Multi-source online monitoring and error compensation additive manufacturing geometric tomography method

Abstract

The invention discloses a multi-source online monitoring and error compensation additive manufacturing geometric tomography method, geometric tomography is realized through a geometric tomography system, the system comprises a molten pool optical tomography device and a powder bed optical tomography online monitoring device, and the method comprises the following steps: S1, collecting a single frame of molten pool optical image frame by frame, a powder bed image after powder spreading is collected after powder spreading of the scraper; s2, perspective transformation correction is carried out on the whole-layer molten pool light image and the powder bed image after powder laying; s3, segmenting a deposition area in the corrected molten pool light image, and obtaining a binarization mask image of a surface layer warping area and a binarization mask image of a drum edge area; s4, obtaining a compensation deposition area mask image through compensation calculation; and S5, contour point cloud of each layer of compensation deposition area is obtained, and a three-dimensional point cloud model of the additive manufacturing structure morphology is generated. According to the invention, geometric feature tomography and three-dimensional visualization of a tiny feature structure and a geometric deviation structure in an online mode can be effectively realized.

Description

Technical Field

[0001] This invention belongs to the field of laser powder bed melting technology, specifically relating to a multi-source online monitoring and error compensation additive manufacturing geometric tomography method. Background Technology

[0002] Laser powder bed fusion (L-PBF) is a widely used additive manufacturing technology that can directly form metal parts with complex structures such as lattice and topology. However, the extreme manufacturing conditions of L-PBF often lead to geometric deviations in the structural morphology. For example, thin rods in lattice structures are prone to geometric errors, and overhanging structures are prone to geometric defects. Currently, the three-dimensional morphology of the formed structure is mainly characterized by offline inspection methods after printing, such as 3D scanners and X-ray computed tomography. This offline inspection mode has obvious limitations: even if geometric defects occur during manufacturing, defect diagnosis can only be performed after the entire structure is printed. If geometric defects or accuracy errors are diagnosed, manufacturing time and costs are lost. Therefore, developing online monitoring and three-dimensional morphology characterization methods for formed structures has significant engineering application value for quality assessment and process optimization in the manufacturing process.

[0003] Current online monitoring of structural morphology primarily utilizes powder bed tomography, capturing visible light images of the powder bed after laser melting, segmenting the contours of the deposition areas layer by layer, and using this information for 3D reconstruction of the formed structural morphology. However, this process faces several challenges. First, segmenting the deposition areas of small-sized structures is extremely difficult. For example, with lattice rods less than 1 mm in diameter, it is difficult to distinguish the deposition areas from the powder areas in visible light images, leading to a lack of structural information in the final 3D reconstructed model. Furthermore, 3D reconstruction based on the segmented deposition areas from the post-melting powder bed images results in a 3D reconstructed model that approximates the design model but lacks features such as geometric defects and geometric deviations. Summary of the Invention

[0004] The purpose of this invention is to provide a multi-source online monitoring and error compensation additive manufacturing geometric tomography method to achieve online monitoring and high-precision three-dimensional reconstruction of normal deposition structures, small feature structures, and geometrically problematic structures.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] A multi-source online monitoring and error compensation additive manufacturing geometric tomography method is disclosed. The method achieves geometric tomography through a geometric tomography system. The geometric tomography system includes a molten pool optical tomography device and a powder bed optical tomography online monitoring device, both located outside the forming chamber of an L-PBF additive manufacturing platform. The molten pool optical tomography device includes a first industrial camera and a filter, with the filter installed at the front end of the lens of the first industrial camera. The powder bed optical tomography online monitoring device includes a second industrial camera.

[0007] The method includes the following steps:

[0008] S1: During the single-layer melting process, a single frame of molten pool optical tomography image is acquired frame by frame using a molten pool optical tomography device. After the powder is spread by the scraper, a powder bed optical tomography image is acquired using a powder bed optical tomography device.

[0009] S2: The single-frame molten pool light images acquired frame by frame are stitched together into a whole-layer molten pool light image, and perspective transformation correction is performed on the whole-layer molten pool light image and the powder bed image after powder spreading.

