Method for determining damage to component subjected to mechanical stress and device for implementing such method

Abstract

The invention relates to a method (1) for determining damage to a part subjected to a mechanical stress on the basis of input data representing a deformation field according to formula (I) obtained by stereoscopically correlating images of a region (S1) to be observed at different times k before and after the mechanical stress. The input data is then transformed to obtain two-dimensional output data. These data are then placed in the form of a series of images Ik representing a deformation field, each of the images comprising a useful area ZUk. Then, the area Ak of each of the useful regions ZUk is estimated to obtain a function f (k) = Ak representing the evolution of the area over time, and the area Ak is defined when f (k) is greater than or equal to the threshold # 1 characteristic of impairment. A first time, ke1, associated with the first impairment may also be simultaneously identified.

Description

Technical Field

[0001] This invention relates to the technical field of methods for determining damage to components subjected to mechanical stress, and in particular to methods for determining damage to components subjected to mechanical stress based on the analysis of digital images obtained by stereo correlation (S-CIN, stereo digital image correlation (Stereo-DIC)). Background Technology

[0002] Many techniques are known for studying damage in materials. The term "damage" refers to any form of deterioration in the properties of a material, which may be associated with changes in its internal structure (e.g., the appearance of microcracks at different scales within the material, fiber / resin debonding, fiber breakage, or delamination). These scales can be the scale of carbon fibers, carbon filaments (thousands of fibers), braided planes, several braided planes, or the entire material. In the case of metallic materials, this can be microcracks between grain boundaries or cracks within a single crystal.

[0003] Digital image correlation (DIC) is a non-destructive testing technique that uses digital image correlation to measure displacement vectors on the surface of a component. This method is based on acquiring an image of the component in its reference state (called the reference image) and an image of the component in its deformed state (called the "deformed state image"), both captured by a single digital sensor (e.g., a camera). Therefore, this method is essentially configured for two-dimensional surface measurement. By applying the principle of grayscale conservation, the displacement vector at each point on the analyzed surface can be determined, ensuring that a point in the reference image matches the point in the deformed state image where displacement has occurred.

[0004] The paper "Strain Distribution and Fatigue Life Estimation of Welded Steel Joints in Low Cycle Fatigue Based on DIC" published by Xiangyun R. et al. in Optics and Lasers in Engineering, Volume 124, 2020, uses digital image correlation to estimate the Wöhler limit of metallic materials by analyzing the fatigue hysteresis surface area during dynamic cycling. The paper "Tensive-Tensive Fatigue Behavior of Hybrid Glass / Carbon and Carbon / Carbon Composites" published by Filipe R. et al. in the International Journal of Fatigue, Volume 146, 2021, introduces the study of CIN fatigue in composite materials.

[0005] In addition, infrared thermal imaging is widely used in surface analysis. This non-destructive surface inspection technique uses an infrared camera placed remotely in front of a component to record the thermal scene. This technique is used to measure the temperature dissipated by the structure during self-heating testing. However, it may be necessary to use a light-shielding curtain to isolate the thermal scene and avoid external radiation interference that could alter the temperature measurement.

[0006] Infrared thermography can also be performed at high acquisition frequencies to identify damage in dynamic impact testing environments. This is described, for example, in the paper "A Quantitative Method for Obtaining Delamination Extension from Thermal Images Recorded During Impact Testing" by Carosena M. et al., published in NDT&E International, Vol. 100, pp. 142-152, 2018. The choice of post-impact images depends on parameters affecting the thermal effects associated with the impact, such as material type, geometry, and impact energy. This choice can be determined by empirical considerations or, as in this paper, by measuring the thermal diffusion time within the material.

[0007] However, all the aforementioned techniques share a common feature: all surface control parameters are limited to the pixel scale (thermal or visual). This severely restricts the detection of damage initiation and the monitoring of damage propagation within the studied component. Furthermore, the same metrics, which can be seen using infrared thermal imaging, may not be used to characterize and monitor damage propagation on a surface for dynamic fatigue and dynamic impact testing.

[0008] Image stereo correlation (ESR) is a nondestructive testing technique used to measure the displacement field on the surface of a component during static or dynamic mechanical testing. Knowing the displacement field, the deformation field on the surface of the component can be determined. Therefore, ESR is increasingly used in experimental mechanics. Typically, an ESR system includes at least two digital sensors (usually cameras), each acquiring images of the component from a different viewpoint. This allows for the measurement of surfaces in three-dimensional space, unlike traditional digital image correlation (DIR). Each digital sensor acquires an image of the component in its reference state. Then, after the component has been subjected to mechanical stress, each sensor acquires one or more images of the component in its successive deformation states.

[0009] Various methods have been developed to obtain reliable data related to the measured displacement. These methods involve calibrating a camera or directly calibrating a model of the camera on the part and measuring the shape of the part. In practice, what is studied is a region on the surface of the part, called the region of interest (ROI). In local methods, calibration is performed by determining the displacement between an image and a reference image in the deformed state, centered on a finite set of pixels within the ROI at that point. This finite set of pixels is called a microimage, subset, or correlation window. Calibration is performed by locating the microimage in the image in the deformed state. If several images in the deformed state exist, calibration should be performed on each of these images. However, even with this calibration, obtaining reliable displacement measurements remains difficult because calibration and shape measurement are still complex to implement. Currently, some algorithms allow obtaining the deformation field using the known dimensions of a single microimage used.

[0010] Traditionally, parameters used to analyze displacement data are based on the analysis of deformation in the direction of mechanical stress, or, for multiaxial stress, on the analysis of deformation in the direction of stress. These parameters can be used to derive the limiting values ​​of deformation at the surface of a component. However, for the reasons given above, these results still require interpretation because there is currently no objective standard to clearly identify the time it takes for a material to transition from a "healthy" state to a "damaged" state. Therefore, it is impossible to determine the conditions under which damage occurs.

