Real-time molten pool monitoring and defect early warning method and system for additive manufacturing

Abstract

The invention provides a real-time molten pool monitoring and defect early warning method and system for additive manufacturing. The method comprises the following steps that real-time molten pool strength data of the full breadth of a printing layer is obtained; the real-time molten pool intensity data are preprocessed, the processed molten pool intensity data are converted into a two-dimensional molten pool intensity cloud picture, the two-dimensional molten pool intensity cloud picture of the printing layer is visually displayed, and a molten pool intensity array picture is generated; according to the layer-by-layer molten pool intensity cloud picture data, a three-dimensional molten pool intensity distribution model is constructed based on a three-dimensional reconstruction algorithm, and a three-dimensional molten pool intensity cloud picture is obtained; carrying out defect identification on the two-dimensional molten pool intensity cloud picture and the three-dimensional molten pool intensity cloud picture respectively, extracting defect areas, marking and sending out early warning; through real-time collection of the full-breadth molten pool strength in the printing process, the molten pool strength cloud picture of each layer is generated, the three-dimensional molten pool strength cloud picture is constructed, dynamic monitoring and defect early warning of the state of the molten pool are achieved, and the quality control capacity in the additive manufacturing process is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of laser powder bed metal additive manufacturing, and particularly to a real-time molten pool monitoring and defect warning method and system for additive manufacturing. Background Art

[0002] Laser powder bed metal additive manufacturing is an advanced technology for manufacturing parts by melting metal powder layer by layer, which has significant advantages in processing complex components, saving materials, and shortening the manufacturing cycle. This technology has been widely applied in the fields of aerospace, medical devices, and high-end industrial manufacturing; however, the dynamic changes of the molten pool and defect problems in the metal additive manufacturing process are the main technical difficulties restricting its further application. In the metal additive manufacturing process, the strength of the molten pool directly affects the final quality and performance of the printed parts. Insufficient molten pool strength may lead to incomplete melting, forming defects such as pores or cracks; excessive molten pool strength may cause overburning or splashing of the material, resulting in shape deviation or performance degradation of the parts; therefore, accurate monitoring and control of the molten pool strength become the key to improving the quality stability and consistency of metal additive manufacturing.

[0003] A multi-angle visual sensing device for the molten pool morphology of metal additive manufacturing based on a single camera with the publication number of CN113843420A, the sensing device includes three parts: a sensing path, a composite filter system, and an image detection system. The first plane mirror receives the light radiated from the molten pool from above and reflects it to the composite filter system, and the width information of the molten pool is obtained after passing through the image detection system; the light radiated from the molten pool and the arc light reflected by the previous deposition layer are reflected by the third plane mirror to reach the second plane mirror, and then pass through the composite filter system, and finally the height information of the molten pool and the height information of the previous deposition layer are obtained through the image detection system.

[0004] The current technologies have certain limitations in terms of spatial resolution, real-time performance, data integration ability, etc., and it is difficult to meet the real-time monitoring requirements of the molten pool strength distribution of the entire printing area; at the same time, the lack of the ability to reconstruct the three-dimensional visualization of the molten pool strength distribution makes it difficult to provide effective support for quality control and defect identification, reducing the quality control ability in the additive manufacturing process. Summary of the Invention

[0005] In view of this, the present invention proposes a real-time molten pool monitoring and defect warning method and system for additive manufacturing. By collecting and analyzing the real-time molten pool strength of the entire printing area during the printing process, a molten pool strength cloud map of each layer is generated, and a three-dimensional molten pool strength cloud map is constructed, so as to realize the dynamic monitoring of the molten pool state and defect warning, and effectively improve the quality control ability in the additive manufacturing process.

