Description
Title of the invention: Method for analyzing the weaving topology of a part made from a woven composite material and associated control method TECHNICAL FIELD OF THE INVENTION
[0001] The technical field of the invention is that of parts made from a woven composite material and more particularly that of methods for controlling parts made from a woven composite material.
[0002] The present invention relates to a method for analyzing the weaving topology of a part made from a woven composite material and in particular to a method for analyzing the weaving topology of a part made from a woven composite material from a three-dimensional image of the part. The present invention also relates to a method for controlling a part made from a woven composite material. TECHNOLOGICAL BACKGROUND OF THE INVENTION
[0003] A woven composite material is an assembly comprising at least one woven textile framework called reinforcement and a binder called matrix, the reinforcement being composed of woven strands. The manufacture of a part in a woven composite material therefore requires a first step of making the reinforcement by weaving, then a second step of assembly with the matrix, for example by injection after shaping in a mold.
[0004] The reinforcement of a part is made so that the weave topology of the reinforcement conforms to a theoretical weave topology allowing the part to have the desired thermo-physical and / or thermo-mechanical properties. However, it is common for deviations to be observed between the actual weave topology of the part and the theoretical weave topology. These deviations, called anomalies or weave indications, often result in a variation in the thermo-physical and / or thermo-mechanical properties of the part compared to what was planned, and can therefore lead to the rejection of the part as defective if the deviation is too great.
[0005] Currently, the detection of weaving anomalies is carried out by visual analysis of three sets of images, each set of images comprising images each corresponding to a section of a three-dimensional image of the part along a section plane perpendicular to an axis, the axis being different for each set of images and the three axes preferably forming an orthogonal reference frame. The analysis consists of manually and successively scrolling through each set of images and thus detecting a difference in the position of the strands, a sign of a potential anomaly. Once detected, an anomaly is then identified and characterized.
[0006] The operation is therefore extremely time-consuming; the analysis of a single LEAP™ blade can take up to 10 hours.
[0007] There is therefore a need to precisely detect each weaving anomaly of a piece in a shorter time. Summary of the invention
[0008] The invention offers a solution to the problems mentioned above, by making it possible to check the conformity of the weaving topology of the reinforcement of a part with the theoretical weaving topology in less than 2 hours.
[0009] A first aspect of the invention relates to a method for analyzing the weaving topology of a part made from a woven composite material, the method being carried out from a three-dimensional image of the part acquired by an imaging system, and comprising the following steps: • For each axis of a set of axes, automatic scrolling of a set of images at constant speed, the set of images comprising a plurality of images each corresponding to a section of at least one part of the three-dimensional image along a section plane perpendicular to the axis and each being associated with a position on the axis, the plurality of images being ordered by increasing position in the set of images; • As soon as a potential weaving anomaly is detected on an image of the set of images, scrolling is stopped, a marker is created at a point in the image belonging to the potential weaving anomaly and then scrolling is resumed; • For each marker created, if the corresponding potential weaving anomaly actually corresponds to a weaving anomaly, identification and characterization of the weaving anomaly.
[0010] Thanks to the invention, the analysis of the weaving topology of a part made from a woven composite material is carried out by scrolling through at least one set of images comprising sections of a three-dimensional image of the part so as to scan the entire volume of the part, and stopping only when a weaving anomaly is suspected. The creation of a reference point then makes it possible to retain the location of the potential weaving anomaly while the images are scrolling and to return to it once the scrolling is complete.
[0011] The scrolling is automatic and carried out at constant speed, that is to say without action from the operator and in a regular manner.
[0012] Once the complete scrolling is completed, each potential weaving anomaly is analyzed to determine whether it actually corresponds to a weaving anomaly, then identified from a set of known and characterized weaving anomalies.
[0013] By limiting the downtime during scrolling and by analyzing all the potential weaving anomalies once scrolling is complete, the duration of the analysis of the weaving topology of the part is considerably reduced, approximately 1h 15 being sufficient to analyze the weaving topology of a LEAP™ blade with the method according to the first aspect of the invention.
[0014] In addition to the characteristics which have just been mentioned in the preceding paragraph, the method according to the first aspect of the invention may have one or more complementary characteristics among the following, considered individually or according to all technically possible combinations.
[0015] According to an alternative embodiment, each image of the set of images corresponds to a section of the three-dimensional image from which a part of the part held during the acquisition of the three-dimensional image by the imaging system has been subtracted.
[0016] Thus, the artifacts generated by the tooling holding the part during the acquisition of the three-dimensional image are not analyzed. Since weaving anomalies are easily visible on the surface of the part, the area is not analyzed. The analysis time is therefore reduced.