[0011] Furthermore, even when damage is known to exist, it can sometimes be difficult to clearly identify the preferred direction of stress (the fiber direction of a composite material or the direction of necking in a metal). The occurrence of damage is a critical event in determining the dynamic fatigue dimensions of a structure, particularly in estimating the Waller limit. Similarly, during dynamic impact testing, knowing the time at which damage appears on the surface of a component and the associated deformation field provides the information needed to determine the dimensions of the impacted structure.

[0012] The paper "Health Monitoring of Airfoil Turbine Blades in Operation Using Three-Dimensional Digital Image Correlation" published by Wu Rong et al. in Mechanical Systems and Signal Processing, Vol. 130, pp. 470-483, May 22, 2019, discloses a method for determining damage to components subjected to mechanical stress.

[0013] This invention aims to overcome at least some of the aforementioned problems and proposes the use of a single index to study damage propagation in dynamic fatigue and dynamic impact testing environments. This index allows for the use of objective criteria to identify the transition of a material from a "healthy" state to a "damaged" state, enabling direct verification of test data related to the presence or absence of damage during testing at the end of the method according to the invention. Summary of the Invention

[0014] Therefore, the present invention proposes a method for determining damage to a component subjected to mechanical stress, wherein the component is confined in a three-dimensional space, the three-dimensional space having a pointing vector. , and The method, which involves the following steps implemented sequentially by a computer, defines orthogonal axes X, Y, and Z: 10) Provide input data, which represents the deformation field, based on... , and The deformation field is obtained by stereo correlation of images of the region of the component to be observed at different times k (k=[0...M], M is an integer) before and after mechanical stress. 20) Transform the input data to obtain two-dimensional output data, and then point to the vector... , The output data is represented in a two-dimensional reference frame to obtain a series of images I representing the deformation field. k Image I k Each image in the dataset includes a useful region ZU. k , 30) Estimating the useful area ZU k The surface area A of each useful region in the diagram k To obtain the function f(k) = A that represents the evolution of surface area over time. k , 40) Determine the threshold at which f(k) is greater than or equal to the damage. One or more surface areas A of the characteristic k And determine f(k) e1 Greater than or equal to The first time k e1 The first time k e1 Associated with the first injury.

[0015] The method according to the invention performs a transformation on input data representing a deformation field, the deformation field being based on... , and The deformation field is obtained through image stereo correlation (S-CIN), and then the method represents the deformation field in a two-dimensional representation reference frame to obtain a series of images I representing the deformation field. k Then, the useful region ZU of each of the resulting images can be estimated. k The surface area A covered by the distribution of points (pixels) in the graph. k Therefore, the evolution of the surface area over time is analyzed so that the spread of damage can be tracked as soon as it occurs.

[0016] By estimating the surface area A of the distribution of points in the useful region. k Objective criteria can then be used to identify the transition of a material from a "healthy" state to a "damaged" state. Surface area calculation is based on a two-dimensional representation of data representing the deformation field. In this way, significant changes in the mechanical behavior of a component can be observed directly by estimating the surface area, rather than having to establish standards as in the prior art. This also makes the method according to the invention reproducible and suitable for studying the emergence and propagation of damage in dynamic fatigue testing environments and dynamic impact testing environments.

[0017] Various features of the present invention that can be used together or separately: - The method also includes the following steps following step 40): 50) Regarding the time k prior to the first injury <k e1 The associated deformation fields are analyzed, and reference parameters representing the deformation fields prior to the first damage are determined. 60) Combine the reference parameters with the time k after the first damage event. e1 The associated deformation fields are compared to determine the physical parameters of the damage to the component. - The method also includes the following steps following step 50) or 60): 70) Compare the reference parameters with pre-existing parameters, where the pre-existing parameters represent the materials used to manufacture the parts; - Mechanical stress is applied during dynamic fatigue testing, and physical parameters to be determined during step 60) are used to characterize the dynamic fatigue of the material; - Step 40) includes the following sub-steps: 42) Determine the function f(k) = A xy,k The maximum value f(k) = A k,max , 44) Based on the maximum value f determined in the previous step max (k) is used to limit the damage threshold. , 46) Determine f(k) e1 Greater than or equal to the damage threshold The first time k e1 ; - Damage threshold defined in step 44) Indicates that between A k,max Between 20% and 60%; - Step 60) includes the following sub-steps: 62) The time k = k for the first injury e1 Image I of the associated deformation field ke1 To perform analysis, the forces, stresses, displacements, or deformations of the material are determined. 64) Based on the first time k e1 Associated surface area A ke1 To estimate the threshold of durability limit characteristic; - Mechanical stress is applied during dynamic impact testing, and the physical parameters to be determined during step 60) are used to characterize the dynamic impact on the material; - Step 40) includes the following sub-steps: 42′)Identification and impact time k=k i (k) i <k e The associated surface area A i , 44′) Limiting the damage threshold The damage threshold Corresponding to surface area A ke1 A ke1 >A i A ke1 The value of A i The difference is at least 25%. 46′) Identify f(k) e1 )= The first time k e1 ; - Step 60) includes the following sub-steps: 62′) The time k = k for the first injury e1 Image I of the associated deformation field ke1 To conduct analysis, 64′) Estimate the damage threshold of the material's energy limit, dynamic load limit, or dynamic deformation limit. characteristic; - Step 20) includes a first sub-step 22), which includes performing a linear regression of the deformable field; - Wherein, step 20) includes a first sub-step 22), which includes performing principal component analysis of the deformation field; - The component is the housing of the fan used in the turbine; - The useful area ZU of each image k The following region of the image is on which the data represented in the reference frame R1 extends.

