Precise forging method of new energy automobile aluminum alloy forging production line

Abstract

The invention relates to the technical field of metal forging, in particular to a precision forging method of a new energy automobile aluminum alloy forging production line. The method comprises the following steps: acquiring digital ray images of a prefabricated member at different angles; obtaining a target frame and a confidence coefficient by taking the defect position as an input, obtaining a gray abnormal value of the target frame based on the gray difference of the target frame, and forming a density abnormal value of the defect area by the confidence coefficient and the gray abnormal value of the defect position at different angles; extracting an edge curve, obtaining an edge difference based on the curvature of the edge curve and the distance of the discrete points, and constructing a volume intersection by taking the discrete points as a core to determine a parting difference between the discrete points; on the basis of powder difference clustering, obtaining a flash correction coefficient according to the distance between the defect area and the clustering cluster and the volume and density abnormal values; calculating the thickness of the gasket; and the flash groove is adjusted through the thickness of the gasket to forge the workpiece. Local adjustability is achieved, and the forging efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of metal forging technology, and in particular to a precision forging method for an aluminum alloy forging production line for new energy vehicles. Background Art

[0002] With the rapid development of the new energy vehicle industry, the proportion of lightweight aluminum alloy parts used in complete vehicles continues to rise. In order to balance structural strength and forming efficiency, aluminum alloy parts often adopt a composite manufacturing process of "casting + forging". First, a near-net-shape preform is quickly obtained through extrusion casting, and then precision forging is performed to improve the material density and mechanical properties. Among them, the flash design in the forging process plays a key role in the quality of the final forging. The existence of flash not only helps the metal fill the mold cavity, but also forms a high three-dimensional compressive stress area in the flash groove, thereby improving the organizational density, fatigue strength and overall mechanical properties of the forging. Therefore, it is of great significance in precision forging.

[0003] In traditional forging processes, the flash size is usually fixed during the mold design stage and cannot be dynamically adjusted according to the individual differences of each die-cast preform. Due to common defects such as uneven casting stress, local voids, and shrinkage in the aluminum alloy die-casting process, different batches or even different workpieces in the same batch may have density differences in local areas. In this case, if a uniform flash bridge thickness is still used, it is often impossible to provide sufficient three-dimensional compressive stress for the low-density area, resulting in the defects not being effectively repaired after forging, which in turn affects the overall mechanical properties. Remanufacturing multiple sets of molds to cope with the differences in different regions is not only costly, but also complicated and inefficient in mold replacement operations. Therefore, there is an urgent need for a method that can quickly adjust the compressive stress distribution in the flash area without changing the mold to adapt to the quality status of different die-cast workpieces, thereby improving the precision forging density and pass rate. Summary of the Invention

[0004] In order to solve the technical problem of low efficiency caused by die flash fixation, this application provides a precision forging method for a new energy vehicle aluminum alloy forging production line. The technical solution adopted is as follows:

[0005] This application proposes a precision forging method for a new energy vehicle aluminum alloy forging production line, which includes the following steps:

[0006] Obtain a preform after die-casting the aluminum alloy, and obtain digital radiographic images of the preform at different angles;

[0007] Digital radiography is used as neural network input to obtain the target frame and confidence level. At each angle, the grayscale outlier of the target frame is obtained based on the grayscale difference between the target frame and the target frame with the minimum grayscale value. The intersection of the defect positions corresponding to different angles is projected into a voxel column in space to obtain the 3D defect area. Based on the confidence level and grayscale outlier of the defect position at all angles, the density outlier of the corresponding defect area is obtained.

[0008] Extract the edge curve of the parting surface in the mold 3D model, divide the edge curve into multiple discrete points, and calculate the edge difference between the discrete points based on the vector difference and distance of each discrete point. Construct a sphere with each discrete point as the core, and obtain the parting difference between the discrete points based on the volume ratio difference and edge difference between the sphere and the 3D model.

[0009] All discrete points are clustered into different clusters using the difference in mold separation as the cluster distance. The flash correction coefficient of each cluster is obtained based on the volume and density outliers of all defect areas and their distance from the cluster. The gasket thickness of different clusters is calculated based on the flash correction coefficient.

[0010] Different gasket thicknesses are selected for different clusters to forge the workpiece.