[0017] According to an alternative embodiment compatible with the previous alternative embodiment, the method according to the first aspect of the invention comprises a step of receiving a weaving report generated at the end of the weaving of the piece by a loom: • If the weaving report has no identified weaving anomalies and no loom-related events, each image in the image set corresponds to a section of the three-dimensional image from which an upper part of the piece has been subtracted.
[0018] Thus, an upper part of the part is not analyzed if no problem was noted during the weaving of the part and the duration of the analysis is therefore reduced.
[0019] According to an alternative embodiment compatible with the preceding alternative embodiments, the identification and characterization step comprises creating a box encompassing the weaving anomaly in the three-dimensional image.
[0020] Thus, obtaining the dimensions of the weaving anomaly is facilitated since the dimensions of the box directly give the dimensions of the weaving anomaly. It is therefore not necessary to carry out a dimension measurement for each direction. The duration of the analysis is therefore reduced.
[0021] According to an alternative embodiment compatible with the preceding alternative embodiments, the identification and characterization step comprises a measurement of a spacing using an interactive tool in the form of a circle.
[0022] Thus, obtaining a spacing measurement is facilitated since the measurement is obtained directly from the circle, the installation of which only requires the positioning of its center and a point on its perimeter. The duration of the analysis is therefore reduced.
[0023] A second aspect of the invention relates to a method for controlling a part made from a woven composite material from a three-dimensional image of the part acquired by an imaging system, comprising the following steps: • For each axis of a set of axes and for each element of a predefined set of elements: • Automatic detection of each occurrence of the element present in each image of a set of images comprising a plurality of images each corresponding to a section of the three-dimensional image according to a section plane perpendicular to the axis, and creation for each image, of a set of occurrences comprising each detected occurrence; • For each image of at least part of the image set, checking each occurrence of the corresponding occurrence set and removing each unchecked occurrence from the occurrence set; • If each set of occurrences is empty, carrying out the steps of the method according to the first aspect of the invention.
[0024] Thus, each element that could lead to rejection of the part is automatically searched for in each image, then each occurrence of the detected element is checked by an operator, and the analysis of the weaving topology of the part is only carried out if no element has been detected in the part after checking.
[0025] According to an alternative embodiment, the predefined set of elements comprises at least one element from among the following: void volume ratio, foreign body,.
[0026] According to an alternative embodiment compatible with the previous alternative embodiment, the verification step is only carried out for a subset of images of the set of images comprising each image of the set of images for which at least one corresponding set of occurrences is non-empty.
[0027] Thus, only images on which at least one element has been automatically detected are verified and the duration of the verification step is therefore reduced.
[0028] According to an alternative embodiment compatible with the preceding alternative embodiments, the detection step comprises a creation for each detected occurrence, of a box encompassing the occurrence in the three-dimensional image, each box created being invisible when carrying out the steps of the method according to the first aspect of the invention.
[0029] According to an alternative embodiment compatible with the preceding alternative embodiments, the method according to the second aspect of the invention further comprises a step of automatically calculating the distance between each identified weaving anomaly.
[0030] Thus, a decision regarding the rejection of the part based on the identified weaving anomalies can be made automatically.
[0031] According to an alternative embodiment compatible with the preceding alternative embodiments, the method according to the second aspect of the invention comprises a step of automatically generating a final report comprising each identified weaving anomaly.
[0032] Thus, no manual operation is necessary to generate the final report.
[0033] A third aspect of the invention relates to a computer program product comprising instructions which, when the program is executed on a computer, cause the latter to implement the steps of the method according to the first or second aspect of the invention.
[0034] A fourth aspect of the invention relates to a computer-readable recording medium comprising instructions which, when executed by a computer, cause the computer to implement the steps of the method according to the first or second aspect of the invention.
[0035] The invention and its various applications will be better understood upon reading the following description and examining the accompanying figures. BRIEF DESCRIPTION OF THE FIGURES
[0036] The figures are presented for information purposes only and in no way limit the invention. • [Fig.l] shows a three-dimensional image of a woven composite material, acquired by X-ray tomography. • [Fig.2] is a block diagram illustrating the sequence of steps of an analysis method according to the invention. • [Fig.3] shows a schematic representation of a third stage and of a fourth step of the analysis method according to the invention. • [Fig.4] is a block diagram illustrating the sequence of steps of a control method according to the invention. DETAILED DESCRIPTION
[0037] Unless otherwise specified, the same element appearing in different figures has a single reference.