[0018] The present invention also relates to an apparatus for implementing the method as described above, the apparatus comprising elements for implementing at least steps 20) to 40) by means of a computer. Attached Figure Description

[0019] Other objects, features, and advantages of the invention will become more apparent from the following description with reference to the accompanying drawings, in which: - Figure 1a This is a schematic diagram illustrating the various steps of the method according to the present invention. - Figure 1b This is a schematic diagram illustrating the various steps of a method according to a first embodiment of the present invention. - Figure 1c This is a schematic diagram illustrating the various steps of a method according to a second embodiment of the present invention. - Figure 2 The component whose damage will be studied using the methods described in this invention is shown; - Figure 3a An example of a possible spot is shown when fatigue and dynamic impact tests are performed on a solid component via S-CIN; - Figure 3b An example of a possible spot is shown when fatigue and dynamic impact testing are performed on a small component using S-CIN; - Figure 4a , Figure 4b and Figure 4c Image I, obtained during the implementation of the method according to the invention, is shown. k Image I k The data were obtained from data representing the deformation field at different times during the dynamic fatigue test. - Figure 5 It shows that in relation to Figure 4a , Figure 4b and Figure 4c In the same test case, the function f(k) = A is represented. k Curve showing evolution over time: The horizontal dashed line indicates the damage threshold. The position of the vertical dashed line indicates the start time of the fatigue test; - Figure 6a , Figure 6b and Figure 6c Image I, obtained during the implementation of the method according to the invention, is shown. k Image Ik The data were obtained from data representing the deformation field at different times during the dynamic impact test. - Figure 7 It shows that in relation to Figure 6a , Figure 6b and Figure 6c In the same test case, the function f(k) = A is represented. k Curve showing evolution over time: The horizontal dashed line indicates the damage threshold. Location; - Figure 8 The results of tests performed on an undamaged component undergoing deformation are shown for the purpose of tomographic imaging studies: Figure 8 Part a) is an image of the component. Figure 8 b、 Figure 8 c and Figure 8 d represents the components of the displacement field of the component along the axis (Z), along the axis (Y), and along the axis (X), respectively. - Figure 9 The results of tests performed on the damaged component from the perspective of tomographic imaging studies are shown: Figure 9 Part a) shows an image of the component. Figure 9 Part b Figure 9 part c and Figure 9 The components d represent the components of the displacement field of the component along the axis (Z), along the axis (Y), and along the axis (X), respectively. Detailed Implementation

[0020] The present invention relates to a method 1 for determining damage to a component subjected to mechanical stress.

[0021] 1. Definition—General Commentary Throughout this application, the term S-CIN refers to digital image stereo correlation technology. Unless otherwise stated, displacement and deformation field data are obtained via S-CIN. While S-CIN can be implemented in both static and dynamic mechanical testing environments, the method according to the invention is applied to data obtained in the dynamic domain, particularly data obtained through dynamic fatigue and dynamic impact testing. This topic will be returned to in the following sections.

[0022] The purpose of these tests can be to characterize the properties of the material or to check its quality. The methods according to the invention can also be applied to data from the tests to verify the behavior of a component, such as a component that has undergone degradation during its production. In this case, depending on the argument requirements and the area where its acceptability needs to be verified, the entire component or a portion thereof can be extracted, and then mechanical tests can be performed on the component or a portion thereof.

[0023] At this stage, it should be emphasized that the testing itself is not the subject of this invention; only the data related to the displacement and / or deformation fields obtained from these tests are useful. In other words, this invention is independent of the testing phase and everything associated with it. In other words, this invention can be implemented independently of the testing phase and the test data collection phase, and particularly independently of the data related to the displacement and / or deformation fields.

[0024] In other words, the following sections present the experimental conditions for performing the tests so that those skilled in the art can reproduce the tests.

[0025] In the context of this invention, S-CIN data is acquired by a CCD camera, which is a camera using a charge-coupled device (CCD) based imaging sensor. CCD cameras have acquisition frequencies suitable for fatigue and dynamic shock testing. CCD cameras capable of acquiring a few images per second are suitable for performing low-frequency dynamic fatigue testing but insufficient for dynamic shock testing. In the latter case, a camera capable of acquiring thousands of images per second is preferably used. That is, the method according to the invention, which will be described in more detail below, is designed to determine the damage of components subjected to mechanical stress in the environment of dynamic fatigue testing and dynamic shock testing. Therefore, a CCD camera capable of acquiring thousands of images per second is used, but it is understood that this is not mandatory.

[0026] The acquisition area is illuminated using LED (light-emitting diode) lights. Depending on the acquisition device in question, the illumination can be continuous or synchronized with the camera. When continuous illumination is used, this allows the frequency of the LED lights to be decoupled from the frequency of the CCD camera.

[0027] The component to be tested can be any component. The component is not limited by its shape, size, color, etc. The component includes an outer surface S, which includes the area S1 to be observed. An example of the embodiment is... Figure 2 As shown in the image. A close-up view of the damage can be seen within the dashed box. The component is confined in three-dimensional space, which has vectors pointing to it. , and The associated orthogonal axes are X, Y, and Z. In the following text, axis X is called the longitudinal axis, axis Y is called the transverse axis, and axis Z is called the orthogonal axis.

[0028] In the context of this invention, the component is preferably made of a woven composite material. For example, the component can be made of a woven continuous carbon fiber composite material. Such a 3D woven composite material is used, for example, in the context of a casing fan for an aircraft engine. This 3D woven composite material comprises a plurality of warp fibers arranged along a longitudinal axis, the warp fibers and weft fibers woven together in one or more weave planes and embedded in a cured polymer matrix. The weft and warp fibers can then be made of carbon, glass, or a mixture of carbon and glass. The weft and warp fibers can also be made of other types of fibers. This invention can be applied to components made of other types of composite materials.

