Ultrasonic detection method for pore defects in component

Abstract

The invention relates to a method for ultrasonic detection of pore area type defects in a to-be-tested workpiece. The method comprises the following steps: receiving an ultrasonic image of the to-be-tested workpiece; receiving a thickness map of at least one part of the to-be-tested workpiece; based on the thickness map of the at least one part of the to-be-tested workpiece, a simulated ultrasonic map of a corresponding area of the non-pore area type defect workpiece is determined, and the component and the thickness map of the corresponding area of the non-pore area type defect workpiece are the same as those of the corresponding area of the to-be-tested workpiece; based on the ultrasonic image of the to-be-tested workpiece and the simulated ultrasonic image, determining a detection image of the to-be-tested workpiece; and determining whether the pore area type defect exists or not in the to-be-tested workpiece based on the detection graph.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ultrasonic non-destructive testing of components, and in particular to a technique for detecting defects by analyzing the attenuation of ultrasonic waves in a workpiece.

[0002] Therefore, the present application relates to an ultrasonic testing method for detecting defects using thickness data of at least one component of a workpiece. BACKGROUND

[0003] In industrial production, some components are manufactured according to a digital model. However, during the production of a workpiece, the digital model of the workpiece (i.e. the “ideal” workpiece representing the desired manufacture) can differ from the actual manufactured workpiece in terms of structure or composition. For example, for a workpiece made of woven composite material, the actual woven structure can differ from the theoretical woven structure provided to the weaving machine. Such differences are not necessarily problematic and do not all need to be detected.

[0004] The manufactured workpiece can also have defects such as cracks, fissures, delaminations or peels (for adhesive materials), which can cause “pore areas” to form inside the workpiece. Such defects can change the mechanical properties of the workpiece, increasing its brittleness, and must therefore be detected.

[0005] In the prior art, ultrasonic testing methods are often used to detect such “pore area” type defects. In an ultrasonic testing method, a transducer is placed on the surface of the workpiece to be tested and emits ultrasonic waves, which propagate inside the workpiece. When the ultrasonic waves encounter the interface between two areas of different acoustic impedance inside the workpiece (in particular when the ultrasonic waves encounter a defect inside the workpiece), part of the ultrasonic waves is reflected and the part that is not reflected is attenuated.

[0006] Conventional ultrasonic testing methods are mainly divided into two categories: reflection method and transmission method (also known as attenuation method). In the reflection method, the same transducer is responsible for both emitting ultrasonic waves and receiving reflected waves (i.e. “echoes” or “return signals” generated by the emitted waves). According to the intensity and arrival time of the return signals, information about the presence and location of defects in the workpiece can be obtained. In the transmission method, the receiver and the transmitter are independent of each other and are placed on another surface of the workpiece (for example, the surface opposite to the surface on which the ultrasonic wave emitting transducer is placed) to receive the attenuated ultrasonic waves. Defect detection is based on the amount of ultrasonic waves that penetrate the workpiece and reach the second surface (unlike the reflection method, in which the ultrasonic waves propagate twice, the transmission method involves only a single pass of the ultrasonic waves).

[0007] Since pore area type defects cause strong attenuation of ultrasonic waves, the transmission method is very suitable for detecting such defects.

[0008] For example, such methods have been applied to the inspection of bonded workpieces, such as aerospace workpieces comprising a composite component with metallic reinforcements bonded to the component. A typical example is the composite fan blade of the LEAP engine, whose titanium alloy reinforcements are fixed by bonding. Defects such as porosities or lack of glue can exist in the bonded area, which are all porosity area type defects. The manufactured workpieces are usually subjected to ultrasonic inspection to detect such defects.

[0009] The three-dimensional woven structure of the composite material causes strong scattering of the ultrasonic waves, and the resin in the composite material causes significant attenuation of the ultrasonic waves. Therefore, the implementation of the reflection method is extremely difficult, and the attenuation method is more preferred in actual detection.

[0010] The ultrasonic detection result can be presented by a C-view, i.e. a so-called "C-scan image", which is a corresponding relationship diagram between the surface detection position of the region to be detected and the amplitude of the ultrasonic transmission signal, and the C-scan image can provide a planar top view of the workpiece.

[0011] In the C-scan image of the manufactured workpiece, the defects are often difficult to directly identify, and therefore the C-scan image usually needs to be processed to highlight the defects. Specifically, a standard piece can be used as a reference, which is of the same type as the workpiece to be detected (for example, a composite blade with a bonded titanium alloy leading edge of a predefined size), and is known to be free of porosity area type defects. The C-scan image of the standard piece is obtained as a reference image.

[0012] For each workpiece to be detected, its C-scan image is determined, and the image is subtracted from the reference image. After subtraction, a "processed" image is obtained, in which the defects and other patterns in the image are shown. Then, thresholding processing is performed on the image to eliminate patterns other than defects, so that only visible defects are retained in the thresholded image. The threshold value used for image thresholding processing is a preset threshold value, which can be adjusted according to the method described below.

[0013] The above defect detection process is shown in Figures la-le. Specifically, Figure 1 a-1e show schematic views of a workpiece to be detected 101. In this example, Figure 1 a shows a workpiece 101 which is a composite fan blade, and a titanium alloy leading edge 101b is bonded to a composite body 101a. Figure 1 b presents a C-scan image 102 of the workpiece 101. The image is conventionally obtained by the attenuation ultrasonic detection method. The C-scan image 102 comprises a plurality of regions 102a, 102b, 102c, 102d corresponding to different ultrasonic attenuation values.

[0014] Figure 1c shows the C-scan image 103 of the standard part. This standard part is of the same type as the workpiece 101 under test (e.g., manufactured based on the same digital model, theoretically identical in size, structure, and composition), but it has been confirmed that it does not contain the type of defects that need to be checked in the workpiece 101 under test; that is, the standard part is a defect-free "ideal" workpiece. In the examples of Figures 1a-1e, the standard part is a blade assembled from a intact prefabricated composite material body and a bonded titanium alloy leading edge with a geometry as close as possible to the nominal geometry. The bonding process is strictly controlled and inspected to ensure that it is defect-free.

[0015] Similar to the C-scan image 102 of the workpiece under test, the C-scan image 103 of the standard part also contains multiple regions 103a, 103b, 103c, and 103d corresponding to different ultrasonic attenuation values.