[0038] A first aspect of the invention relates to a method for analyzing the weaving topology of a part made from a woven composite material.
[0039] A composite material is an assembly comprising a textile reinforcement and a matrix, the matrix being a binder, often made of plastic.
[0040] In a woven composite material, the textile reinforcement comprises strands woven using a loom, following a theoretical weaving topology defined according to at least two orientations, also called reinforcement axes, each strand being composed of a plurality of fibers, often carbon or glass.
[0041] For three-dimensional woven composite materials, the theoretical weave topology is defined over several layers linked by strands called binders.
[0042] The term "analysis of the weaving topology of a part made from a woven composite material" means the verification of the conformity of the actual weaving topology of the part with the corresponding theoretical weaving topology, a deviation between the actual weaving topology and the theoretical weaving topology being called a weaving anomaly and being able to lead to the rejection of the part.
[0043] A variation between the actual weaving topology of the part and the corresponding theoretical weaving topology can be qualified as a deviation if the value of each criterion of a set of criteria linked to the variation is greater than a predefined threshold.
[0044] The set of criteria includes, for example, a displacement expressed in number of layers and / or a spacing expressed in millimeters.
[0045] The part can be divided into several zones and the threshold associated with each criterion can vary depending on the zone of the part in which the weaving anomaly is located.
[0046] The analysis method according to the invention is carried out from a three-dimensional image of the part acquired by an imaging system.
[0047] The imaging system is for example an X-ray tomography imaging system.
[0048] When acquiring the three-dimensional image, the part is held by a tool that comes into contact with only a part of the part.
[0049] [Fig-1] [Fig.l] shows a three-dimensional I3D image of a composite material, acquired by X-ray tomography.
[0050] In [Fig.l], the composite material is a three-dimensional woven composite material having a reinforcement 300 consisting of strands 302 woven according to a weaving topology defined on several layers 301 arranged according to two orthogonal reinforcement axes X and Y, conventionally called warp and weft, the layers 301 being superimposed according to the reinforcement axis Z.
[0051] The range of gray levels of the three-dimensional image I3D is for example limited to a predefined range of gray level values.
[0052] A set of images is obtained from the three-dimensional image I3D for each axis of a set of axes comprising at least one axis, each image of the set of images being a section of at least a part of the three-dimensional image I3D along a section plane perpendicular to the axis.
[0053] The set of images comprises, for example, 1500 images.
[0054] [Fig.3] [Fig.3] shows a schematic representation of a three-dimensional image I3D of a part made from a woven composite material and of an image I corresponding to a section of the three-dimensional image I3D along a section plane comprising the X and Y axes, perpendicular to the Z axis.
[0055] The section planes of the images I of the set of images are parallel to each other and not coincident, that is to say that the section planes are spaced apart from each other along the axis to which they are perpendicular. Each plane can therefore be associated with a position on the axis to which it is perpendicular.
[0056] The images I of the image set are ordered by increasing position along the axis in the image set.
[0057] Three sets of images are for example obtained from the three-dimensional image I3D for three distinct axes if the part is made from a three-dimensional woven composite material.
[0058] The three axes form, for example, an orthogonal reference frame.
[0059] The three axes correspond for example to the reinforcement axes X, Y and Z or to the three main axes of the room.
[0060] Each image I of the set of images is for example a section of the three-dimensional image I3D from which the part of the part held during the acquisition of the three-dimensional image I3D has been subtracted.
[0061] The boundaries of each area of the part may appear on the I images of the image set.
[0062] The part is for example an aeronautical part, such as an engine blade, for example the LEAP™ blade.
[0063] [Fig.2] [Fig.2] is a block diagram illustrating the sequence of steps of the analysis method 100 according to the invention.
[0064] The analysis method 100 according to the invention may comprise a first step 101 of receiving a weaving report.
[0065] The weaving report was generated after weaving the reinforcement 300 of the part and includes each weaving anomaly identified during weaving and each loom-related event that occurred during weaving.
[0066] An event related to the loom is for example the blocking of the loom or the breakage of a thread during weaving.
[0067] If the weaving report does not include any identified weaving anomalies and no loom-related events, each image I of the set of images is, for example, a section of the three-dimensional image I3D from which an upper part of the piece has been subtracted.
[0068] The upper part of the part comprises, for example, the part of the part held during the acquisition of the three-dimensional image I3D.
[0069] A second step 102 of the analysis method 100 according to the invention consists of scrolling each set of images at constant speed.
[0070] Scrolling is done automatically, i.e. without human intervention.
[0071] The speed value can be adjustable and is for example between 17 and 24 images per second.