[0029] Typically, S-CIN measurements are performed using spots deposited on the surface of a component whose damage propagation is to be studied. Measuring displacement and / or deformation fields requires images acquired by digital sensors to have random, contrasting textures and sizes suitable for the component in question. Few regions naturally possess the conditions required to perform these measurements. Spots are used to impart a suitable texture to the surface of the component. Different types of spots exist. In the context of this invention, black and white spots are used to provide a suitable texture to the component in question.

[0030] Figure 3a and Figure 3b This is an image illustrating an example of possible spots when performing fatigue and dynamic impact tests using S-CIN. Figure 3a In the middle, spots have been deposited on the solid parts, while Figure 3b In the middle, spots have been deposited on small parts. "Solid parts" are defined as parts whose maximum size is between 1 meter and several meters. For example, Figure 3a A portion of a composite impact specimen is shown. "Small component" refers to a component whose maximum dimension is less than 1 meter. As an example, Figure 3b A standard laboratory test tube is shown.

[0031] Before actual testing, a calibration procedure can be performed. This calibration procedure may include calibrating the CCD camera parameters and measuring the actual shape of the part in question. Different methods exist. For example, the CCD camera parameters can be calibrated using a grayscale minimization method, and the shape of the part can be measured using the principle of grayscale conservation. The calibration procedure can be implemented based on these well-known methods.

[0032] During dynamic impact testing, the component is subjected to firing. This firing generates mechanical stresses that can lead to damage (the term "damage" is defined in the preamble of this specification). In the event of dynamic impact and damage, when the damage occurs can be visually detected. In addition to allowing monitoring of damage propagation as soon as it occurs, this method can also be used to define a threshold above which the material used to manufacture the component in question will be damaged. This threshold is called… Damage threshold. It should be noted that the method according to the present invention can also be applied to situations involving dynamic impacts that do not cause damage.

[0033] In the illustrative examples of the invention, dynamic impact tests have been performed on specimens or test plates of real parts to ensure good mechanical strength during certification verification and to simulate flight conditions such as blade loss, inhalation of birds or debris during flight.

[0034] In dynamic fatigue testing, a component is subjected to periodic loads to estimate the durability limit of the materials used to manufacture the component. This durability limit is referred to below as the durability threshold. The durability limit can be known from the first cycle of the test, thus eliminating the need for multiple cycles to subject the component to load. In any case, the method according to the invention allows the durability limit to be estimated in a single test, unlike the current practice in the literature where several loads are used in a series of tests to estimate the durability limit. Similar to dynamic impact testing, dynamic fatigue testing may or may not cause material damage. In this way, the method according to the invention not only allows for the definition of the material's durability threshold... Furthermore, it allows for monitoring the spread of damage as soon as it occurs.

[0035] In an illustrative example of the invention, dynamic fatigue tests are performed to characterize the mechanical strength associated with the engine's operating environment. When turbine components are attached to an aircraft, these tests occur, for example, during takeoff, landing, cruise, or rapid descent phases.

[0036] 2. Description of General Embodiments refer to Figure 1a Furthermore, as indicated above, the present invention relates to a method 1 for determining damage to a component subjected to mechanical stress. As already mentioned, the mechanical stress can be applied in a fatigue test or dynamic impact test environment.

[0037] According to the first step 10 of the method 1 of the present invention, input data is provided, the input data representing the deformation field, the deformation field being based on , and The deformation field is obtained by stereo correlation of images of the region S1 (of the component) to be observed at different times k (k=[0...N], N is an integer) before and after mechanical stress.

[0038] Based on S-CIN , and The displacement and / or deformation fields reside in a three-dimensional domain, and are therefore inherently more accurate than point or average information obtained using methods such as those employing existing techniques. The deformation field can be obtained in two ways. It can be directly measured by a computer program designed for this purpose. In this respect, it is advantageous to input the dimensions of a micro-image or subset used for S-CIN into the computer program, allowing the program to directly calculate the deformation field. However, this is not mandatory. The deformation field can also be calculated from the displacement field determined by the computer program. Regardless of the method used to obtain the deformation field, it is based on... , and The deformation fields, each constituting input data, are used to implement the following steps according to method 1 of the present invention.

[0039] In practice, the S-CIN computer program can directly provide the input data related to the deformation field. If this is not the case, the deformation field can be recalculated based on the displacement field, which must have stereo-related data. Please note that the following formula should be used to calculate the deformation: .

[0040] This formula is a general example of calculating deformation in the most common cases. If deformation must be calculated, one approach is to use a finite difference scheme. It is advantageous if the scheme used is consistent with the data. If you only want the deformed data, you can restrict yourself to a template of two or three (the number of data points used for derivation), with the graph centered in the nominal case; or, if you are at the edge of the image, off-center upstream or downstream. These schemes allow the entire image to be preserved and avoid altering the size of the area being observed.

[0041] The purpose of this invention is not only to identify damage on the outer surface S of a component, but also to monitor the propagation of damage within the component in question. In this regard, it is important to ensure that images are acquired at different times before and after mechanical stress during the testing phase. As will be seen later, the input data representing the deformation field relative to the time before mechanical stress characterizes the region S1 to be observed when region S1 is healthy. Region S1 is considered "healthy" because the mechanical properties of the material have not yet undergone changes due to damage. The input data representing the deformation field relative to the time after mechanical stress allows this input data to characterize the deformation and any damage the material has undergone. Therefore, this input data is necessary when analyzing a component in a deformed state.