[0010] S3: Use the threshold segmentation method to segment the deposition area in the corrected molten pool light image, and use the trained image segmentation neural network model to segment the surface warping area and the bulging edge area in the powder bed image after powder spreading to obtain the binarized mask image of the surface warping area and the binarized mask image of the bulging edge area.

[0011] S4: Using the binarized mask image of the surface warped region and the binarized mask image of the bulging edge region, compensation calculation is performed on the filtered binarized mask image of the deposition region to obtain the compensated deposition region mask image.

[0012] S5: Extract the edges of the mask image of each layer of compensation deposition area, perform point cloud processing on the edge pixels to obtain the contour point cloud of each layer of compensation deposition area, and stitch the contour point clouds of the compensation deposition area layer by layer along the manufacturing direction to generate a three-dimensional point cloud model of the additive manufacturing structure morphology.

[0013] Furthermore, step S1 specifically includes:

[0014] S11: Set the molten pool optical tomography device to capture one single-frame molten pool optical image every n seconds, and the exposure time for each single-frame molten pool optical image is n seconds, where n > 0;

[0015] S12: When the current printing layer starts printing, the molten pool optical tomography device starts taking pictures, and when the current printing layer finishes printing, the molten pool optical tomography device stops taking pictures. During this period, the molten pool optical tomography device continuously acquires multiple single-frame molten pool optical images during the single-layer melting process.

[0016] S13: After the current printing layer is finished, the squeegee spreads the powder for the next printing layer on top of the part. After the powder spreading is completed, a powder bed tomography image is taken using a powder bed tomography device.

[0017] Furthermore, the characteristic is that n is 2.

[0018] Furthermore, step S2 specifically includes:

[0019] S21: Use the time integration method to stitch together all the single-frame molten pool light images acquired during the single-layer melting process into a single single-layer molten pool light image;

[0020] S22: The perspective transformation method is used to correct the light image of the entire molten pool and the powder bed image after powder spreading. After correction, the image border is completely superimposed on the substrate edge.

[0021] Step S21 specifically includes:

[0022] S211: Use the OpenCV open-source computer vision library to convert all single-frame molten pool light images into grayscale images, with the brightness value of each pixel ranging from 0 to 255;

[0023] S212: Create a blank image with the same size as the single-frame molten pool light image. Calculate the brightness value of each pixel in the blank image. The calculation method is: the brightness value of the pixel is equal to the sum of the brightness values ​​of the pixel in all single-frame molten pool light images collected during the single-layer melting process. If the calculated brightness value exceeds 255, set it to 255. After calculating the brightness value of each pixel, a full-layer molten pool light image is finally obtained.

[0024] Furthermore, step S3 specifically includes:

[0025] S31: Use the threshold segmentation method to segment the deposition area in the corrected molten pool light image. Set the brightness value of the pixels with brightness values ​​greater than or equal to the threshold to 255, and set the brightness value of the pixels with brightness values ​​less than the threshold to 0, thereby obtaining the binarized mask image of the deposition area.

[0026] S32: The erosion and dilation operations in the morphological processing method are used to filter the noise in the binary mask image of the deposition area. After filtering, the filtered binary mask image of the deposition area is obtained.

[0027] S33: Using a trained image segmentation neural network model, segment the surface warping region and the bulging edge region in the powder bed image after powder spreading correction, and obtain the binarized mask image of the surface warping region and the binarized mask image of the bulging edge region.

[0028] Furthermore, the threshold in step S31 is set to 30.

[0029] Furthermore, the size of the structuring element in the filtering process in step S31 is set to 3×3.

[0030] Furthermore, the training method for the image segmentation neural network model in step S3 is as follows: based on Labelme data annotation software, the surface warping and bulging edge regions in the powder bed image after correction are segmented out, and binary mask image labels of the surface warping and bulging edge regions are generated to create an image segmentation neural network model training dataset; the image segmentation neural network model is trained using the image segmentation neural network model training dataset.

[0031] Furthermore, step S4 specifically includes:

[0032] S41: Perform a Boolean subtraction operation between the filtered deposition area mask image and the binarized mask image of the surface warped area to obtain the deposition area mask image with the surface warped area removed;

[0033] S42: Perform a Boolean addition operation on the mask image of the deposition area with the surface warped region removed and the binarized mask image of the bulging edge region to obtain the mask image of the compensated deposition area.