[0072] A third step 103 of the analysis method 100 according to the invention is carried out as soon as a potential weaving anomaly is detected on an image I of the set of images during the scrolling carried out in the second step 102.
[0073] The potential weaving anomaly is for example detected either by an operator or automatically, for example using an artificial neural network.
[0074] The term “potential weaving anomaly” means a part of the reinforcement 300 of the part likely to include a weaving anomaly.
[0075] The third step 103 consists of stopping the scrolling carried out in the second step 102 on the image I of the set of images on which the potential weaving anomaly was detected.
[0076] The third step 103 then consists of creating a marker 1031 at a point in the image I belonging to the potential weaving anomaly, then resuming scrolling.
[0077] In [Fig.3], the reference 1031 is represented by a cross.
[0078] A fourth step 104 of the analysis method 100 according to the invention is carried out for each marker 1031 created during the third step 103 if a first condition C1 is verified for the marker 1031.
[0079] The first condition C1 is verified if the potential weaving anomaly on which the marker 1031 has been placed is indeed a weaving anomaly, that is to say if a deviation from the theoretical weaving topology exists in the zone in which the marker 1031 is placed.
[0080] The verification of the first condition Cl is for example carried out by an operator.
[0081] The fourth step 104 then consists of identifying and characterizing the weaving anomaly.
[0082] The term "identifying a weaving anomaly" means assigning a name to the weaving anomaly from a known set of weaving anomaly names.
[0083] The known set of weaving anomalies includes, for example, undulations and multi-layer interlacing anomalies.
[0084] The term “characterizing a weaving anomaly” means determining characteristics of the weaving anomaly, such as for example its length, width and height in the part, or the spacing of strands 302 composing it.
[0085] The fourth step 104 comprises for example the creation of a box 1041 encompassing the weaving anomaly in the three-dimensional image.
[0086] The box 1041 is for example a rectangular parallelepiped each side of which is in contact with at least one point of the external contour of the weaving anomaly.
[0087] The fourth step 104 comprises for example at least one measurement of the spacing between two strands 302 of the weaving anomaly.
[0088] The spacing measurement is for example carried out using an interactive tool 1042 in the shape of a circle as illustrated in [Fig.3] on two strands 302 of the weaving anomaly corresponding to the mark 1031 in image I.
[0089] To carry out a measurement of the spacing between two strands 302 using the interactive tool 1042, it is sufficient to position a point corresponding to the center of the circle then a point located on the perimeter of the circle so that the circle is generated automatically, the diameter or the radius of the circle then providing a measurement of the spacing.
[0090] In [Fig.3], the point corresponding to the center of the circle is placed between the two strands 302 and the point located on the perimeter of the circle is placed on one of the strands 302, the diameter of the circle giving the spacing between the two strands 302.
[0091] Alternatively, the point corresponding to the center of the circle can be placed on one of the strands 302 and the point located on the perimeter of the circle on the other strand 302, the radius of the circle giving the spacing between the two strands 302.
[0092] A second aspect of the invention relates to a method for controlling the part, produced from the three-dimensional image I3D of the part.
[0093] [Fig.4] [Fig.4] is a block diagram illustrating the sequence of steps of the control method 200 according to the invention.
[0094] A first step 201 of the control method 200 according to the invention consists, for each element of a predefined set of elements, in automatically detecting each occurrence of the element in each image I of each set of images and in creating, for each image I, a set of occurrences comprising each occurrence of the element detected.
[0095] The detection is for example carried out by thresholding on the gray levels of the image I.
[0096] The set of elements comprises, for example, at least one element among the following: void volume ratio, foreign body.
[0097] The term “void volume ratio” means a quantity of air, in the form of an air bubble created in the part during the injection of the matrix, included in a defined rectangular parallelepiped.
[0098] “Foreign body” means any element other than the reinforcement 300 and the matrix of the part.
[0099] The first step 201 comprises, for example, the creation for each detected occurrence of a box 1041 encompassing the occurrence in the three-dimensional image I3D.
[0100] A second step 202 of the control method 200 according to the invention consists, for each image I of at least a part of each set of images, in verifying each occurrence of the set of occurrences obtained for the image I in the first step 201 and in deleting each unverified occurrence from the set of occurrences.
[0101] By "verifying each occurrence of an element" is meant the analysis of each detected occurrence to ensure that it is indeed an occurrence of the element.
[0102] The second step 202 is for example carried out only for a subset of images of each set of images comprising each image I of the set of images for which at least one corresponding set of occurrences is non-empty.