[0011] In the above scheme, this application addresses the technical bottleneck of traditional die flash structures being fixed and lacking adaptability to individual differences in die-cast parts. By introducing X-ray nondestructive testing of die-cast parts, this method acquires internal defect distribution data for die-cast and forged workpieces. On this basis, a regional division mechanism for the parting surface edge curve is established. In combination with density anomaly information in the defective area, the flash correction coefficient is used to guide the adjustment of the flash bridge thickness. By installing shims of varying thicknesses in the flash groove, the local adjustability and dynamic optimization of the die flash structure are achieved. Not only does this method achieve differentiated adjustment of the flash structure without replacing the entire die set, significantly improving the die's ability to adapt to different die-casting conditions, but it also enhances the three-dimensional compressive stress effect during the forging process through regional density analysis, improving the plastic flow effect and microstructure density of the material, thereby effectively reducing forging defects and improving structural strength and fatigue life. Furthermore, combined with a quick die assembly and disassembly design and standardized shim assembly, this method enables rapid response and low-cost adaptation, greatly improving production efficiency. This method is suitable for applications such as new energy vehicles that require lightweight, high-performance, and high-consistency structural components, and has significant engineering practical value and promotion prospects.

[0012] In one embodiment, the digital radiographs are acquired at ±30°, ±60°, and 90° of the preform, respectively.

[0013] In one embodiment, the method for obtaining the grayscale outlier value of the target frame based on the grayscale difference between the target frame and the target frame with the minimum grayscale value is:

[0014] g z,r It represents the grayscale mean of the rth defect position at the zth angle, min() represents the minimum function, represents the grayscale mean of all defect positions at the zth angle, l z,r It represents the grayscale abnormal value of the rth defect position at the zth angle, and each target box is a defect position.

[0015] In one embodiment, the density anomaly of the defect area is positively correlated with the confidence level and grayscale anomaly value of the defect location.

[0016] In one embodiment, the edge difference is positively correlated with the vector difference and the distance between discrete points.

[0017] In one embodiment, the vector difference is the difference between the vectors of two discrete points, and the vector of each discrete point is a vector formed by the discrete point pointing to the next discrete point.

[0018] In one embodiment, the mold separation difference is positively correlated with the volume share difference and the edge difference respectively; the volume share difference is the absolute value of the difference in volume share of two discrete points; the volume share of the discrete points is the ratio of the intersection volume of the sphere corresponding to each discrete point and the 3D model to the volume of the sphere.

[0019] In one embodiment, the flash correction coefficient is positively correlated with the volume and density anomaly of the defect area, and negatively correlated with the distance between the defect area and the cluster.

[0020] In one embodiment, the distance between the defect area and the cluster is the minimum value of the distance between the matching point in the defect area and the discrete points in the cluster, and the matching point is the point corresponding to each discrete point after the mold and the preform are matched.

[0021] In one embodiment, the method for calculating the gasket thickness of different clusters based on the flash correction coefficient is:

[0022] ψ u represents the flash correction coefficient of the u-th cluster, norm() represents the normalization function, ceil{} represents the rounding function, Δh u Indicates the thickness of the gasket corresponding to the discrete point in the u-th cluster.

[0023] The beneficial effects of this application are:

[0024] This application addresses the technical bottleneck of traditional die flash structures being fixed and lacking adaptability to individual differences in die castings. By introducing X-ray nondestructive testing of die castings, this method acquires internal defect distribution data for die-cast and forged workpieces. Based on this, a regional division mechanism for the parting surface edge curve is established. In combination with density anomaly information in the defective area, the flash correction coefficient is used to guide the adjustment of the flash bridge thickness. By installing shims of varying thicknesses in the flash groove, the die flash structure is locally adjustable and dynamically optimized. This method not only achieves differentiated adjustment of the flash structure without replacing the entire die, significantly improving the die's ability to adapt to different die casting conditions, but also enhances the three-dimensional compressive stress effect during the forging process through regional density analysis, improving the plastic flow effect and microstructure density of the material, thereby effectively reducing forging defects and improving structural strength and fatigue life. Furthermore, the combination of a quick die assembly and disassembly design and standardized shim assembly enables rapid response and low-cost adaptation, significantly improving production efficiency. This method is suitable for applications such as new energy vehicles that require lightweight, high-performance, and high-consistency structural components, and has significant engineering practical value and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0026] Figure 1 A flow chart of a precision forging method for a new energy vehicle aluminum alloy forging production line provided in one embodiment of the present application;

[0027] Figure 2 It is a schematic diagram of the flash groove;

[0028] Figure 3 Schematic diagram of flash groove gasket. DETAILED DESCRIPTION

[0029] To further illustrate the technical means and effects adopted by this application to achieve the intended invention objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effects of a precision forging method for a new energy vehicle aluminum alloy forging production line proposed in this application. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable form.