[0034] Furthermore, step S5 specifically includes:

[0035] S51: Use a contour detection algorithm to obtain the contour of the deposition region in the mask image of each layer of compensated deposition region;

[0036] S52: Use the OpenCV open-source computer vision library to obtain the X and Y coordinates of each pixel on the contour of the deposition region, and generate a point cloud based on the position information of the pixels.

[0037] S53: Add Z coordinate information to each point cloud on the contour of the deposition area. The Z axis is the printing direction, and calculate the Z coordinate value = the number of layers of the powder bed image after powder spreading × the powder spreading thickness set in the printing process.

[0038] S54: Calculate the XYZ coordinate data of the point cloud of the contour of all compensated deposition regions from the first printing layer to the last printing layer, and generate a three-dimensional point cloud model of the additive manufacturing structure morphology.

[0039] Compared with the prior art, the technical effects achieved by the present invention are as follows:

[0040] (1) This invention designs a three-dimensional reconstruction system for multi-source online monitoring and error compensation of additive manufacturing structure morphology, which can reconstruct the three-dimensional morphology model of the formed structure online. Compared with the traditional online three-dimensional reconstruction technology of formed structure based on single-source monitoring method of powder bed optical tomography, this invention adds the function of online monitoring of molten pool optical tomography. By using dual-source online monitoring signals, small-sized feature structures less than 1 mm can be effectively segmented and reconstructed, solving the problem of loss of small-sized feature structures in traditional reconstruction methods. (2) This invention proposes an error compensation strategy for parts with geometric deviations and geometric defects. It can perform error compensation on the deposition area based on online monitoring data and perform three-dimensional reconstruction based on the compensated deposition area. The reconstructed model can effectively characterize the morphology, size and distribution of geometric problems such as geometric deviations and geometric defects. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the multi-source online monitoring and error compensation additive manufacturing geometric tomography method of the present invention.

[0042] Figure 2 This is a schematic diagram of the online monitoring system for molten pool optical tomography and powder bed optical tomography.

[0043] Figure 3 This is a schematic diagram of the image acquisition and image stitching method for molten pool optical tomography.

[0044] Figure 4 This diagram illustrates a method for segmenting feature regions in multi-source online monitoring images.

[0045] Figure 5 This diagram illustrates the error compensation method for the sedimentation area.

[0046] Figure 6 The flowchart describes the process of compensating for point cloudification in the mask image of the deposition area and stitching it layer by layer into a 3D reconstruction model. Detailed Implementation

[0047] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0048] A method for multi-source online monitoring and error compensation of three-dimensional reconstruction of additive manufacturing structural morphology is disclosed. The method utilizes a geometric tomography system to achieve geometric tomography. Outside the forming chamber of the L-PBF additive manufacturing platform, a molten pool optical tomography device and a powder bed optical tomography online monitoring device are mounted off-axis. Figure 2 This is a schematic diagram of the molten pool optical tomography and powder bed optical geometric tomography system constructed according to the present invention;

[0049] The molten pool optical tomography device consists of a visible light industrial camera and a neutral density filter with a transmittance of 1%. The filter is installed at the front end of the visible light industrial camera lens, and the powder bed optical tomography online monitoring device is an industrial camera.

[0050] Combination Figure 1 The method includes the following steps:

[0051] S1: During the single-layer melting process, a single frame of molten pool optical tomography image is acquired frame by frame using a molten pool optical tomography device. After the powder is spread by the scraper, a powder bed optical tomography image is acquired using a powder bed optical tomography device.

[0052] S2: The single-frame molten pool light images acquired frame by frame are stitched together into a whole-layer molten pool light image, and perspective transformation correction is performed on the whole-layer molten pool light image and the powder bed image after powder spreading.