[0103] The second step 202 is for example carried out by an operator.
[0104] Once the second step 202 is completed, the steps of the analysis method 100 according to the invention are then carried out if each set of occurrences is empty.
[0105] If a box 1041 has been created for each occurrence detected in the first step 201, each box 1041 created is for example invisible when carrying out the analysis method 100 according to the invention, that is to say that no box 1041 created appears on an image I of the set of images when carrying out the analysis method 100 according to the invention.
[0106] The control method 200 according to the invention may comprise a third step 203 consisting of automatically calculating the distance between each weaving anomaly identified at the end of the analysis method 100 according to the invention, that is to say each weaving anomaly identified during the fourth step 104 of the analysis method 100 according to the invention and, where appropriate, each weaving anomaly identified during weaving and included in the weaving report.
[0107] The control method 200 according to the invention may comprise a fourth step 204 consisting of automatically generating a final report comprising each weaving anomaly identified at the end of the analysis method 100 according to the invention and if the third step 203 is carried out, each distance calculated in the third step 203.
Claims
Claims
1. Method (100) for analyzing the weaving topology of a part made from a woven composite material, the method (100) being carried out from a three-dimensional image (I3D) of the part acquired by an imaging system, and comprising the following steps: Detection of a set of potential weaving anomalies by an operator, by automatic scrolling of a set of images at constant speed, the set of images comprising a plurality of images (I) each corresponding to a section of at least a part of the three-dimensional image (I3D) along a section plane perpendicular to an axis (X, Y, Z) of the part, and each being associated with a position on the axis (X, Y, Z), the plurality of images being ordered by increasing position in the set of images (102); • As soon as a potential weaving anomaly is detected by the operator on an image of the set of images, the scrolling is stopped, a marker (1031) is created at a point in the image (I) belonging to the potential weaving anomaly and then the scrolling is resumed (103); If the set of potential weaving anomalies is non-empty, for each marker (1031) created, if the corresponding potential weaving anomaly actually corresponds to a weaving anomaly (Cl), identification and characterization of the weaving anomaly (104).
2. Method (100) according to claim 1, in which each image (I) of the set of images corresponds to a section of the three-dimensional image (I3D) from which a part of the part held during the acquisition of the three-dimensional image (I3D) by the imaging system has been subtracted.
3. Method (100) according to any one of the preceding claims, comprising a step (101) of receiving a weaving report generated at the end of the weaving of the piece by a loom: - If the weaving report does not include any identified weaving anomaly and no event linked to the loom, each image (I) in the image set corresponds to a section of the three-dimensional image (I3D) from which an upper part of the part has been subtracted.
4. Method (100) according to any one of the preceding claims, wherein the step (104) of identifying and characterizing comprises creating a box (1041) encompassing the weaving anomaly in the three-dimensional image (I3D).
5. A method (100) according to any preceding claim, wherein the identifying and characterizing step (104) comprises measuring a gap using an interactive circle-shaped tool (1042).
6. Method (200) for inspecting a part made of a woven composite material from a three-dimensional image (I3D) of the part acquired by an imaging system, comprising the following steps: - For each axis (X, Y, Z) of a set of axes and for each element of a predefined set of elements: • Automatic detection of each occurrence of the element present in each image (I) of a set of images comprising a plurality of images (I) each corresponding to a section of the three-dimensional image (I3D) along a section plane perpendicular to the axis (X, Y, Z), and creation for each image (I), of a set of occurrences comprising each detected occurrence (201); • For each image (I) of at least a part of the set of images, verification of each occurrence of the corresponding set of occurrences and deletion of each unverified occurrence from the set of occurrences (202);- If each set of occurrences is empty, carrying out the steps of the method (100) according to any one of the preceding claims.;
7. The method (200) of claim 6, wherein the predefined set of elements comprises at least one of the following: void volume ratio, foreign body,.
8. The method (200) of claim 6 or 7, wherein the verification step (202) is performed only for a subset of images of the set of images comprising each image (I) of the set of images for which at least one corresponding set of occurrences is non-empty.
9. Method (200) according to any one of the preceding claims, in which the detection step (201) comprises a creation for each detected occurrence, of a box (1051) encompassing the occurrence in the three-dimensional image (I3D), each box (1051) created being invisible when carrying out the steps of the method (100) according to any one of claims 1 to 5.
10. Method (200) according to any one of the preceding claims, further comprising a step (203) of automatically calculating the distance between each identified weaving anomaly.
11. Method (200) according to any one of the preceding claims, comprising a step (204) of automatically generating a final report comprising each identified weaving anomaly.