[0006] The technical solution of the present invention is realized as follows: In the first aspect, the present invention provides a real-time molten pool monitoring and defect warning method for additive manufacturing, including the following steps:

[0007] S1. Obtain the real-time molten pool intensity data of the full print layer;

[0008] S2. Preprocess the real-time molten pool intensity data, convert the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually display the two-dimensional molten pool intensity cloud map of the print layer to generate a molten pool intensity array map;

[0009] S3. According to the layer-by-layer molten pool intensity cloud map data, construct a three-dimensional molten pool intensity distribution model based on a three-dimensional reconstruction algorithm to obtain a three-dimensional molten pool intensity cloud map;

[0010] S4. Perform defect identification on the two-dimensional molten pool intensity cloud map and the three-dimensional molten pool intensity cloud map respectively, extract the defect area and issue a warning.

[0011] Based on the above technical solution, preferably, obtaining the real-time molten pool intensity data of the full print layer in step S1 includes: using an optical sensor as the data acquisition device, and the optical sensor is installed at a paraxial position of the laser scanning path. The optical sensor moves along with the laser scanning path to collect the molten pool intensity data of the entire print area, and obtain the full-area intensity distribution of the molten pool during the printing process in real time. The optical sensor is set at an inclination angle of 15-20° with the horizontal plane.

[0012] Based on the above technical solution, preferably, the data acquisition device further includes a narrowband pass filter, and the spectral range of the narrowband pass filter is 880nm - 930nm, so that the optical sensor receives the corresponding band optical signal radiated by the molten pool and reduces noise interference.

[0013] Based on the above technical solution, preferably, the preprocessing of the real-time molten pool intensity data in step S2, converting the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually displaying the two-dimensional molten pool intensity cloud map of the print layer to generate a molten pool intensity array map includes the following sub-steps:

[0014] S21. Clean, denoise, and correct the collected real-time molten pool intensity data, and perform smoothing processing on the collected full-area molten pool intensity distribution data using an image processing algorithm to obtain the standard molten pool intensity data;

[0015] S22. Convert the standard molten pool intensity data into a grayscale image to obtain the two-dimensional matrix map of the molten pool intensity of each layer;

[0016] S23. Traverse the elements in the two-dimensional matrix diagram of the molten pool intensity in sequence using color mapping, match the corresponding colors according to the intensity values of each pixel, and fill them into the corresponding grayscale image to obtain a two-dimensional molten pool intensity cloud map.

[0017] S24. Display the two-dimensional molten pool intensity cloud map of each layer on the interface to show the distribution of the molten pool intensity on the printing surface.

[0018] Based on the above technical solutions, preferably, in step S3, according to the data of the molten pool intensity cloud map layer by layer, and based on a three-dimensional reconstruction algorithm, a three-dimensional molten pool intensity distribution model is constructed, which includes the following sub-steps:

[0019] Perform standardization, denoising, and smoothing processing on the two-dimensional molten pool intensity cloud map collected layer by layer, and use bilateral filtering to remove noise.

[0020] Use a three-dimensional interpolation method to fill the data missing areas in the denoised two-dimensional molten pool intensity cloud map to obtain a standard map of the molten pool intensity array.

[0021] According to the standard map of the molten pool intensity array layer by layer, and based on a three-dimensional reconstruction algorithm, a three-dimensional molten pool intensity distribution model is constructed to obtain a three-dimensional molten pool intensity cloud map. The three-dimensional molten pool intensity distribution expression is:

[0022]

[0023] In the formula, I(x, y, z) is the molten pool intensity value at the position (x, y, z) in the three-dimensional molten pool intensity cloud map, (x, y, z) is a three-dimensional coordinate system, x and y are the two-dimensional plane coordinates of the printing surface, z is the number of printing layers, I k (x, y) is the molten pool intensity data of each layer, N is the total number of printing layers, d is the layer thickness, and δ(z - k·d) is the mapping of the molten pool intensity data of the kth layer in the z direction.

[0024] Combine the molten pool intensity array maps into a continuous three-dimensional model, and enhance the visual effect through pseudo-color mapping.

[0025] Based on the above technical solutions, preferably, step S4 includes defect detection of the two-dimensional molten pool intensity cloud map, which includes the following steps:

[0026] Calculate the average intensity μ and standard deviation σ of the two-dimensional molten pool intensity cloud map, and set the abnormal area threshold according to the average intensity μ and standard deviation σ.