[0053] S3: Use a threshold segmentation method to segment the deposition area in the corrected molten pool image. Use a trained image segmentation neural network model to segment the surface warping region and the bulging edge region in the powder bed image after powder spreading. Obtain the binarized mask images of the surface warping region and the bulging edge region. Figure 4 This is a schematic diagram illustrating the feature region segmentation effect of multi-source online monitoring images;

[0054] S4: Using the binarized mask images of the surface warped region and the bulging edge region, compensation calculations are performed on the filtered binarized mask image of the deposition region to obtain the compensated deposition region mask image. Figure 5 A schematic diagram of the error compensation method for the sedimentary region;

[0055] S5: Extract the edges of the mask image of each layer of compensation deposition area, perform point cloud processing on the edge pixels to obtain the contour point cloud of each layer of compensation deposition area, and stitch the contour point clouds of the compensation deposition areas layer by layer along the manufacturing direction to generate a three-dimensional point cloud model of the additive manufacturing structure morphology. Figure 6 The flowchart describes the process of compensating for point cloudification in the mask image of the deposition area and stitching it layer by layer into a 3D reconstruction model.

[0056] Furthermore, step S1 specifically includes:

[0057] S11: Set the molten pool optical tomography device to capture a single frame of molten pool optical image every 2 seconds, with an exposure time of 2 seconds for each single frame of molten pool optical image;

[0058] S12: When the current printing layer starts printing, the molten pool optical tomography device starts taking pictures, and when the current printing layer finishes printing, the molten pool optical tomography device stops taking pictures. During this period, the molten pool optical tomography device continuously acquires multiple single-frame molten pool optical images during the single-layer melting process.

[0059] S13: After the current printing layer is finished, the squeegee spreads the powder for the next printing layer on top of the part. After the powder spreading is completed, a powder bed tomography image is taken using a powder bed tomography device.

[0060] Furthermore, step S2 specifically includes:

[0061] S21: Use the time integration method to stitch together all the single-frame molten pool light images acquired during the single-layer melting process into a single single-layer molten pool light image;

[0062] S22: The perspective transformation method is used to correct the light image of the entire molten pool and the powder bed image after powder spreading. After correction, the image border is completely superimposed on the substrate edge.

[0063] Step S21 specifically includes:

[0064] S211: Use the OpenCV open-source computer vision library to convert all single-frame molten pool light images into grayscale images, with the brightness value of each pixel ranging from 0 to 255;

[0065] S212: Create a blank image with the same size as the single-frame molten pool light image. Calculate the brightness value of each pixel in the blank image. The calculation method is: the brightness value of this pixel is equal to the sum of the brightness values ​​of this pixel in all single-frame molten pool light images acquired during the single-layer melting process. If the calculated brightness value exceeds 255, set it to 255. After calculating the brightness value of each pixel, a complete molten pool light image is finally obtained. Figure 3 This is a schematic diagram of the image acquisition and image stitching method for molten pool optical tomography.

[0066] Furthermore, step S3 specifically includes:

[0067] S31: Use a threshold segmentation method to segment the deposition area in the corrected molten pool light image. The threshold is set to 30, the brightness value of pixels with a brightness value greater than or equal to 30 is set to 255, and the brightness value of pixels with a brightness value less than 30 is set to 0, thereby obtaining a binarized mask image of the deposition area.

[0068] S32: The erosion and dilation operations in the morphological processing method are used to filter the noise in the binary mask image of the deposition area. The size of the structuring element in the filtering process is set to 3×3. After filtering, the filtered binary mask image of the deposition area is obtained.

[0069] S33: Using a trained image segmentation neural network model, segment the surface warping region and the bulging edge region in the powder bed image after powder spreading correction, and obtain the binarized mask image of the surface warping region and the binarized mask image of the bulging edge region.

[0070] Furthermore, the training method for the image segmentation neural network model in step S3 is as follows: Based on Labelme data annotation software, the surface warping and bulging edge regions in the powder bed image after correction are segmented out, and binary mask image labels for the surface warping and bulging edge regions are generated to create an image segmentation neural network model training dataset; the image segmentation neural network model is trained using the image segmentation neural network model (e.g., U-Net image segmentation neural network model, Mask R-CNN image segmentation neural network model) training dataset.

[0071] Furthermore, step S4 specifically includes:

[0072] S41: Perform a Boolean subtraction operation between the filtered deposition area mask image and the binarized mask image of the surface warped area to obtain the deposition area mask image with the surface warped area removed;

[0073] S42: Perform a Boolean addition operation on the mask image of the deposition area with the surface warped region removed and the binarized mask image of the bulging edge region to obtain the mask image of the compensated deposition area.