[0016] Figure 1d shows the processed image 104, obtained by calculating the difference between the scanned image 102 of the workpiece C to be tested and the scanned image 103 of the standard part C. For example... Figure 1 As shown in Figure d, the range of ultrasonic attenuation values ​​in the processed image 104 is significantly reduced compared to C-scan images 102 and 103. However, the processed image 104 contains regions 104a, 104b, and 104c, referred to as "ultrasonic indicator areas." These regions correspond to attenuation peak areas and may be porosity-type defects. To identify the defects, the processed image 104 needs to be thresholded, resulting in the thresholded image 105 shown in Figure 1e. Specifically, a threshold is set for the attenuation values ​​of the processed image 104. In particular, this threshold can be defined as a percentage of the maximum amplitude of the received signal (i.e., amplitudes below this maximum amplitude are "truncated"). For example, a preset threshold can correspond to a signal attenuation value of -6dB. That is, when the attenuation value corresponding to the difference between the workpiece under test and the standard part in C-scan image 103 is greater than -6dB, the difference is determined to be a porosity-type defect (especially an adhesive defect). The steps for determining this threshold will be explained in detail below. Figure 1 In the threshold-processed image 105 of e, defect 105a can be clearly identified.

[0017] It can be seen that by calculating the difference between the scanned image 102 of the workpiece C and the scanned image 103 of the standard part C, the interference of complex geometric shapes and thickness differences of the workpiece can be eliminated, thereby improving the detection accuracy of defects in the porosity region.

[0018] The threshold applied to the processed image 104 can be determined or adjusted in an additional step based on a so-called "defect" part. This part is of the same type as the workpiece to be tested, but it contains "controlled" defects, i.e. the position, size and type of each defect in the defect part are known. This defect part is usually machined from a reference part (i.e. a defect-free part) in which elements having the same behavior as the porosity-like defects are inserted. For example, Teflon elements can be inserted, given that the overall behavior of the Teflon elements with respect to the ultrasonic signal is similar to that of a bubble (the signal attenuation is roughly the same when the ultrasonic wave passes through a Teflon element as when it passes through a bubble or air gap). With the help of the defect part, it is possible to determine the detection threshold in decibels (dB), for example, according to the tolerance range of the actual size of the defects. That is, since the position of the defects in the defect part is known, it is possible to find the best compromise between "false positives" (detection of a defect on the threshold image but which is not actually a defect to be tested) and "false negatives" (presence of a defect in the workpiece but not shown on the threshold-processed image) by adjusting the detection threshold.

[0019] The flowchart of the prior art defect workpiece detection method described above is shown in Figure 2.

[0020] Step 210: acquisition of the C-scan image of a defect-free reference part (also called "gain calibration blade") (i.e. the C-scan image 103 in Figures la-le); Step 220: acquisition of the C-scan image of the workpiece to be tested (i.e. the C-scan image 102 in Figures la-le); Step 230: difference operation between the C-scan image of the workpiece to be tested and the C-scan image of the defect-free reference part, resulting in a so-called processed image (i.e. the image 104 in Figures la-le).

[0021] Step 240: acquisition of the C-scan image of the defect part (the defect information being known as described above); Step 250: determination of the optimal threshold for defect detection based on this scan, as described above.

[0022] Step 260: application of the threshold determined in Step 250 to the processed image obtained in Step 230. Step 270: identification of the defects in the threshold-processed image. In fact, the difference between the workpiece to be tested and the defect-free reference part (i.e. the ultrasonic indication zone) is theoretically only the adhesive defect, i.e. the porosity-like defect.

[0023] The C-scan images of the reference part, the workpiece to be tested and the defect part are usually obtained by a transmission ultrasonic detection method.

[0024] It should be noted that steps 210, 220 and 240 can be executed in parallel, or in any order (however, in general, steps 210 and 240 related to the standard and defective parts are completed before the actual detection step of the workpiece, i.e. step 220).

[0025] In the above prior art method, the role of the non-defective standard part is to provide a C-scan image as uniform as possible, so as to eliminate the differences in ultrasonic attenuation caused by the material itself of the workpiece and the thickness of the workpiece (especially the thickness of different materials such as composite materials and titanium alloys in the previous example).

[0026] However, not only the defects to be detected can cause ultrasonic indication zones (i.e. regions of ultrasonic attenuation peaks). Other factors can also cause ultrasonic indication zones, especially the differences in the structure or thickness of the workpiece (for example, for a composite material blade with a bonded titanium alloy leading edge, differences in the geometry of the leading edge and / or differences in the woven reinforcement layer can cause such indication zones). These differences are difficult to predict when the detection scheme is developed, and the reasons for their occurrence can include differences from the same product supplier, changes in geometry over time, differences in manufacturing processes, and relaxation of the geometric acceptance criteria for the workpiece (or intermediate parts, such as the leading edge and the preform in the previous example). These differences can cause false detection marks (false defects) in the threshold-processed image.

[0027] One solution is to set different thresholds for different parts of the workpiece to be tested, but this solution is difficult to implement and requires the pre-determination of regions where differences in attenuation are allowed, and the setting of corresponding acceptable thresholds for each region. In addition, this solution can increase the risk of missed detection, i.e. more defects are not detected (especially those located in the regions of the workpiece where larger differences in attenuation are allowed).

[0028] Therefore, there is an urgent need for an improved method for detecting defects in the porosity region of a workpiece, which can eliminate irrelevant ultrasonic indication zones (i.e. regions that appear as defects in the threshold-processed image but are not actually defects). SUMMARY

[0029] The present application solves the above-mentioned problems in the prior art by using the thickness data of the workpiece to be tested to construct a simulated ultrasonic image of a virtual workpiece, and using this simulated ultrasonic image as a reference image instead of the ultrasonic image of the non-defective standard part in the prior art. The use of this simulated ultrasonic image can effectively reduce irrelevant ultrasonic indication zones caused by geometric differences between the manufactured workpiece and its digital model of manufacture.

[0030] One aspect of the present application relates to a computer-implemented method for ultrasonic testing of defects in the porosity region of a workpiece to be tested, the method comprising the following steps:

[0031] Obtain the ultrasonic image of the workpiece to be tested;

[0032] Obtain thickness maps of at least a portion of the workpiece to be measured;

[0033] Based on the thickness map of at least a portion of the workpiece to be tested, a simulated ultrasonic map of the corresponding region of the workpiece with a non-porous defect is determined. The composition and thickness map of the corresponding region of the non-porous workpiece are completely consistent with those of the workpiece to be tested.

[0034] Based on the ultrasonic wave image and simulated ultrasonic wave image of the workpiece to be tested, the detection pattern of the workpiece to be tested is determined;

[0035] By applying a preset threshold to the inspection image of the workpiece under test, it is determined whether there are porosity-type defects in the workpiece under test.

[0036] The "porosity defects" mentioned in this article refer to defects in a workpiece whose effect on ultrasonic waves (especially the ultrasonic wave attenuation coefficient) is similar to that of pores in the workpiece. When such defects are present, the attenuation of ultrasonic waves is usually greater than the attenuation when passing through other areas of the workpiece. For example, bubbles or cracks belong to this category of defects.