[0027] Compare the abnormal area threshold with the intensity of the two-dimensional molten pool intensity cloud map. If it is not within the range of the abnormal area threshold, highlight the defect area on the two-dimensional molten pool intensity cloud map and send out a warning message. The expression is:

[0028] Tlow ≤I k (x, y) ≤ T high ;

[0029] Wherein, T low is the lower limit of the abnormal area threshold range, T low = μ - 3σ; I k (x, y) is the molten pool strength data of each layer, T high is the upper limit of the abnormal area threshold range, T high = μ + 3σ.

[0030] On the basis of the above technical solutions, preferably, step S4 includes defect detection of the three-dimensional molten pool strength cloud map, including the following steps:

[0031] S41. Obtain the three-dimensional molten pool strength cloud map data, calculate the spatial gradient of the molten pool strength by using the numerical differentiation method, and perform edge detection by using the Laplace operator to obtain the molten pool edge image. The expression is:

[0032]

[0033] Wherein, is the result of the molten pool strength value at the position (x, y, z) in the three-dimensional molten pool strength cloud map after being acted on by the Laplace operator, is the change rate of the three-dimensional molten pool strength distribution model in the x direction, is the change rate of the three-dimensional molten pool strength distribution model in the y direction; is the change rate of the three-dimensional molten pool strength distribution model in the z direction;

[0034] S42. Extract the abnormal area from the molten pool edge image by using three-dimensional morphological filtering;

[0035] S43. Perform voxel analysis on the extracted abnormal area, and calculate its volume and density. The expression is:

[0036]

[0037] Wherein, v is the volume of the abnormal area, R is the abnormal area, δI(x, y, z) is the volume of a single voxel in the abnormal area; ρ is the density of the abnormal area, m is the total mass of the abnormal area;

[0038] S44. Judge the defect according to the volume and density of the abnormal area. If it is a defect area, highlight it on the three-dimensional molten pool strength cloud map and send out a warning message.

[0039] Based on the above technical solutions, preferably, in step S44, the defect is judged according to the volume and density of the abnormal area. If it is a defect area, it is highlighted on the three-dimensional molten pool intensity cloud map and a warning message is sent, including:

[0040] A preset low density threshold, a high density threshold, and a volume threshold are set. The volume and density of the abnormal area are compared with the density threshold and the volume threshold. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is less than the low density threshold, it is an incomplete fusion defect, and it is highlighted in blue on the three-dimensional molten pool intensity cloud map and a warning message is sent. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is greater than the high density threshold, it is an overburn area, and it is highlighted in red on the three-dimensional molten pool intensity cloud map and a warning message is sent. Otherwise, it is in a normal state.

[0041] In a second aspect, the present invention also provides a real-time molten pool monitoring and defect warning system for additive manufacturing, which is implemented by using a real-time molten pool monitoring and defect warning method for additive manufacturing as described above. The system includes:

[0042] An acquisition module for acquiring real-time molten pool intensity data of the full width of the printing layer;

[0043] A two-dimensional cloud map construction module for preprocessing the real-time molten pool intensity data, converting the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually displaying the two-dimensional molten pool intensity cloud map of the printing layer to generate a molten pool intensity array map;

[0044] A three-dimensional cloud map construction module for constructing a three-dimensional molten pool intensity distribution model based on the layer-by-layer molten pool intensity cloud map data and obtaining a three-dimensional molten pool intensity cloud map based on a three-dimensional reconstruction algorithm;

[0045] A defect recognition module for respectively performing defect recognition on the two-dimensional molten pool intensity cloud map and the three-dimensional molten pool intensity cloud map, extracting the defect area and marking and sending a warning.

[0046] In a third aspect, the present invention also provides a computer-readable storage medium, on which a program for a real-time molten pool monitoring and defect warning method for additive manufacturing is stored. When the program for a real-time molten pool monitoring and defect warning method for additive manufacturing is executed, it implements a real-time molten pool monitoring and defect warning method for additive manufacturing.