[0030] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0031] An embodiment of a precision forging method for a new energy vehicle aluminum alloy forging production line:

[0032] The specific scheme of the precision forging method of the aluminum alloy forging production line for new energy vehicles provided by this application is described in detail below with reference to the accompanying drawings.

[0033] See also Figure 1 , which shows a flow chart of a precision forging method for a new energy vehicle aluminum alloy forging production line provided by one embodiment of the present application, the method comprising the following steps:

[0034] Step S001: Obtain a preform after die-casting the aluminum alloy, and obtain digital radiographic images of the preform at different angles.

[0035] Aluminum alloy plays a huge role in many parts of new energy vehicles. Aluminum alloy can be used as different workpieces in new energy vehicles such as steering knuckles, subframes, and wheels.

[0036] The forging process for aluminum alloy workpieces is as follows: the aluminum material is heated to a molten state in a furnace in preparation for casting; the molten aluminum is poured into a mold and rapidly cooled and formed under high pressure to form the initial workpiece shape; after the die-casting is completed, the gate part of the workpiece used to deliver the molten aluminum needs to be removed; the die-cast workpiece is then heated to the forging temperature and finely forged to further shape the workpiece's shape and size; finally, the workpiece is precision machined to achieve the precise size and surface quality requirements of the final product.

[0037] In traditional forging processes, the same flash groove parameters are used. However, due to different workpieces and different installation positions, the extrusion force is uneven. Insufficient extrusion force in some areas can lead to excessively large grains and affect mechanical properties. The flash groove is a gap groove structure set around the final forging die cavity of the hammer forging. It consists of a flash bridge (bridge part) connected to the die cavity and a peripheral flash bin (bin part).

[0038] To assess the density and morphology of aluminum alloy workpieces after die-casting, the system needs to inspect the preforms (semi-finished workpieces) after die-casting. Based on the inspection of the preforms after die-casting, the mold can be adjusted.

[0039] Industrial X-ray inspection equipment is used to perform full-area radiographic imaging of the die-cast preform to obtain information on its internal structural defects, including local defects such as pores and cracks. The die-cast preform is scanned based on the DR digital radiographic imaging results obtained at multiple different angles, and DR digital radiographic images of the preform at ±30°, ±60°, and 90° are obtained respectively.

[0040] At this point, DR ray digital imaging of the prefabricated part at different angles has been obtained.

[0041] Step S002: Get the target frame and confidence, get the grayscale outlier of the target frame based on the grayscale difference of the target frame, and combine the confidence and grayscale outlier of the defect position at different angles to form the density outlier of the defect area.

[0042] The DR radiographic digital image at each angle is used as input and the YOLOv8 neural network outputs the target box of the defect area, the center point coordinates, the category label, and the confidence level. The target box represents the defect location.

[0043] Die-casting molds are prone to defects such as pores, shrinkage, cold shut, and flow marks due to the mold fixed structure and internal flow channel design that limit the filling and exhaust efficiency of liquid metal.

[0044] However, considering that in subsequent forging, the main consideration is the situation where the local density caused by holes and cracks is lower than that of normal areas, and in DR images, the defective area has a shorter X-ray penetration path due to internal holes, resulting in a higher grayscale in the defective area than in the normal area. Therefore, the grayscale anomaly value of each defect position is calculated based on the grayscale difference in each defective area.

[0045] For DR digital radiography at each angle, the difference between the grayscale mean of each defect position and the minimum grayscale mean in all target frames is taken as the grayscale outlier value of each defect position.

[0046] Preferably, in this embodiment, the expression of the grayscale outlier value is:

[0047] g z,r It represents the grayscale mean of the rth defect position at the zth angle, min() represents the minimum function, represents the grayscale mean of all defect positions at the zth angle, l z,r Indicates the grayscale abnormal value of the rth defect position at the zth angle. The role of is to normalize and eliminate the dimension. The larger the grayscale mean value of the defect location, the larger the grayscale outlier value of the defect location, and the lower the local density of the defect location.