[0042] However, it should be noted that the input data provided during the first step 10) of the method according to the invention does not necessarily represent all the input data that can be obtained (first configuration) or actually all the input data obtained from the measurement performed by S-CIN (second configuration).

[0043] Regarding the first configuration, as an example, during dynamic fatigue testing, S-CIN analysis can be limited to a series of images acquired at every 10% of the total cycles. In other words, for a test with 100,000 cycles, analysis can be performed at 10,000 cycles, 20,000 cycles, 30,000 cycles, and so on. These values ​​can be modified according to the operator's needs and computational capabilities. In the case of dynamic impact testing, we can actually limit ourselves to a single data acquisition before the impact, and can further limit ourselves to up to ten data acquisitions after the impact begins, at which point damage to the outer surface can be observed after the test.

[0044] Regarding the second configuration, as an example, if no damage is observed on the outer surface during dynamic impact testing, it may not be necessary to analyze all the data obtained. We can limit ourselves to the data with the highest degree of deformation.

[0045] According to the second step 20 of method 1 of the present invention, the input data is transformed to obtain two-dimensional output data, and the direction vector is... , These output data are represented in a two-dimensional reference frame R1 to obtain a series of images I representing the deformation field. k Image I k Each image in the dataset includes a useful region ZU. k .

[0046] Therefore, the transformation of the data related to the deformation field can be performed in the first sub-step 22) of the second step 20). The three-dimensional data provided in step 10) is reduced to two-dimensional data. Therefore, the first sub-step 22) is merely a reduction in the size of the input data. In this respect, different variable reduction methods can be implemented. Depending on the deformation field... , and Each data point corresponds to a variable, and there is redundancy among all these variables because they are linked together (correlated). Therefore, the purpose of this first sub-step 22) is to extract data whose sum encompasses all the information contained in the input data provided in the first step 10).

[0047] According to a first specific implementation, the first sub-step 22 can be achieved by applying linear regression to the deformable field. This reduction method is well-known, and its application to the deformable field does not cause any specific problems. According to a second implementation, the first sub-step 22 can be achieved using principal component analysis (PCA). PCA is a relatively well-known decomposition method that ensures that each new variable, called a principal component, constitutes a vector orthogonal to the system composed of the other principal components. In the context of this invention, the number of principal components is determined by the model's ability to represent the entire deformable field.

[0048] The two variable reduction methods presented above, namely linear regression and principal component analysis, provide output data that objectively represents the input data while reducing the number of variables. Therefore, the ultimate goal of the first sub-step 22) is to obtain the best possible representation of the data related to the deformation field in two-dimensional form.

[0049] In the second sub-step 24) of the second step 20), the output data is represented in a two-dimensional reference frame. This produces multiple images I. k This characterizes the surface deformation at different times before and after mechanical stress. Image I8 refers to the image associated with the input data acquired at time eight, meaning that time is not necessarily equal to 8 seconds. Depending on the camera used and the input data chosen by the operator for analysis, this time can be 1.45 milliseconds (ms) or 1 second after the start of the test. Therefore, the second sub-step 24) is a simple representation step. Thus, the challenge lies in the first sub-step 22), because the input data must be qualitatively reduced to obtain a basis... , and The best representation of the deformation field.

[0050] Image I k Each image in the dataset includes a useful region ZU. kThe useful region ZU0 specifies the useful region of image I0, the useful region ZU1 specifies the useful region of image I1, and so on. This useful region is not necessarily the entire image. This useful region corresponds to an area on which the points (pixels of the image) required to achieve the third step 30) of method 1 according to the invention extend, these points being merely a graphical representation of the output data obtained during the second step 20). In other words, the useful region ZU... k Image I k The useful region ZU is the region defined by endpoints corresponding to the output data represented in the image. k This includes the extreme points (pixels) themselves, but also the points (pixels) located between these extreme points, specifically all points located between these extreme points. Therefore, this useful region can be the entire image I. k Or simply image I k This is part of the equation, depending on the situation. Useful areas can be identified by associating pixels with values, and the absence of a pixel can be represented by a value of 0.

[0051] In step 30), the useful region ZU is estimated. k The surface area A of each useful region in the diagram k To obtain the function f(k) = A that represents the evolution of surface area over time. k .

[0052] Based on the two-dimensional output data, the useful area ZU is measured. k The surface area A of each useful region in the diagram k This is used to obtain indicators, thus providing a snapshot of the evolution of the outer surface deformation over time. Therefore, when the surface area A k When a threshold is reached (e.g., the characteristics of damage), it can be identified by measuring the surface area of ​​the image I. k This determines the time associated with reaching that threshold. However, the surface area may not be used in this way. To estimate whether the surface area is a characteristic of surface damage, it is necessary to determine the threshold at which damage is considered to have occurred. This topic will be returned to below. In any case, the surface area is an objective indicator that can be used to monitor the spread of damage on the outer surface S of the component under discussion.

[0053] In the fourth step (40), the threshold for f(k) to be greater than or equal to the damage is determined. One or more surface areas A of the characteristic k And determine f(k) e1 Greater than or equal to The first time k e1 The first time k e1 Associated with the first injury.

[0054] Damaged Threshold characteristics (or more simply, The damage threshold is a surface area; anything above this threshold indicates damage to the outer surface. In other words, the damage threshold... This is the surface area value. When the surface area exceeds this value, the outer surface is damaged; that is, the outer surface has damage. The definition of damage is given in the preface of the specification and will not be repeated here. Therefore, like the surface area itself, there is a damage threshold. It is expressed in units of surface area.