[0047] The real-time molten pool monitoring and defect warning method and system for additive manufacturing of the present invention have the following beneficial effects compared with the prior art:

[0048] (1) By collecting and analyzing the full - scale molten pool intensity in real - time during the printing process, a molten pool intensity cloud map for each layer is generated, and a three - dimensional molten pool intensity cloud map is constructed. The three - dimensional cloud map visually shows the overall distribution of the molten pool intensity during the printing process, thereby realizing the dynamic monitoring of the molten pool state and defect warning, and effectively improving the quality control ability in the additive manufacturing process;

[0049] (2) By installing an optical sensor at the paraxial position of the laser scanning path, the real - time collection of the full - scale molten pool intensity during the printing process is achieved, and the collection optical path is optimized to improve the signal quality and eliminate ambient light interference, making the actual intensity data of the molten pool clearer;

[0050] (3) Through the set defect recognition algorithm and warning mechanism, potential defects such as insufficient or excessive molten pool intensity can be quickly detected, and a warning is sent to the operator in real - time; the operator can adjust the process parameters in a timely manner according to the warning information, thereby reducing the defect rate during the printing process. Brief Description of the Drawings

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following - described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0052] Figure 1 It is the flowchart of the method of the present invention;

[0053] Figure 2 It is the overall structure schematic diagram of the system of the present invention;

[0054] Figure 3 It is the installation schematic diagram of the optical sensor structure of the present invention;

[0055] Figure 4 It is the data collection structure schematic diagram of the optical sensor of the present invention;

[0056] Figure 5 It is the schematic diagram of the two - dimensional molten pool intensity cloud map of the present invention;

[0057] Figure 6 It is the reconstruction schematic diagram of the three - dimensional molten pool intensity cloud map of the present invention;

[0058] Figure 7 It is the schematic diagram of defect recognition and warning of the present invention;

[0059] Figure 8 It is the schematic diagram of the user interface of the present invention. Detailed Embodiments

[0060] Next, in combination with the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0061] As Figure 1 shown, the present invention provides a real-time molten pool monitoring and defect warning method for additive manufacturing, including the following steps:

[0062] S1, obtaining real-time molten pool intensity data of the entire printing layer.

[0063] As Figure 2 and Figure 3 , in step S1 of this embodiment, it includes: using an optical sensor as the data acquisition device, and the optical sensor is installed at a paraxial position of the laser scanning path. The optical sensor moves along with the laser scanning path to collect data on the molten pool intensity of the entire printing area, and obtains the full-area intensity distribution of the molten pool during the printing process in real time. The optical sensor is set at an inclination angle of 15 - 20° with the horizontal plane.

[0064] It should be noted that the optical sensor can be selected as a CMOS camera. The optical sensor is installed at a paraxial position of the laser scanning path to ensure that it can cover the entire printing area. In order to improve the data quality, the system has optimized the acquisition optical path. The sensor is configured with a suitable filter and lens to improve the quality of the monitoring signal and eliminate ambient light interference.

[0065] Among them, the lens is selected as a low-distortion industrial-grade optical lens, which reduces edge distortion and astigmatism effects compared with traditional lenses, ensuring the accuracy of imaging. And the traditional vertical top-down acquisition method is easily interfered by laser reflection. In this system, the camera is installed at an inclination angle of 15 - 20°, reducing the direct reception of laser reflection light by the lens and improving the image signal quality. Moreover, it also includes the optimization of the filter. A narrow-bandpass filter is used, and the spectral range of the narrow-bandpass filter is 880nm - 930nm, enabling the optical sensor to receive the corresponding band light signals radiated by the molten pool, reducing noise interference by filtering out ambient light, and improving the signal quality. The narrow-bandpass filter adopts a high-transmittance coating process to reduce signal loss and make the actual intensity data of the molten pool clearer.

[0066] As Figure 4As shown, the optical sensor needs to have a high sampling rate and high resolution to ensure that it can clearly capture the subtle changes in the molten pool intensity. A high-speed data acquisition device is used to sample the molten pool intensity data captured by the sensor in real time, ensuring that the acquisition frequency is synchronized with the laser scanning speed to ensure the temporal and spatial consistency of the monitoring data. The data storage device uses a solid-state drive or other high-speed storage devices to ensure the real-time storage of high-frequency data for subsequent analysis and processing.