[0037] The "workpiece to be tested" mentioned in this article refers to a mechanical workpiece that needs to be inspected for defects. It should be noted that this method can be applied to both complete workpieces and partial areas of workpieces (not complete workpieces). For the sake of simplicity, the complete workpiece and partial areas of workpieces will be collectively referred to as "workpiece" or "workpiece to be tested" below.

[0038] The "ultrasonic map" described in this article is a graphical representation of the ultrasonic characteristics at different locations on a workpiece. For example, an ultrasonic map can be a two-dimensional image projected onto a predetermined plane, reflecting the amplitude of the ultrasonic signal received at various points on the projection plane after the ultrasonic wave passes through the workpiece. Specifically, each point on the map can be associated with a set of data reflecting the ultrasonic characteristics exhibited by the component along an axis perpendicular to the projection plane passing through that point. For example, the associated data set could be the amplitude value of the workpiece's output signal, reflecting the attenuation experienced by the ultrasonic wave as it passes through the component. Because regional defects attenuate the ultrasonic signal penetrating them, these attenuation phenomena will be visible in the ultrasonic map of the workpiece.

[0039] Therefore, the "ultrasonic image of the workpiece under test" is a graphic obtained after the entire workpiece (or a part of the workpiece) has been manufactured.

[0040] A "thickness map" as described herein is a graphical representation of the thickness of a workpiece in a particular direction. Typically, the thickness map is in the same plane as the projection plane of the ultrasonic map of the workpiece, and each point in the projection plane corresponds to a set of data representing the thickness of the workpiece in the direction perpendicular to the projection plane at that point. For example, the thickness map can be a two-dimensional image composed of a plurality of pixels, each pixel having a value calculated based on the thickness of the workpiece at the point corresponding to the pixel

[0041] According to the present method, the thickness map obtained can be a thickness map of the entire workpiece to be inspected, or a thickness map of only a portion of the workpiece. For example, when the workpiece to be inspected is an engine blade formed by bonding a titanium alloy leading edge to a composite body, the thickness map obtained can be a thickness map of only the leading edge, while the ultrasonic map obtained is an ultrasonic map of the entire blade.

[0042] A "simulated ultrasonic map" of a workpiece (or a portion of a workpiece) as described herein refers to a graph that "simulates" the ultrasonic characteristics of the workpiece (or the portion of the workpiece), but the graph is not (at least not entirely) obtained by performing ultrasonic inspection on the workpiece or the portion of the workpiece. Specifically, the simulated ultrasonic map in the present method is determined based on the thickness map. That is, the thickness map is "converted" into an equivalent ultrasonic map that would be obtained if an ultrasonic method were performed on the workpiece or the portion of the workpiece to investigate its ultrasonic characteristics and derive an ultrasonic map, assuming that the workpiece or the portion of the workpiece does not have any defects.

[0043] The simulated ultrasonic map corresponds to an ultrasonic map of a "virtual" workpiece (or a virtual portion of a workpiece) that has similar characteristics to the workpiece (or the portion of the workpiece) being investigated, but does not have any porosity region-like defects. The core principle is that, in the absence of defects, the degree of attenuation of an ultrasonic wave depends on the average attenuation coefficient of the medium through which it propagates and the path length of the ultrasonic signal (i.e. the thickness of the workpiece). By obtaining the attenuation data of a workpiece made of the same medium (which can be a non-homogeneous medium, for example, through which the ultrasonic wave propagates sequentially through a plurality of media), and the actual thickness data of the workpiece to be inspected (i.e. the actual thickness of the workpiece as manufactured, including any thickness deviations from the digital model of the workpiece as manufactured), a simulated ultrasonic map of a workpiece that is free of defects but has the same thickness as the workpiece to be inspected can be derived. Figure 1 Therefore, the simulated ultrasonic map can be used as a reference map of the ultrasonic characteristics of the workpiece in the absence of defects.

[0044] Specifically, the thickness map can be composed of a plurality of pixels, each pixel corresponding to a respective value, and when determining the simulated ultrasonic map, the value of each pixel can be multiplied by a respective coefficient.

[0045] As used herein, the term "component" refers to a set of parameters that determine the structure and composition of a workpiece (or a portion thereof). For example, for a workpiece made of woven composite material, the components specifically include the type of fabric, the type of warp and weft yarns, the ratio of warp to weft yarns, etc. Thus, a "defect-free part" that is identical to the "component and thickness map" of the workpiece under test refers to a virtual workpiece that is identical to the workpiece under test in terms of component and dimensions, but assumes no void area type defects.

[0046] As used herein, the term "detection map" refers to a map used to determine defects on the projection surface of a workpiece. In the present method, the detection map is obtained from the (actual) ultrasonic map of the workpiece under test and the (reference) simulated ultrasonic map.

[0047] The above method has the advantage that the simulated ultrasonic map is constructed using the actual thickness data of the workpiece under test as a reference to determine the presence of void area type defects. Compared to the conventional method in the prior art that uses the ultrasonic map of an "ideal" (physical) workpiece as a reference, the present method is more accurate because the reference part in the prior art can have a thickness difference from the workpiece under test.

[0048] In one or more embodiments, the detection map of the workpiece under test can be composed of a plurality of pixels, each pixel corresponding to a respective numerical value. The step of determining whether the workpiece under test has a void area type defect can include:

[0049] determining whether there is a group of adjacent pixels in the detection map of the workpiece under test, all of the numerical values associated with the pixels in the group being greater than or less than a predetermined threshold value;

[0050] if there is a group of adjacent pixels in the detection map of the workpiece under test, and the numerical values associated with the pixels in the group are lower or higher than the predetermined threshold value, then it is determined that there is a void type defect;

[0051] if there is no group of adjacent pixels in the detection map of the workpiece under test, such that the numerical values associated with the pixels in the group are lower or higher than the predetermined threshold value, then it is determined that there is no void type defect.

[0052] In these embodiments, the detection map is an image composed of a plurality of pixels, each pixel being associated with a respective numerical value. The numerical value associated with a pixel is, for example, a grayscale value or a luminance value. It should be noted that the present application is not limited to each pixel corresponding to only one numerical value. For example, each pixel can correspond to a set of numerical values, each representing the luminance value of the pixel in the red, green, and blue channels.

[0053] As used herein, “adjacent pixels” refer to pixels connected to each other by a connectivity relationship (e.g., 4-connectivity). An “adjacent pixel group” refers to a group of pixels in which each pixel is adjacent to at least one other pixel in the group. The pixel group can include only one pixel, or two or more pixels. In some embodiments, a limit condition can be imposed on the number of pixels in the pixel group (i.e., only an adjacent pixel group with a number of pixels greater than or equal to a preset minimum value can be identified as an “adjacent pixel group”).