[0048] Based on the target frames in the DR images at different angles, the defect outline detected at each viewing angle is "back-projected" into a voxel column in space. The intersection of these back-projected volumes at multiple viewing angles is the potential defect area. In other words, the same defect area at multiple angles is mapped through voxels to obtain a 3D defect area. There is a one-to-one correspondence between the defect area and the defect location at each angle.

[0049] The density anomaly of the corresponding defect area is obtained based on the confidence and grayscale anomaly values ​​of the defect position at all angles. The density anomaly is positively correlated with the confidence and grayscale anomaly values ​​respectively.

[0050] It should be noted that positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large; the specific relationship is determined by actual application and this application does not impose any special restrictions.

[0051] Preferably, in this embodiment, the expression of the density abnormality value of the defect area is:

[0052] represents the confidence level of the rth defect position at the zth angle, l z,r Indicates the grayscale abnormal value of the rth defect position at the zth angle, L r represents the density outlier value of the r-th defect area.

[0053] At this point, the density anomaly value of each defect area is obtained.

[0054] Step S003: extract edge curves, obtain edge differences based on the curvature of the edge curves and the distances between discrete points, and construct volume intersections with the discrete points as cores to determine the sub-modulus differences between the discrete points.

[0055] In the production process, forging is performed directly after die-casting. The three-dimensional compressive stress during the forging process can effectively compact micro-defects such as shrinkage and pores inside the casting, thereby improving the density of the structure and reducing the risk of failure such as fracture and fatigue caused by voids and inclusions. In order to adapt to the efficiency and structural stability of mass production, traditional molds use a method of pre-setting flash grooves to improve the material fluidity and density of the forging process. However, since the workpieces after die-casting are not necessarily applied to the same part, and even the workpieces are different, this will result in uneven extrusion force at different positions of different workpieces during the subsequent forging process. Insufficient extrusion force in some areas will lead to excessively large grains and large differences in local density, affecting mechanical properties.

[0056] Since traditional molds are divided into upper and lower molds, the contact surface of the upper and lower molds is the parting surface of the workpiece, which is also the last closed interface of the metal material during the forging process. Traditional flash is usually located around the parting surface. It is the path for the material to overflow after filling the mold cavity, that is, the outlet area where the metal finally flows. It is used to form a flash groove to guide the metal flow and generate three-dimensional compressive stress.

[0057] During hammer forging, the dovetails of the upper and lower forging dies are fastened to the hammer head and lower die base using wedges. The upper and lower dies squeeze the die-cast preform, enhancing the workpiece's strength. Considering that different workpieces experience differences in local thickness and metal flow resistance around the parting surface during forging, the edge curve of the parting surface is divided into several discrete regions to achieve regionally adaptive forging pressure regulation.

[0058] A 3D model of the mold is obtained, and an edge curve of the parting surface is extracted from the 3D model. The edge curve is an edge line of the parting surface close to the inner side of the mold.

[0059] Discrete points are collected on the edge curve through "spatial equidistant sampling" so that the distance between adjacent discrete points is the same. In this embodiment, the distance between adjacent discrete points is 1 mm. It can be adjusted according to the size of different workpieces.

[0060] When two discrete points are close in distance and have similar curvature variations, they are spatially continuous and the edge contour is smooth and consistent within this region. Each discrete point is connected to the next discrete point to form a vector for that discrete point. The difference between the two discrete points' vectors can be used to determine the curvature variation. The edge dissimilarity between the two discrete points is determined based on the vector difference and distance between the two discrete points.

[0061] The edge difference is positively correlated with the vector difference and the distance respectively.

[0062] Preferably, in this embodiment, the expression of edge difference is:

[0063] d i,j represents the distance between the i-th discrete point and the j-th discrete point, D represents the maximum distance in the edge curve, which serves as normalization, v i Represents the vector from the i-th discrete point to the i+1-th discrete point, v j,j+1 Represents the vector from the jth discrete point to the j+1th discrete point, C i,j Indicates the edge difference between the i-th discrete point and the j-th discrete point.

[0064] When the distance between two discrete points is closer and the curvature changes are closer, the similarity between the two discrete points is greater, and the two discrete points are more likely to have similar parting modes.

[0065] If the internal volume characteristics of the mold cavity (local thickness of the workpiece) corresponding to the two discrete points are similar, it reflects that the stress state and filling behavior of the area during the forging process are similar. At this time, the two discrete points adopt the same flash groove thickness, and the corresponding pressure during extrusion is more uniform and the corresponding density is better.