[0055] A significant advantage is that the damage threshold can be predetermined. If the properties of the material used to manufacture the component are known and the method according to invention 1 has been performed at least once, the damage threshold can be determined in advance. Above the damage threshold If this occurs, damage to the outer surface S has already taken place. In this case, the first implementation of Method 1 can be considered a calibration. However, it is preferable to determine the damage threshold based on relevant test data. On the one hand, this restricts the implementation of Method 1 to once per test; on the other hand, it avoids the slippage effect that can be observed after a certain number of tests. In this case, the fourth step 40 of Method 1 should begin by determining this threshold. As will be seen later, this can be accomplished in two very simple sub-steps 42) and 44).

[0056] Known damage threshold This is sufficient to, in the first or third sub-step (depending on the situation), reduce the surface area A. k The value is compared with a threshold and the damage value greater than or equal to that threshold is extracted. The value of A. Once the relevant surface area value A has been extracted. k In the second or fourth sub-step, identify those measured to be greater than or equal to the damage threshold. First surface area A ke1 Image I k This allows for the determination of when the damage occurred. ke1 .

[0057] If the operator's goal is simply to know when the damage occurred or to know that the damage has occurred, then there is no need to perform the fifth step (50) and the sixth step (60), which will be described below. According to the method 1 of the invention, the process can be stopped at the end of the fourth step (40). Since steps (50) and (60) are optional, they are indicated by dashed boxes in the figures.

[0058] If the objective is to obtain the physical parameter characteristics of the damage, then advantageously, method 1 according to the invention may include the following steps.

[0059] In step 50), the time k before the first injury is... <k e1 The associated deformation fields are analyzed, and reference parameters representing the deformation fields prior to the first damage are determined.

[0060] This fifth sub-step is used to estimate the displacement and deformation fields of the material before mechanical stress. Therefore, this data can be used to determine the state of the outer surface S of the component while the component is still healthy. Thus, they provide a reference parameter for the material.

[0061] The reference parameters determined in this step depend on the test being performed. In the case of dynamic fatigue, for example, forces, stresses, displacements, or deformations can be determined to estimate the material's durability limit, thus knowing when damage occurs. In this case, examining the conditions applied at the time in question may be useful. In the case of dynamic impact testing, the analysis is simplified. In the case of external surface damage, this means that the local mechanical capacity of the material has been greatly exceeded, leading to cracking. Then, as with dynamic fatigue testing, the limiting dynamic energy or load or limiting dynamic deformation before damage can be derived. This topic will be returned to below.

[0062] Then, in step 60), the reference parameters and the time k after the first damage can be used. e1 The associated deformation fields are compared to determine the physical parameters of the damage to the component.

[0063] Simultaneously with or after step 60), step 70) may also be performed to compare the reference parameters with pre-existing parameters to determine the physical parameter characteristics of the damage to the component, wherein the pre-existing parameters represent the material used to manufacture the component.

[0064] These pre-existing parameters of the material may have been determined empirically or theoretically. The purpose of this comparison is solely to examine the qualitative values ​​of the parameters determined in step 50), and thus to check their reliability. Therefore, this comparison with external data allows for the establishment of standards for the consistency of the data obtained through method 1.

[0065] The method according to the invention can be implemented to determine damage to various composite material components. The composite material component can be a turbine housing, such as the housing of a turbine fan. Advantageously, the composite material component can be made of carbon fiber. Alternatively, the composite material component can be made of glass fiber. The composite material component can also be a mixture of carbon fiber or glass fiber.

[0066] It should be noted that this method is also applicable to a variety of possible deformations. Using S-CIN, the principal deformation at each point in the studied region is typically obtained. Method 1 according to the invention can be applied to principal deformations 1 and 2, rather than principal deformations having deformation fields along X and Y. These principal deformations can be obtained based on automatic analysis of the moiré circles of the deformations. Depending on the examples studied, the displayed distribution may differ visually, but the conclusions will be similar to those described in detail above.

[0067] 3. The first embodiment of the present invention—dynamic fatigue All comments made above regarding general embodiments apply, except for examples relating to dynamic impact situations.

[0068] Now, for reference Figure 1b In this embodiment, mechanical stress is applied during dynamic fatigue testing. The physical parameters determined during step 60) can then be used to characterize the dynamic fatigue of the material. In the case of dynamic fatigue, damage propagation is relatively slow and gradual. The function f(k) will change slowly and gradually over time.

[0069] According to this first embodiment of method 1, step 40) includes the following sub-steps: 42) Determine the function f(k) = A xy,k The maximum value f(k) = A k,max , 44) Based on the maximum value f determined in the previous step max(k) To limit the damage threshold , 46) Determine f(k) e1 Greater than or equal to the damage threshold The first time k e1 .

[0070] Damage threshold is calculated based on data related to dynamic fatigue testing. The input data is obtained from dynamic fatigue testing. The maximum value of the function f(k) is f(k) = A. k,max This represents the maximum magnitude of deformation experienced by the material. Therefore, the maximum value of the function f(k) can be used to objectively establish the damage threshold. Preferably, the damage threshold defined in step 44) Indicates that between A k,max Between 20% and 60%, i.e., the damage threshold defined in step 44). Having between A k,max Values ​​between 20% and 60%. In this way, greater than or equal to the established damage threshold. Any surface area value will be associated with damage to the material. This serves as a reminder that the material may undergo multiple damage events during testing. After identifying one or more surface areas that meet this criterion, the remaining task is to identify the image I associated with that surface area. k To determine when one or more injuries occurred.

[0071] Still according to the first embodiment of method 1, step 60) includes the following sub-steps: 62) The time k = k for the first injury e1 Image I of the associated deformation field ke1 To perform analysis, the forces, stresses, displacements, or deformations of the material are determined. 64) Based on the first time k e1 Associated surface area A ke1 To estimate the threshold of durability limit characteristic.