[0054] In these embodiments, the pixel group can be a pixel set with a value greater than a preset threshold, or a pixel set with a value less than a preset threshold. The two embodiments are equivalent.

[0055] When the value of a pixel in the image is greater than (or less than) the preset threshold, it indicates that the ultrasound wave attenuates more severely when passing through the region of the workpiece, which can mean that there is a defect in the region.

[0056] In some embodiments, the thickness map can be a thickness map of the entire workpiece. The workpiece thickness map is composed of a plurality of pixels, each pixel corresponding to a respective value. When determining the simulated ultrasound image, the value of at least one pixel in the plurality of pixels in the thickness map can be corrected.

[0057] For example, the value of each pixel in the plurality of pixels in the thickness map of the workpiece to be measured can be multiplied by a respective coefficient, thereby obtaining the simulated ultrasound image. The coefficient can be the same value for all pixels, or each pixel can be assigned a dedicated coefficient.

[0058] It should be noted that the present application is not limited to this embodiment. For example, a simulated thickness map can be automatically generated based on the thickness map of the workpiece by a machine learning algorithm (e.g., a generative adversarial network).

[0059] In these embodiments, the thickness map is for the entire workpiece (rather than a partial region of the workpiece). By converting the thickness map, a simulated ultrasound image is obtained, which is equivalent to an ultrasound image corresponding to a workpiece with similar characteristics to the workpiece to be measured but without defects.

[0060] In some embodiments, the method can further include the following steps:

[0061] re-calibrating the thickness map of the workpiece to be measured based on the ultrasound image of the workpiece to be measured;

[0062] converting the re-calibrated thickness map.

[0063] As used herein, “re-calibration” refers to matching pixels at the same position on the projection surface of the workpiece in two images. The information in the two images is integrated.

[0064] In these embodiments, the recalibration step of the thickness map is performed before the thickness map is converted into the simulated ultrasonic image.

[0065] Alternatively, the method can also comprise the steps of:

[0066] re-calibrating the simulated ultrasonic image based on the ultrasonic image of the workpiece;

[0067] and determining the defect map based on the ultrasonic image of the workpiece and the re-calibrated simulated ultrasonic image of the workpiece.

[0068] In these embodiments, the recalibration step of the simulated ultrasonic image is performed after the thickness map is converted into the simulated ultrasonic image.

[0069] In other alternative embodiments, the workpiece under test is assembled from a first component and a second component by adhesion, wherein the second component is manufactured according to a predetermined digital model. The thickness map is a thickness map of the first component before assembly with the second component, and the ultrasonic image of the workpiece under test is an ultrasonic image of the workpiece under test after assembly. The method can also comprise the steps of:

[0070] obtaining a reference ultrasonic image of the second component of the workpiece under test, the reference ultrasonic image of the second component of the workpiece under test being an ultrasonic image of a second reference component manufactured according to the predetermined digital model and free of the area of porosity type defect;

[0071] determining the simulated ultrasonic image based on the thickness map of the first component of the workpiece under test and the reference ultrasonic image of the second component of the workpiece under test.

[0072] In these embodiments, the workpiece under test comprises at least two components connected by adhesion. The adhesion area usually contains the area of porosity type defect, such as defects caused by lack of glue or presence of bubbles in the adhesion layer. The ultrasonic image is always for the complete (assembled) workpiece, but the thickness map is only for the first component before assembly with the second component. In addition, in these embodiments, a "reference" ultrasonic image of the second component is also obtained, which is usually an ultrasonic image obtained for the second component of a workpiece similar in characteristics to the second component of the workpiece under test but confirmed to be free of defects (same principle as obtaining a reference image for the entire workpiece in the prior art, but here only for the second component of the workpiece).

[0073] Such embodiments are particularly suitable for scenarios where the workpiece comprises a component with small thickness variation and a component with large thickness variation.

[0074] In these embodiments, the thickness map of the first component of the workpiece under test is composed of a plurality of pixels, each pixel being associated with a corresponding numerical value. The method can also comprise the steps of:

[0075] correcting a value of at least one pixel in the plurality of pixels in the first component thickness map of the workpiece under test to obtain a converted thickness map of the first component of the workpiece under test;

[0076] determining the simulated ultrasonic map based on the reference ultrasonic map of the second component of the workpiece under test and the converted thickness map of the first component of the workpiece under test.

[0077] For example, the correction can be done by multiplying the value of each pixel in the plurality of pixels in the first component thickness map by a respective coefficient. The coefficient can be the same for all pixels or can be different for each pixel.

[0078] It is noted that the present application is not limited to this embodiment. For example, the converted thickness map can be automatically generated based on the first component thickness map by a machine learning algorithm, such as a generative adversarial network.

[0079] Thus, the thickness map of the first component of the workpiece is converted into an equivalent ultrasonic map in a defect-free state; the simulated ultrasonic map is determined by fusing the reference ultrasonic map of the second component and the equivalent ultrasonic map of the first component.

[0080] In some embodiments, the method can further comprise the steps of:

[0081] re-calibrating the thickness map of the first component of the workpiece and the ultrasonic map of the second component of the workpiece based on the ultrasonic map of the workpiece;

[0082] converting the re-calibrated thickness map of the first component of the workpiece;

[0083] wherein the simulated ultrasonic map of the workpiece can be obtained by summing the re-calibrated ultrasonic map of the second component of the workpiece and the map obtained by applying the transformation to the re-calibrated thickness map of the workpiece.

[0084] Optionally, the method can further comprise:

[0085] - re-calibrating the map obtained by applying the transformation to the re-calibrated thickness map of the workpiece and the ultrasonic map of the second component of the workpiece based on the ultrasonic map of the workpiece;

[0086] wherein the simulated ultrasonic map of the workpiece can be obtained by summing the re-calibrated ultrasonic map of the second component of the workpiece and the map obtained by applying the transformation to the re-calibrated thickness map of the workpiece.

[0087] In one or more embodiments, the transformation can be a multiplication by a predetermined coefficient.

[0088] In particular, the method can comprise a pre-computation step of a predetermined coefficient, the pre-computation step of the predetermined coefficient comprising:

[0089] for a set of workpieces having the same composition as the workpiece under test and no porosity region-like defects:

[0090] obtaining a thickness map of the workpiece, the thickness map comprising a plurality of pixels each associated with a respective value;

[0091] obtaining an ultrasonic map of the workpiece, the ultrasonic map comprising a plurality of pixels each associated with a respective value;

[0092] a ratio of the value of a thickness map pixel to the value of a corresponding ultrasonic map pixel;

[0093] based on the determined plurality of ratios, calculating a coefficient.

[0094] In some embodiments, the detection map is obtained by a difference operation between the ultrasonic map of the workpiece under test and the simulated ultrasonic map.