[0066] Therefore, a sphere is constructed with each discrete point as the core, and the intersection volume between the sphere and the 3D model of the workpiece is calculated. The ratio of the intersection volume of the sphere and the 3D model corresponding to each discrete point to the volume of the sphere is recorded as the volume proportion of each discrete point.

[0067] The mold separation difference of the two discrete points is obtained based on the volume ratio difference of the two discrete points and the edge difference of the two discrete points.

[0068] The mold separation difference is positively correlated with the volume ratio difference and the edge difference, respectively.

[0069] Preferably, in this embodiment, the expression of the mold separation difference is:

[0070] D i,j =|ψ i -ψ j |×C i,j , C i,j Indicates the edge difference between the i-th discrete point and the j-th discrete point, ψ i represents the volume proportion of the i-th discrete point, ψ j represents the volume ratio difference of the jth discrete point, D i,j Indicates the difference in the split mode between the i-th discrete point and the j-th discrete point.

[0071] The smaller the difference in volume proportion, the more similar the volume features of the two discrete points are; the smaller the edge difference, the more similar the two discrete points have contour features, that is, the two discrete points have the same burr groove thickness.

[0072] The differences in the sub-modes between all discrete points are calculated, and all discrete points are divided into different clusters using a clustering algorithm. The clustering method used in this embodiment is a K-means clustering algorithm, where the K value is obtained by the traditional silhouette coefficient.

[0073] At this point, all discrete points in the edge curve are divided into different clusters.

[0074] Step S004: Based on the powder difference clustering, the flash correction coefficient is obtained according to the distance between the defect area and the cluster, as well as the volume and density abnormal values; and the gasket thickness is calculated.

[0075] The preform and the mold are matched to obtain the matching point of each discrete point in the preform.

[0076] Based on the density anomalies of each defective region, the distance between each defective region and the edge curve corresponding to all clusters after the above division is calculated on the 3D model during the subsequent forging process. The impact of the defective region on local density during the subsequent forging process is analyzed. The distance between the defective region and the edge curve is the minimum distance between the matching point in the defective region and the discrete points in the cluster.

[0077] When there are multiple defective areas close to the same cluster, there are more holes or gaps around the edge of the u-th cluster, which leads to low local density during the forging process. Therefore, the flash correction coefficient of each cluster is obtained based on the volume, density anomaly values ​​and distance of all defective areas from the cluster.

[0078] The flash correction coefficient is positively correlated with the volume and density anomaly of the defect area, and negatively correlated with the distance between the defect area and the cluster.

[0079] It should be noted that negative correlation means that when one variable increases, the other variable decreases accordingly, and the two variables change in opposite directions. When one variable changes from large to small or from small to large, the other variable also changes from small to large or from large to small. The specific relationship is determined by actual application and this application does not impose any special restrictions.

[0080] Preferably, the expression of the flash correction coefficient is:

[0081] L r Indicates the density anomaly of the rth defect area, V r represents the volume of the rth defect area, d u,r represents the distance to the rth defective region in the uth cluster, N represents the number of defective regions, ψ u Indicates the flash correction coefficient of the u-th cluster.

[0082] When ψ u When the value is large, there are many local voids or gaps in the die-cast preform around the u-th cluster. Therefore, based on the traditional die flash groove, it is considered to reduce the thickness of the flash bridge in the flash groove by adding a gasket, thereby reducing the speed of internal material overflow and improving the local density after forging. Figure 2 shown.

[0083] Among them, after the semi-finished product is heated, under the forging action of the forging press, the internal material passes through the flash groove bridge part. During the final forging, the flash area will produce strong three-dimensional compressive stress due to shape limitations to improve the plasticity and density of the metal. The thickness of the flash bridge in the traditional flash groove is 2 to 4 mm, but due to the abnormal density inside the preform, the flash bridge thickness of the initial mold is set to 4 mm in this embodiment, and then under different clusters, gaskets of different thicknesses are selected to reduce the flash bridge thickness and increase local compressive stress. The schematic diagram of the flash groove gasket is shown as follows Figure 3 As shown. Figure 3 In the figure, h1 and h2 represent the thickness of the flash bridge and the thickness of the flash bridge after adding the flash groove gasket, respectively.