[0074] Furthermore, step S5 specifically includes:

[0075] S51: Use a contour detection algorithm to obtain the contour of the deposition region in the mask image of each layer of compensated deposition region;

[0076] S52: Use the OpenCV open-source computer vision library to obtain the X and Y coordinates of each pixel on the contour of the deposition region, and generate a point cloud based on the position information of the pixels.

[0077] S53: Add Z coordinate information to each point cloud on the contour of the deposition area. The Z axis is the printing direction, and calculate the Z coordinate value = the number of layers of the powder bed image after powder spreading × the powder spreading thickness set in the printing process.

[0078] S54: Calculate the XYZ coordinate data of the point cloud of the contour of all compensated deposition regions from the first printing layer to the last printing layer, and generate a three-dimensional point cloud model of the additive manufacturing structure morphology.

[0079] This invention presents a multi-source online monitoring and error compensation method for 3D reconstruction of additive manufacturing structural morphology, capable of reconstructing a 3D morphology model of the formed structure online. Compared to traditional online 3D reconstruction technology based on single-source monitoring using powder bed optical tomography, this invention adds a molten pool optical tomography online monitoring module. Utilizing dual-source online monitoring signals, it can effectively segment and reconstruct small-sized feature structures below 1 mm, achieving a data integrity of 99.1% for a 0.3 mm rod diameter lattice structure, thus solving the problem of lost small-sized feature structures in traditional reconstruction methods. For parts exhibiting geometric deviations or defects, this invention proposes an error compensation strategy. It can perform error compensation calculations on the deposition area online and perform 3D reconstruction based on the compensated deposition area. The reconstructed model can effectively characterize the morphology, size, and distribution of geometric problems such as geometric deviations and defects. For a lattice rod with a design size of 1.5 mm and an actual formed size of 2.09 mm exhibiting geometric deviations, the 3D reconstruction accuracy of this invention can reach 96.7%. For a suspended structure with geometric defects that has a design size of 3.6 mm and an actual formed size of 2.9 mm, the three-dimensional reconstruction accuracy of the present invention can reach 94.6%.

[0080] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A multi-source online monitoring and error compensation additive manufacturing geometric tomography method, characterized in that, The method realizes geometric tomography through a geometric tomography system, the geometric tomography system comprises a molten pool optical tomography device and a powder bed optical tomography online monitoring device which are arranged outside a forming cabin of an L-PBF additive manufacturing platform, the molten pool optical tomography device comprises a first industrial camera and a filter, the filter is mounted at the front end of the lens of the first industrial camera, and the powder bed optical tomography online monitoring device comprises a second industrial camera; The method comprises the following steps: S1: using the molten pool optical tomography device to collect single-frame molten pool optical images frame by frame during the single-layer melting process, and using the powder bed optical tomography device to collect a powder bed image after powder spreading after the scraper spreads powder; S2: splicing the single-frame molten pool optical images collected frame by frame into a whole-layer molten pool optical image, and correcting the whole-layer molten pool optical image and the powder bed image after powder spreading through perspective transformation; S3: using a threshold segmentation method to segment the deposited area in the corrected molten pool optical image, using a trained image segmentation neural network model to segment the surface warping area and the drum edge area in the powder bed image after powder spreading to obtain a surface warping area binary mask image and a drum edge area binary mask image; S4: using the surface warping area binary mask image and the drum edge area binary mask image to compensate the binary mask image of the filtered deposited area to obtain a compensated deposited area mask image; S5: extracting the edge of each layer of the compensated deposited area mask image, performing point cloud processing on the edge pixels to obtain the contour point cloud of each layer of the compensated deposited area, and splicing the contour point clouds of the compensated deposited areas layer by layer along the manufacturing direction to generate an additive manufacturing structure topography three-dimensional point cloud model.

2. The multi-source online monitoring and error-compensated additive manufacturing geometric tomography method according to claim 1, wherein, Step S1 specifically comprises: S11: setting the molten pool optical tomography device to take a single-frame molten pool optical image every n seconds, and the exposure time of each single-frame molten pool optical image is n seconds, n>0; S12: starting to take images by the molten pool optical tomography device when the current printing layer starts printing, and stopping to take images by the molten pool optical tomography device when the current printing layer ends printing, during which the molten pool optical tomography device continuously collects to obtain multiple single-frame molten pool optical images during the single-layer melting process; S13: after the current printing layer ends printing, the scraper spreads the powder of the next printing layer above the part, and after the powder spreading ends, a powder bed image after powder spreading is taken by using the powder bed optical tomography device.