[0095] For example, the workpiece under test can be an aeronautical workpiece. In particular, when the workpiece comprises two parts, the workpiece can be an engine blade, wherein the first part is a metal leading edge and the second part is a woven composite body.

[0096] In some embodiments, the porosity region-like defects include cracks, fissures, delaminations or peeling.

[0097] In some embodiments, each of the above-mentioned maps is a C-scan.

[0098] Another aspect of the application relates to a device for ultrasonic detection of porosity region-like defects in a workpiece, the device comprising:

[0099] an input interface configured to:

[0100] receive an ultrasonic map of the workpiece under test;

[0101] receive a thickness map of at least a portion of the workpiece under test;

[0102] a circuit configured to:

[0103] based on the thickness map of at least a portion of the workpiece under test, determine a simulated ultrasonic map of a corresponding portion of a workpiece having no porosity region-like defects, the workpiece having the same composition and thickness map as the corresponding portion of the workpiece under test;

[0104] based on the ultrasonic map of the workpiece under test and the simulated ultrasonic map, determine a detection map of the workpiece under test;

[0105] by applying a predetermined threshold to the detection map of the workpiece under test, determine whether the workpiece under test has a porosity region-like defect.

[0106] For a computer program that can be installed and used on existing equipment and can implement some or all of the above method steps, it has significant advantages itself.

[0107] Therefore, the present application also relates to a computer program product, which comprises a series of instructions that can implement some steps of the above method when the program is executed on a processor.

[0108] The program can be written in any programming language (such as object-oriented language or other language), and its form can be interpretable source code, partially compiled code or fully compiled code.

[0109] The figure 3 to be described in detail below can be used as a general algorithm flowchart of the computer program.

[0110] The present application and its various application scenarios can be more fully understood in combination with the following description and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0111] Other features and advantages of the present application can be further understood by reading the following description in conjunction with the drawings, which are merely illustrative and not limiting.

[0112] Fig. 1a-1e are schematic diagrams of some steps of the existing art method for detecting air hole area type defects;

[0113] Fig. 2 is a flowchart of the existing art method for detecting air hole area type defects;

[0114] Fig. 3 is a flowchart of the method for detecting air hole area type defects in one embodiment of the present application;

[0115] Fig. 4 is a flowchart of the method for detecting air hole area type defects in another embodiment of the present application;

[0116] Fig. 5a-5c are schematic diagrams of the determination steps of the simulated ultrasonic wave of the workpiece in one embodiment of the present application;

[0117] Fig. 6 is a flowchart of the method for detecting air hole area type defects in an embodiment of the present application, which includes a threshold value determination step;

[0118] Fig. 7 is an example diagram of the device for detecting air hole area type defects in an embodiment of the present application. DETAILED DESCRIPTION

[0119] Fig. 3 is a flowchart of the method for detecting air hole area type defects in one embodiment of the present application.

[0120] In this embodiment, the workpiece is detected as a whole (while in the embodiment shown in Fig. 4, the workpiece is regarded as an assembly composed of two bonded parts).

[0121] Step 220 is the same as that described in FIG. 2, obtaining an ultrasonic image (e.g., a C-scan image) of the workpiece. The ultrasonic image is typically an image comprising a plurality of pixels, and the brightness value of each pixel represents the amplitude of the received signal (or, the attenuation value of the ultrasonic wave when propagating through the workpiece, also referred to as the ultrasonic attenuation value) of the workpiece region corresponding to the pixel. That is, the value of each pixel reflects the attenuation of the transmitted ultrasonic wave when propagating through the workpiece. Therefore, the region of the workpiece where a pore or similar defect exists will typically exhibit a stronger ultrasonic signal attenuation in the corresponding pixel brightness value.

[0122] Step 310: Obtain a thickness map of the workpiece. The thickness map of the workpiece is typically an image comprising a plurality of pixels, and the brightness value of each pixel represents the thickness of the workpiece in a particular direction. For example, if the workpiece is in a coordinate system (X, Y, Z), the thickness map along the Z-axis direction represents the distribution of the thickness of the workpiece in the Z-axis direction on the (X, Y) plane.

[0123] The thickness map can be obtained by a size detection method in the prior art. For example, a three-dimensional measuring machine (TMM) can be used to obtain measurement data, and then a thickness map is generated. The three-dimensional measuring machine moves the probe and / or optical sensor through the measuring arm to determine the size, shape and position of the object to be measured in a particular coordinate system, thereby obtaining the thickness data of the object in multiple directions. Subsequently, the measurement data can be processed by, for example, metrology software to model the object and detect any dimensional deviations of the workpiece relative to its digital manufacturing model.

[0124] The ultrasonic image of the workpiece obtained in step 220 and the thickness map of the workpiece obtained in step 310 are typically images of the same size (i.e., the number of pixels in the two coordinate axes of the image is consistent). It should be noted that steps 220 and 310 can be executed in any order or in parallel.

[0125] In one or more embodiments, a calibration step 320 can be performed to recalibrate the ultrasonic image and the thickness map of the workpiece. In fact, due to the different ways of obtaining the two types of images, the images can not correspond completely. For example, the pixel coordinates corresponding to a point on the workpiece in the ultrasonic image and in the thickness map are different. Typically, there can be a rotation and / or translation relationship between the images. In order to match the two types of images and integrate the information of each of them, one of the images needs to be transformed to achieve "recalibration" with the other image.

[0126] Image re-calibration techniques between images are well known to those skilled in the art. In particular, re-calibration techniques can be used based on fiducials of known locations on the workpiece. Pixels corresponding to these fiducials are identified in both images and a transformation is determined based on these pixels to apply to re-calibrate. For example, if M points on the workpiece correspond to pixels p (with coordinates (x,y)) in the ultrasound image and pixels p' (with coordinates (x',y')) in the thickness image, then a transformation T needs to satisfy T (x',y') = (x,y) (or T (x,y) = (x',y')). Such techniques are well known in the art and will not be described in detail herein. Figures 5a-5c illustrate the process of re-calibration in one embodiment of the present application.

[0127] In one or more embodiments, step 330 performs a transformation operation on the thickness image to modify the intensity values of the pixels in the thickness image to obtain a so-called "simulated" ultrasound image of a workpiece of similar characteristics (i.e. same composition, structure, dimensions) to the workpiece under consideration but without defects. After step 330 is performed, a simulated ultrasound image of the workpiece is obtained. The term "simulated" as used herein means that the image is generated (at least in part) from the thickness image and has the effect of a "simulated" ultrasound image. It is important to note that "simulated" in this document does not mean a map generated entirely by computer simulation.