[0072] Substeps 62) and 64) are used to estimate the material's durability limit. Therefore, to obtain the reference parameters mentioned above—namely, the material's force, stress, displacement, or deformation—it is first necessary to determine the time of damage occurrence, i.e., the time of the first damage. An estimable threshold can be established. The durability threshold. As seen in the introduction of this specification, current methods may not reliably estimate this durability limit because the data obtained by these methods are subject to interpretation. In fact, no metric in the literature uses objective standards to clearly distinguish the transition of a material from a "healthy" state to a "damaged" state. Method 1 according to the invention allows for a reliable estimation of this durability limit.

[0073] In other words, this does not mean that the durability limit of a material is higher than its damage. In most cases, Durability limit threshold less than Damage threshold, because durability occurs before damage.

[0074] 4. Second embodiment - dynamic impact All comments made above regarding general embodiments apply, except for examples relating to dynamic fatigue.

[0075] Now, for reference Figure 1c In a second embodiment of the invention, mechanical stress is applied during the dynamic impact test. The physical parameters to be determined in step 60) are used to characterize the dynamic impact on the material. In the case of dynamic impact, damage propagation is rapid and sudden. Therefore, the function f(k) changes by an order of magnitude at the instant of damage, thus the timing of the damage can be easily identified.

[0076] In this respect, according to the second embodiment of method 1, step 40) includes the following sub-steps: 42′)Identification and impact time k=k i Associated surface area A i , 44′) Limiting the damage threshold The damage threshold Corresponding to surface area A ke1 A ke1 >A i A ke1 The value of A i The difference is at least 25%. 46′) Identify f(k) e1 )= The first time k e1 .

[0077] Therefore, in this embodiment, the damage threshold Indicating surface area A k The order of magnitude change. Therefore, the damage threshold Used as a standard to reflect the deformation damage limit of a material during dynamic impact. Preferably, and as indicated above, relative to the time k=k of the impact. i (k) i <k e The associated surface area A i To assess this order of magnitude change, the surface area corresponding to the "healthy" state of the object's outer surface S is used as a reference surface area. In this case, when this value is compared with A... i A difference of at least 25% indicates surface damage. This signifies a change in the order of magnitude of the surface area, resulting in a significant increase. This is unlikely to be accidental. Such a sudden increase is physically illogical and must correspond to damage to the outer surface.

[0078] Therefore, the surface area A can be measured by identifying it. e1 Image I k The associated time is used to precisely determine when the first damage occurs. Therefore, the surface area A e1 and damage threshold The same applies. Therefore, as mentioned above, this threshold is used as a standard to assess the deformation damage limit of a material during dynamic impact.

[0079] Still according to this second embodiment of method 1, step 60) includes the following sub-steps: 62′) The time k = k for the first injury e1 Image I of the associated deformation field ke1 To conduct analysis, 64′) Estimate the damage threshold of the material's energy limit, dynamic load limit, or dynamic deformation limit. characteristic.

[0080] Substeps 62') and 64') are used to estimate the dynamic energy or load of the material. Therefore, to obtain these reference parameters, it is first necessary to determine the time of damage occurrence, i.e., the time of the first damage. Thus, the damage threshold of the material's energy limit, dynamic load limit, or dynamic deformation limit can be reliably estimated. characteristic.

[0081] 5. An example of an implementation of method 1 according to the present invention. For dynamic fatigue testing and dynamic impact testing, braided composite material components are used.

[0082] 5.1 Example 1 - Dynamic Fatigue Under dynamic fatigue conditions, tests were performed stepwise. Since the material under study is a braided composite, S-CIN was used to observe damage that could lead to significant changes in mechanical properties. When stress conditions change very little (or not at all), the initial damage is likely absorbed by other components of the material (other fibers or braided elements). Therefore, the damage will be observed from one image I. k Appears in another image I k There are jumps in the surface area, and there is a general trend of increasing.

[0083] Figure 4a , Figure 4b and Figure 4c Image I is shown after implementing the method according to the invention. k Image I k The data were obtained from the deformation field data representing different time points during the dynamic fatigue test. The evolution of the distribution at different time points can be observed. Here, only images from nine different time points are shown. k . Figure 5 It shows that in relation to Figures 4a to 4c In the same test case, the function f(k) = A is represented. k Curve showing evolution over time: The horizontal dashed line indicates the damage threshold. The position is indicated by the vertical dashed line, while the starting time of the fatigue test is indicated by the vertical dashed line.

[0084] In the first image I1 (time k = 1.26 ms), all points are clustered together. No deformation was detected. Therefore, the useful region ZU1 has a small surface area ( Figure 5(First point). During dynamic fatigue testing, the surface area of ​​the useful region increases. Here, mechanical stress propagates to the specimen. When damage begins, the useful region begins to expand. Isolated points or groups of isolated points can be seen. This is clearly the case, given the time k=1.34ms associated with I4. The distributed surface area A4 increases significantly, thus reflecting the occurrence of damage. Figure 5 In this case, the surface area clearly exceeds the damage threshold. The value is undoubtedly associated with the first damage. Therefore, three regions of mechanical behavior can be identified: 1) the start of the test associated with the mechanical load; 2) the healthy region corresponding to the minimum value once the load is applied; and 3) the damaged region, which can be identified by the first peak in the surface area as the start of the damage.

[0085] Therefore, once the time of damage is known, all that is needed is to examine the conditions applied at that precise time during the test to deduce the force, stress, displacement, or deformation, thereby enabling an estimation of the material's durability limit.

[0086] 5.2 Example 2 - Dynamic Impact In the case of dynamic impact, the analysis is simplified. In the case of external surface damage, this means that the local mechanical capacity of the material has been greatly exceeded, resulting in cracks. As previously seen, the deformed surface area will increase significantly.