[0084] By lowering the fixed column of the flash groove, the flash groove gasket can be embedded in the flash bin of the lower die to improve the stability during the forging process. During different forging processes, the edge correction coefficient can be analyzed through the inspection results of the preform, and different gaskets can be selected to adjust the flash bridge.

[0085] For discrete points of different clusters, the gasket thickness is calculated based on the flash correction coefficient. The expression of the gasket thickness is:

[0086] ψ u represents the flash correction coefficient of the u-th cluster, norm() represents the normalization function, ceil{} represents the rounding function, Δh u Indicates the thickness of the gasket corresponding to the discrete point in the u-th cluster.

[0087] The thickness of the gasket is between 0 and 2 mm. Based on the above steps, the gasket thickness of each discrete point in each cluster is obtained.

[0088] At this point, the gasket thickness corresponding to the discrete points in each cluster is obtained.

[0089] Step S005, forging the workpiece by adjusting the flash groove through the shim thickness.

[0090] The thickness of the gasket in the flash groove is obtained through the above steps, and gaskets of different thicknesses are selected based on different discrete points. That is, before forging, the thickness of the die flash bridge can be quickly adjusted without replacing the original die, and different thicknesses are used for different workpieces.

[0091] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

[0092] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A precision forging method for a new energy vehicle aluminum alloy forging production line, characterized in that: The method comprises the following steps: Obtain a preform after die-casting the aluminum alloy, and obtain digital radiographic images of the preform at different angles; Digital radiography is used as neural network input to obtain the target frame and confidence level. At each angle, the grayscale outlier of the target frame is obtained based on the grayscale difference between the target frame and the target frame with the minimum grayscale value. The intersection of the defect positions corresponding to different angles is projected into a voxel column in space to obtain the 3D defect area. Based on the confidence level and grayscale outlier of the defect position at all angles, the density outlier of the corresponding defect area is obtained. Extract the edge curve of the parting surface in the mold 3D model, divide the edge curve into multiple discrete points, and calculate the edge difference between the discrete points based on the vector difference and distance of each discrete point. Construct a sphere with each discrete point as the core, and obtain the parting difference between the discrete points based on the volume ratio difference and edge difference between the sphere and the 3D model. All discrete points are clustered into different clusters using the difference in mold separation as the cluster distance. The flash correction coefficient of each cluster is obtained based on the volume and density outliers of all defect areas and their distance from the cluster. The gasket thickness of different clusters is calculated based on the flash correction coefficient. Different gasket thicknesses are selected for different clusters to forge the workpiece.

2. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The digital radiographic images are acquired at ±30°, ±60°, and 90° of the preform respectively.

3. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The method for obtaining the grayscale outlier value of the target frame based on the grayscale difference between the target frame and the target frame with the minimum grayscale value is: g z,r It represents the grayscale mean of the rth defect position at the zth angle, min() represents the minimum function, represents the grayscale mean of all defect positions at the zth angle, l z,r represents the grayscale outlier value of the rth defect position at the zth angle; each target box is a defect position.

4. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The density anomaly of the defect area is positively correlated with the confidence level and grayscale anomaly value of the defect position.

5. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The edge difference is positively correlated with the vector difference and the distance between discrete points.

6. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 5, characterized in that: The vector difference is the difference between the vectors of two discrete points, and the vector of each discrete point is a vector formed by the discrete point pointing to the next discrete point.

7. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The mold separation difference is positively correlated with the volume share difference and the edge difference respectively; the volume share difference is the absolute value of the difference in volume share of two discrete points; the volume share of a discrete point is the ratio of the intersection volume of the sphere corresponding to each discrete point and the 3D model to the volume of the sphere.

8. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The flash correction coefficient is positively correlated with the volume and density anomaly of the defect area, and negatively correlated with the distance between the defect area and the cluster.

9. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 8, characterized in that: The distance between the defect area and the cluster is the minimum value of the distance between the matching point in the defect area and the discrete points in the cluster, and the matching point is the point corresponding to each discrete point after the mold and the preform are matched.

10. The precision forging method for a new energy vehicle aluminum alloy forging production line according to claim 1, characterized in that: The method for calculating the gasket thickness of different clusters based on the flash correction coefficient is: ψ u represents the flash correction coefficient of the u-th cluster, norm() represents the normalization function, ceil{} represents the rounding function, Δh u Indicates the thickness of the gasket corresponding to the discrete point in the u-th cluster.

Images