[0128] For example, in step 330, the intensity value of each pixel in the thickness image can be multiplied by a same predetermined coefficient to obtain the simulated ultrasound image. The coefficient can be determined prior to performing the method of the present application, for example, by selecting a workpiece having a plurality of regions of known thickness and obtaining an ultrasound image of the workpiece. For example, a wedge block having a plurality of regions of different thicknesses (e.g. 5 steps, the first step, second step,... have thicknesses of 1 mm, 2 mm,... 5 mm respectively, the number of steps and thicknesses are examples only and do not limit the present application) can be used. The thickness of each step is known. The ultrasound attenuation values of each step region (e.g. the first step has an attenuation value of -3 dB, the second step has an attenuation value of -4 dB,... the fifth step has an attenuation value of -7 dB) are obtained from the ultrasound image. Then, a relationship between the thickness value and the corresponding attenuation value is established, for example, the relationship Att = (Ep x n) + z is obtained, where Att represents the ultrasound attenuation, Ep represents the thickness, and n and z are two parameters determined based on the wedge block experiment. In addition to the linear model described above, other relationship models between the ultrasound attenuation and the thickness can also exist.

[0129] Alternatively, different coefficients can be set for different regions of the thickness image of the workpiece. For example, different regions correspond to regions of different materials of the workpiece (and therefore have different average attenuation coefficients), the coefficients applied to the pixels in each region can be determined in the manner described above.

[0130] The coefficients can also be determined by other methods, such as machine learning methods. For example, a machine learning model can be trained using a training database comprising pairs of maps, each pair consisting of a thickness map and an ultrasonic map of the same workpiece, so that the model is able to determine the coefficients to be applied to generate an "equivalent" ultrasonic map from a thickness map. Such a model can be, for example, a neural network, in particular a generative adversarial network (GAN), but the application is not limited to this example.

[0131] At the end of step 330, the resulting simulated ultrasonic map has a similarity with the actual ultrasonic map, i.e. the intensity values of the pixels in the simulated ultrasonic map are approximately equivalent to the intensity values of the corresponding pixels in the actual ultrasonic map. That is, by the conversion operation, a conversion of thickness values into equivalent ultrasonic attenuation values is achieved.

[0132] The conversion operation in step 330 effectively improves the comparability of the intensity values of the pixels in the ultrasonic map and the thickness map, so as to integrate the information of the two maps.

[0133] The step 330 described above is a conversion operation performed on a thickness map of a workpiece. However, the application is not limited thereto, or a conversion operation can also be performed on an ultrasonic map of a workpiece to correct the intensity values of the pixels thereof to match the intensity values of the corresponding pixels in a thickness map of the workpiece, similar to the process described above.

[0134] In step 340, the simulated ultrasonic map determined in step 330 is subtracted from the ultrasonic map received in step 220. Of course, it is also possible to subtract the ultrasonic map from the simulated ultrasonic map. The "subtraction" described herein refers to pixel-by-pixel subtraction, i.e. the value of each pixel in the subtracted image corresponds to the difference between the values of the pixels at the same coordinates in the two images.

[0135] The image obtained after step 340 is called a "detection map", which clearly presents the areas of porosity-like defects. These areas indeed correspond to higher pixel values. This is because, in the absence of defects, the ultrasonic attenuation value is directly related to the thickness of the workpiece; when the ultrasonic signal passes through a porosity-like defect, the attenuation is significantly enhanced. Therefore, the value of the ultrasonic map for a pixel corresponding to a defect (including the effect of the defect) and the equivalent attenuation value of the thickness map (in the absence of defects) will be significantly different. Therefore, for a pixel corresponding to a defect, there will be a significant difference between the value of the ultrasonic map (including the effect of the defect) and the value of the thickness map (representing the thickness value "equivalent" to the attenuation value of the workpiece in the absence of defects).

[0136] The false detection rate (i.e. false ultrasonic indication regions) of the detection image obtained by the present method step 340 is lower than that of the processed image obtained by the prior art step 230 shown in Fig. 2, because the detection image is greatly reduced in the effect of material difference. Therefore, the reliability of the detection image is significantly better than the prior art.

[0137] Similar to the method shown in Fig. 2, a threshold processing step 260 can be performed on the detection image obtained by step 340. Specifically, only the pixels with values higher than a preset threshold value can be retained. In one or more embodiments, the threshold value can be determined by the method described in Fig. 2 (step 250). For example, the threshold value can be set to correspond to a decay percentage of -6dB.

[0138] In one embodiment, the pixels with values higher than the threshold value can be set to a first reference value, such as a gray value of 0 (i.e. black pixels); and the pixels with values lower than the threshold value can be set to a second reference value, such as a gray value of 255 (i.e. white pixels). The threshold detection image is a black and white image, in which the defects are shown in black and the background is white. Alternatively, the original values of the pixels with values higher than the threshold value can be retained, and the pixels with values lower than the threshold value can be set to the reference value (such as a gray value of 255, i.e. white pixels). In this case, the background of the detection image is white, and the defect regions retain their brightness values calculated in step 260. Other threshold processing methods can also be used in the present application.

[0139] Similar to the method shown in Fig. 2, step 270 can determine whether there is a pore region type defect based on the threshold detection image obtained by step 260. According to one embodiment, when all the pixel values in a group of adjacent pixels exceed the predetermined threshold value used in step 260, the region is determined to be a defect. The "adjacent pixels" described herein refer to a set of pixels connected to each other by a connection relationship (such as 4-connection). In some embodiments, a minimum number of adjacent pixels Nmin can be set for the group of adjacent pixels. According to these embodiments, the region to be determined as a defect must contain at least Nmin adjacent pixels. If there is no group of adjacent pixels with values exceeding the predetermined threshold value, it indicates that the workpiece being tested does not have a defect.

[0140] Fig. 4 is a flow chart of a pore region type defect detection method according to another embodiment of the present application.

[0141] In this embodiment, the workpiece to be tested is assembled by bonding two components (a first component and a second component). For example, the workpiece to be tested can be an engine blade including a body and a leading edge; the body is made of woven composite material, and the leading edge is made of metal (e.g., titanium alloy). The body and the leading edge are respectively manufactured and then assembled by bonding to form the engine blade. For such a workpiece including bonded components, porosity-like defects are prone to occur in the bonding area (e.g., due to lack of glue, no glue, or presence of porous area, etc.), which reduces the strength of the workpiece, and thus such defects must be detected. It is noted that this embodiment can be extended to a workpiece assembled by bonding more than two components.

[0142] Step 220: Obtain an ultrasonic image (e.g., a C-scan image) of the complete workpiece after assembly (i.e., the components of the workpiece have been bonded). This step is the same as step 220 described in FIG. 2 or FIG. 3. Taking the engine blade as an example, this step 220 obtains an ultrasonic image of the blade after assembly.