3. The multi-source online monitoring and error-compensated additive manufacturing geometric tomography method according to claim 2, wherein, n is 2.

4. The multi-source online monitoring and error-compensated additive manufacturing geometric tomography method according to claim 2 or 3, characterized in that, Step S2 specifically comprises: S21: using a time integration method to splice all single-frame molten pool optical images collected during the single-layer melting process into a single-layer molten pool optical image; S22: correcting the whole-layer molten pool optical image and the powder bed image after powder spreading through perspective transformation, so that the image frame and the substrate edge are completely coincident after correction; Step S21 specifically comprises: S211: using the OpenCV open computer vision library to convert all single-frame molten pool optical images into grayscale images, and the brightness value of each pixel point ranges from 0 to 255; S212: Create a blank image with the same size as the single-frame molten pool light image, calculate the brightness value of each pixel point in the blank image, and the calculation method is: the brightness value of the pixel point is equal to the sum of the brightness values of the pixel point in all single-frame molten pool light images collected during the single-layer melting process. If the calculated brightness value exceeds 255, it is set to 255. After calculating the brightness value of each pixel point, a whole-layer molten pool light image is obtained.

5. The multi-source online monitoring and error-compensated geometric tomography method of claim 4, wherein, Step S3 specifically includes: S31: using a threshold segmentation method to segment the deposition area in the corrected molten pool light image, setting the brightness value of the pixel point with a brightness value greater than or equal to the threshold value to 255, and setting the brightness value of the pixel point with a brightness value less than the threshold value to 0, thereby obtaining a binary mask image of the deposition area; S32: using the erosion and dilation operations in the morphological processing method to filter the noise of the binary mask image of the deposition area, and obtaining a filtered binary mask image of the deposition area after filtering; S33: using the trained image segmentation neural network model to segment the surface warping area and the drum edge area in the powder bed image after powder correction, and obtaining a binary mask image of the surface warping area and a binary mask image of the drum edge area.

6. The multi-source online monitoring and error-compensated geometric tomography method according to claim 5, wherein, The threshold value in step S31 is set to 30.

7. The multi-source online monitoring and error-compensated geometric tomography method of claim 5, wherein, The size of the structure element in the filtering process in step S31 is set to 3x3.

8. The multi-source online monitoring and error-compensated geometric tomography method of claim 5, wherein, The training method of the image segmentation neural network model in step S3 is: based on the Labelme data labeling software, the surface warping and drum edge areas in the powder bed image after powder correction are segmented to generate binary mask image labels of the surface warping and drum edge areas, and the image segmentation neural network model training data set is made; the image segmentation neural network model is trained using the image segmentation neural network model training data set.

9. The multi-source online monitoring and error-compensated additive manufacturing geometric tomography method according to any one of claims 5-8, characterized in that, Step S4 specifically includes: S41: performing a Boolean subtraction operation on the filtered deposition area mask image and the surface warping area binary mask image to obtain a deposition area mask image without the surface warping area; S42: performing a Boolean addition operation on the deposition area mask image without the surface warping area and the drum edge area binary mask image to obtain a compensation deposition area mask image.

10. The multi-source online monitoring and error-compensated additive manufacturing geometric tomography method of claim 9, wherein, Step S5 specifically includes: S51: using a contour detection algorithm to obtain the deposition area contour in each layer of the compensation deposition area mask image; S52: using the OpenCV open source computer vision library to obtain the X and Y coordinates of each pixel point on the deposition area contour, and generating a point cloud according to the position information of the pixel points; S53: adding Z coordinate information to each point cloud on the deposition area contour, the Z axis being the printing direction, and calculating the Z coordinate value = the layer number of the current powder bed image after powdering x the powdering thickness set in the printing process; S54: calculating the XYZ coordinate data of all compensation deposition area contour point clouds from the first printing layer to the last printing layer to generate an additive manufacturing structure morphology three-dimensional point cloud model.

Images