[0087] Figure 6a , Figure 6b and Figure 6c Image I is shown after implementing the method according to the invention. k Image I k The data were obtained from the deformation field at different times during the dynamic impact test. We can see how the distribution changes at different times. As in the previous example, only images from nine different times are shown. k . Figure 7 It shows that in relation to Figures 6a to 6c In the same test case, the function f(k) = A is represented. k Curve showing evolution over time: The horizontal dashed line indicates the damage threshold. The location.

[0088] In the first image (time k = 1.32 ms), all points are clustered together. No deformation was detected. Therefore, the useful region ZU1 has a small surface area ( Figure 7(First point). This situation remains the same for images I2, I3, and I4. During the dynamic impact test, the surface area of ​​the useful region increases. As expected, this increase is significant and sharp. Isolated points or groups of isolated points are visible. This is clearly the case given the time k=1.42ms associated with I5. The distributed surface area A5 increases significantly, reflecting the occurrence of damage. Figure 7 In this case, the surface area significantly exceeds the damage threshold, which is undoubtedly associated with the first type of damage.

[0089] Similar to fatigue testing, external conditions at that time can be identified to deduce the energy or dynamic load limit or dynamic deformation limit prior to damage during the test.

[0090] Finally, the present invention relates to an apparatus for implementing method 1 as described above. The apparatus includes elements for implementing at least steps 20) to 40) via a computer. For example, these resources are a processor and memory.

[0091] The device may also include elements for measuring the displacement of a component by stereo correlation of an image of at least one region S1 of the component to be observed, as defined in section "1. Definitions—General Commentary" of the specification. Advantageously, such resources are: - A component, comprising an outer surface S, the outer surface S comprising an area S1 to be observed, preferably comprising spots deposited on the surface of the component. - At least two CCD cameras, each capable of acquiring several images per second, more preferably several thousand images per second, to acquire S-CIN data. - Several light-emitting diode (LED) lights, which are used to illuminate the acquisition area. - Optionally, an element that generates a firing action on the component and thus produces mechanical stress.

Claims

1. A method (1) for determining damage to a component subjected to mechanical stress, said component being confined in a three-dimensional space having a pointing vector. , and The method, which involves the following steps implemented sequentially by a computer, defines orthogonal axes (X, Y, Z): 10) Provide input data, the input data representing a deformation field, the deformation field being based on , and The deformation field is obtained by stereo correlation of images of the region (S1) of the component to be observed at different times k (k=[0...M], M is an integer) before and after the mechanical stress. 20) Transform the input data to obtain two-dimensional output data, and in the direction vector , The output data is represented in a two-dimensional reference frame (R1) to obtain a series of images I representing the deformation field. k The image I k Each image in the dataset includes a useful region ZU. k , 30) Estimate the useful region ZU k The surface area A of each useful region in the diagram k To obtain a function f(k) = A that represents the evolution of the surface area over time. k , 40) Determine the threshold at which f(k) is greater than or equal to the damage. One or more surface areas A of the characteristic k And determine f(k) e1 Greater than or equal to The first time k e1 The first time k e1 Associated with the first injury.

2. The method according to claim 1 (1), further comprising the following steps after step 40): 50) For the time k prior to the first damage <k e1 The associated deformation field is analyzed, and reference parameters representing the deformation field prior to the first damage are determined. 60) Combine the reference parameters with the time k after the first damage. e1 The associated deformation fields are compared to determine the physical parameters of the damage to the object.

3. The method (1) according to claim 2, further comprising the following steps after step 50) or step 60): 70) Compare the reference parameter with a pre-existing parameter, wherein, The pre-existing parameters represent the materials used to manufacture the component.

4. The method according to any one of claims 2 and 3, wherein, The mechanical stress is applied during the dynamic fatigue test, and the physical parameters to be determined during step 60) are used to characterize the dynamic fatigue of the material.

5. The monitoring method (1) according to claim 4, wherein, Step 40) includes the following sub-steps: 42) Determine the function f(k)=A xy,k The maximum value f(k) = A k,max , 44) Based on the maximum value f determined in the previous step max (k) to limit the damage threshold , 46) Determine f(k) e1 () is greater than or equal to the damage threshold The first time k e1 .

6. The method (1) according to claim 5, wherein, The damage threshold defined in step 44) Indicates that between A k,max Between 20% and 60%.

7. The method (1) according to any one of claims 4 to 6, wherein, Step 60) includes the following sub-steps: 62) For the time k=k of the first injury e1 Image I of the associated deformation field ke1 An analysis is performed to determine the forces, stresses, displacements, or deformations of the material. 64) Based on the first time k e1 Associated surface area A ke1 To estimate the threshold of durability limit characteristic.

8. The method (1) according to any one of claims 2 and 3, wherein, The mechanical stress is applied during the dynamic impact test, and the physical parameters to be determined during step 60) are used to characterize the dynamic impact on the material.

9. The method (1) according to claim 8, wherein, Step 40) includes the following sub-steps: 42′)Identification and impact time k=k i (k) i <k e The associated surface area A i , 44′) Defines the damage threshold The damage threshold Corresponding to surface area A ke1 A ke1 >A i A ke1 The value of A i The difference is at least 25%. 46′) Identify f(k) e1 )= The first time k e1 .

10. The method (1) according to any one of claims 8 and 9, wherein, Step 60) includes the following sub-steps: 62′) for the time k=k of the first damage e1 Image I of the associated deformation field ke1 To conduct analysis, 64′) Estimate the damage threshold of the material's energy limit, dynamic load limit, or dynamic deformation limit. characteristic.

11. An apparatus for implementing the method (1) according to any one of the preceding claims, comprising elements for implementing at least steps 20) to 40) by means of a computer.

Images