[0143] Step 302: Obtain a reference ultrasonic image (e.g., a C-scan image) of the first component of the workpiece. The “reference ultrasonic image” described herein does not necessarily have to be an ultrasonic image of the first component of the workpiece to be tested, but can also be an ultrasonic image of the first component of another workpiece of the same type as the first component of the workpiece to be tested.

[0144] Taking the engine blade as an example, the first component of the workpiece is usually a composite material body. Such a body is usually manufactured according to a digital reference model specific to a certain type of blade, and this model is common to all blades manufactured based on the certain type of blade. Therefore, the reference ultrasonic image obtained in step 302 is of a blade body associated with the blade to be tested and manufactured according to the digital reference model. Generally, the thickness of the first component of the workpiece varies relatively little between different workpieces (components made of woven composite material belong to this category).

[0145] It can be seen that before the detection method shown in FIG. 4 is performed, a same reference ultrasonic image of the first component of the workpiece can be determined in advance, and this reference image can be used to detect multiple different workpieces manufactured based on the same digital reference model specific to the first component of this type of workpiece (i.e., this reference image can be used multiple times in step 302 of the detection method shown in FIG. 4 for multiple workpieces to be tested).

[0146] It is noted that the reference ultrasonic image of the first component of the workpiece obtained in step 302 is an ultrasonic image of the first component of the workpiece before being bonded with the second component.

[0147] Step 304: Acquire a thickness map of the second part of the piece under test before bonding with the first part. In the example above, this is thus the thickness map of the titanium alloy nose of the piece under test, before it is bonded to the woven composite material blade body. In fact, in an engine blade comprising a bonded assembly of a woven composite material body and a titanium alloy nose, it is the titanium alloy nose that exhibits the greatest thickness variability between parts. It is thus this part of the piece under discussion that needs to be known with precision. Typically, the second part of the piece is chosen as the part of the piece for which the thickness variability is greatest between parts (and thus also with respect to the reference model according to which the second part of the piece was manufactured).

[0148] It is thus noted that step 302 relates to a reference ultrasound map, while step 304 relates to a thickness map of the second part of the piece under discussion (the piece under test).

[0149] The thickness map of the second part of the piece under test is typically obtained by a piece thickness measurement technique as described above, with reference to step 310 of the method of Figure 3 .

[0150] In step 325, the different maps can be recalibrated. This recalibration allows to "register" (i.e. reassign to the same coordinates so that all maps correspond to the same image of the piece) the pixels of the different maps that correspond to the same points on the piece.

[0151] Figure 5a , 5b and 5c illustrate one example of such a recalibration 325. Figure 5a The ultrasound map 510 represents the piece under test as it is being assembled. The crosses on the ultrasound map 510 represent "reference" points used for the recalibration between different images. In fact, these reference points are placed according to physical elements visible on the different maps of the piece under discussion before assembly. The ultrasound map 510 of the piece under test as it is being assembled shows two parts 512 and 514 of the piece, i.e. in the blade example, the nose and the body respectively.

[0152] Figure 5a The reference ultrasound map 520 of the second part of the piece 514 (here the blade body) is also represented, as well as the reference points (again represented by crosses).

[0153] The reference points of the maps 510 and 520 are matched to determine a transformation that allows to recalibrate the reference ultrasound map 520 of the second part of the piece with the part 514 of the ultrasound map 510 (the part 512 being ignored in this recalibration step). In the illustrated case, the transformation is a combination of a translation, a rotation and a scaling. Once the transformation is determined and applied to all pixels of the reference ultrasound map 520 of the second part of the piece, the transformed reference ultrasound map 530 of the second part of the piece is obtained.

[0154] Figure 5b represents the same, assembled, workpiece under test ultrasonic image 510. Figure 5a represents the same, assembled, workpiece under test ultrasonic image 510. Figure 5b Also represented are a thickness map 540 of the workpiece first component (here, the leading edge), as well as fiducials (again, represented by crosses).

[0155] As previously described, the fiducials of the maps 510 and 540 are matched to re-calibrate (step 325) the thickness map 540 of the workpiece first component to the ultrasonic image 510 of the workpiece under test being assembled. In addition, a transformation can be applied to the thickness map 540 of the workpiece first component (either before or after the re-calibration) so that the light intensity levels of the pixels of the workpiece first component thickness map 540 generally correspond to the light intensity levels of the corresponding pixels (i.e., representing the same points on the workpiece after the re-calibration), as previously described with reference to Figure 3 Step 330. After the re-calibration and transformation (e.g., multiplying the pixel values of the workpiece first component thickness map 540 by one or more coefficients), a simulated ultrasonic image 550 of the workpiece first component is obtained that is re-calibrated to the ultrasonic image 510 (more precisely, the portion 512 of the image 510) of the complete workpiece.

[0156] Figure 5c represents a simulated ultrasonic image 560 of the workpiece obtained from the transformed fiducial ultrasonic image 530 of the workpiece second component and the simulated ultrasonic image 550 of the workpiece first component. For example, this simulated ultrasonic image 560 of the complete workpiece is obtained by pixel-wise adding the images 530 and 550. Note here that the word "simulated" is used to designate the fact that the image 560 is a simulated ultrasonic image, and it is derived in part from thickness data (for the workpiece first component, i.e., the titanium alloy leading edge in the previous example).

[0157] Thus, at the end of step 330, a simulated ultrasonic image 560 of the workpiece as represented by the element 560 in Figure 5c is obtained.

[0158] Figure 4 Steps 340, 260 and 270 in Figure 3 are similar to steps 340, 260 and 270 in Thus, in step 340, a detection map is obtained by subtracting (or in another order) the simulated ultrasonic image determined in step 330 from the ultrasonic image received in step 220. Then, a threshold can be applied to the detection map obtained in step 340 in step 260, and it can be determined in step 270 whether a porosity region-type defect is present based on the thresholded detection map obtained in step 260.

[0159] Figure 6 represents a flowchart of a method of detecting a porosity region-type defect, the method including determining a threshold to be applied according to an embodiment of the present application.

[0160] During step 330, which replaces step 210 in Figure 2 , a simulated ultrasonic image of the workpiece is obtained. Step 330 can be implemented, for example, as previously described with reference to Figure 3 and Figure 4 . In step 220, an ultrasonic image of the workpiece to be tested is obtained. The simulated ultrasonic image of the workpiece can be subtracted from the C-scan image of the workpiece to be tested (step 340) to obtain a detection image.

[0161] As described in Figure 2 , during step 240, a C-scan image of a defective workpiece (the defects being known, as described above) is obtained and used to determine an optimal threshold (step 250) to detect defects, as described above. In step 260, the threshold determined in step 250 can be applied to the processed image obtained in step 340. Defects can then be identified in the thresholded image in step 270.

[0162] Steps 330, 220 and and 240 can be implemented in parallel or in any order (although typically, steps 330 and 240 of the reference workpiece and of the defective workpiece are implemented before starting the actual testing of the workpiece, i.e. step 220).

[0163] Thus, the method of Figure 6 is similar to the method of Figure 2 , but instead of using an ultrasonic image of a reference workpiece, a simulated ultrasonic image of the workpiece is used, the image being determined from the thickness data of the manufactured workpiece. The thickness variability related to the manufacturing of the workpiece (and thus not necessarily indicative of the presence of defects) is thus advantageously taken into account and integrated into the detection method. The resulting detection method is thus more accurate than the prior art method described with reference to Figure 2 , and produces fewer false positives related to the manufacturing variability of the workpiece.

[0164] Figure 7 represents an example of a porosity area type defect detection device according to an embodiment of the application.

[0165] In these embodiments, the device comprises a computer 700 comprising a memory 701 for storing instructions implementing the method, different images from which the detection method is implemented, and temporary data for executing the different steps of the detection method previously described.

[0166] The computer 700 also comprises a circuit 702. This circuit can be, for example, a processor capable of interpreting instructions in the form of a computer program, an electronic card on which the steps of the method of the application are described in silicon, or a programmable electronic chip, such as a Field-Programmable Gate Array (FPGA) chip.

[0167] The computer 700 comprises an input interface 703 for receiving the ultrasonic image and / or the thickness map, and an output interface 704 for providing the detection map or one or more information related to the defect detection (e.g. an indication about the presence or absence of a defect, and / or, when a defect is detected, an indication about the location of the defect). Finally, the computer can comprise a screen 705 and a keyboard 706 to facilitate the interaction with the user. Of course, the keyboard is optional, especially in the scope of computers in the form of touch screen tablets.

[0168] Moreover, Figure 3 and Figure 4 The functional block diagram set out in the description of the figures is a typical example of a program for which some instructions can be executed using the device. Thus, Figure 3 The general algorithm flowchart can correspond to one computer program for the purposes of the application.

[0169] Of course, the application is not limited to the embodiments described by way of example. It extends to other variants. For example, the method described previously is advantageously applicable to the case of ultrasonic testing of industrial workpieces having complex geometries and thickness variations that can lead to false indications. Such industrial workpieces can comprise bonded or welded assemblies for which it is desirable to guarantee the quality of the bonding or welding, in particular to determine whether they present areas of material deficiency, porosities or areas with foreign bodies. Such industrial workpieces can also comprise homogeneous workpieces that are subjected to material health checks in which volumetric defects such as delaminations or inclusions are sought.

[0170] In order to obtain equivalent inspection results throughout the inspection area and to reliably detect any defects, it is currently necessary to correct the variations in attenuation related to the different thicknesses. This correction can be digital, by multiple gain control, or physical, by using a reference workpiece.

[0171] In these cases, the application makes it possible to reduce the acquisition time by performing a single gain acquisition and also to eliminate false indications caused by the differences between the production workpiece and the reference workpiece.

Claims

1. A computer-implemented method for ultrasonically detecting porosity area type defects in a workpiece under test, the method comprising: - receiving an ultrasonic image of the workpiece under test; - receiving a thickness map of at least one portion of the workpiece under test; - determining a simulated ultrasonic image of a corresponding area of a porosity area type defect free workpiece based on the thickness map of at least one portion of the workpiece under test, the porosity area type defect free workpiece having the same composition and thickness map as the corresponding area of the workpiece under test; - determining a detection image of the workpiece under test based on the ultrasonic image of the workpiece under test and the simulated ultrasonic image; - determining the presence or absence of porosity area type defects in the workpiece under test by applying a predefined threshold to the detection image of the workpiece under test.

2. The method of claim 1, wherein, The thickness map is a thickness map of the entire workpiece; wherein the thickness map of the workpiece under test comprises a plurality of pixels, each of the plurality of pixels being associated with a respective value; wherein the simulated ultrasonic image is determined by modifying the value of at least one of the plurality of pixels of the thickness map of the workpiece under test.

3. The method of claim 1, wherein, The workpiece under test comprises a first portion and a second portion assembled by adhesion, the second portion being manufactured according to a predefined digital model; wherein the thickness map is a thickness map of the first portion of the workpiece under test prior to assembly with the second portion of the workpiece under test; wherein the ultrasonic image of the workpiece under test is an ultrasonic image of the assembled workpiece under test, the method further comprising: - receiving a reference ultrasonic image of the second portion of the workpiece under test, the reference ultrasonic image of the second portion of the workpiece under test corresponding to an ultrasonic image of a second reference workpiece manufactured according to the predefined digital model and not containing any area type defects; wherein the simulated ultrasonic image is determined based on the thickness map of the first portion of the workpiece under test and the reference ultrasonic image of the second portion of the workpiece under test.

4. The method of claim 3, wherein, The thickness map of the first portion of the workpiece under test comprises a plurality of pixels, each of the plurality of pixels being associated with a respective value, the method further comprising: - modifying the value of at least one of the plurality of pixels of the thickness map of the first portion of the workpiece under test to obtain a so-called transformed thickness map of the first portion of the workpiece under test; - wherein the simulated ultrasonic image is determined based on the reference ultrasonic image of the second portion of the workpiece under test and the transformed thickness map of the first portion of the workpiece under test.

5. The method according to any of the preceding claims, wherein, The detection image is obtained by subtracting the ultrasonic image of the workpiece under test from the simulated ultrasonic image.

6. The method according to any one of the preceding claims, wherein, The workpiece is an aeronautical piece.

7. The method according to the preceding claim in combination with one of claims 3 or 4, wherein, The workpiece is an engine blade, wherein the first portion of the workpiece is a metal leading edge and wherein the second portion of the workpiece is a main body made of woven composite material.

8. The method of any of the preceding claims, wherein, The porosity area type defects are the presence of cracks, fissures, delaminations or peeling.

9. A device for ultrasonically detecting porosity area type defects in a workpiece under test, the device comprising: - an input interface configured to: o receive an ultrasonic image of the workpiece under test; o receive a thickness map of at least one portion of the workpiece under test; - a circuit configured to: o determining a simulated ultrasonic map of a region of the workpiece under test corresponding to the region of the workpiece under test based on the thickness map of at least one portion of the workpiece under test, the workpiece under test having the same composition and thickness map as the region of the workpiece under test; o determining a detection map of the workpiece under test based on the ultrasonic map of the workpiece under test and the simulated ultrasonic map; o determining the presence or absence of a region of porosity type defect in the workpiece under test by applying a predefined threshold to the detection map of the workpiece under test.

10. A computer program product comprising instructions for implementing the method according to any one of claims 1 to 8 when the program is executed by a processor.

Images