[0067] In this embodiment, a high-precision optical sensor and high-speed data acquisition technology are adopted to realize the real-time monitoring of the full-width molten pool intensity during the printing process, and can accurately record the spatial distribution and dynamic changes of the molten pool intensity, providing a reliable guarantee for the printing quality of parts.

[0068] S2. Preprocess the real-time molten pool intensity data, convert the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually display the two-dimensional molten pool intensity cloud map of the printing layer to generate a molten pool intensity array map.

[0069] In step S2 of this embodiment, the following sub-steps are included:

[0070] S21. Clean, denoise, and correct the collected real-time molten pool intensity data, and use an image processing algorithm to smooth the collected full-width molten pool intensity distribution data to obtain the standard molten pool intensity data.

[0071] S22. Convert the standard molten pool intensity data into a grayscale image to obtain the two-dimensional matrix map of the molten pool intensity for each layer.

[0072] S23. Traverse the elements in the two-dimensional molten pool intensity matrix map in turn using color mapping, match the corresponding color according to the intensity value of each pixel, and fill it into the corresponding grayscale image to obtain the two-dimensional molten pool intensity cloud map.

[0073] S24. Use the PictureBox control to display the two-dimensional molten pool intensity cloud map of each layer on the interface to show the distribution of the molten pool intensity on the printing surface.

[0074] It should be noted that after each layer of printing is completed, the molten pool intensity data matrix is automatically saved to record the complete intensity distribution information of the current layer. The collected data is preprocessed, including denoising, calibration, and data enhancement. Denoising uses methods such as median filtering and bilateral filtering to eliminate environmental noise and data acquisition errors. Calibration is based on calibration data to correct the geometric distortion of the image to ensure the accuracy of the intensity distribution. Data enhancement optimizes the continuity of the intensity data through a smoothing algorithm to eliminate local data deviations caused by changes in the scanning speed, improving the clarity and accuracy of the molten pool intensity distribution map.

[0075] Such as Figure 5As shown, the processed molten pool intensity data is converted into a two-dimensional cloud map to visually display the distribution of the molten pool intensity on the printing surface. The molten pool intensity cloud map represents the intensity magnitude through color gradients, such as the color change from blue to red, and marks the abnormal intensity regions, such as insufficient intensity or excessive intensity.

[0076] Specifically, in this embodiment, the visualization of the two-dimensional molten pool intensity cloud map uses the C#.NET Framework platform and utilizes the PictureBox control for interface display; First, the system stores the collected molten pool intensity data in the form of a grayscale image and converts it into temperature or intensity data; To enhance visual recognition, a pseudo-color mapping method is adopted to make the color distribution of different intensity regions more intuitive.

[0077] During the image processing, the Bitmap class is used to store and process the data, and the LockBits method is used to optimize the pixel access speed to improve the data processing efficiency; The color mapping adopts the Jet or Parula pseudo-color mapping scheme, traverses the molten pool data matrix, matches the corresponding color according to the intensity value of each pixel, and fills it into the Bitmap object; Finally, the processed cloud map is displayed on the user interface through the PictureBox.Image property, providing real-time molten pool intensity distribution information for the operator.

[0078] Since additive manufacturing is carried out layer by layer, the system automatically refreshes the PictureBox after each layer is printed to ensure that the latest molten pool intensity cloud map can be presented in real time; To improve the display smoothness, double-buffering technology is adopted to reduce the flicker phenomenon during image refreshing.

[0079] S3. According to the layer-by-layer molten pool intensity cloud map data and based on the three-dimensional reconstruction algorithm, construct a three-dimensional molten pool intensity distribution model to obtain a three-dimensional molten pool intensity cloud map.

[0080] The steps in step S3 of this embodiment include the following sub-steps:

[0081] Perform standardization, denoising, and smoothing processing on the layer-by-layer collected two-dimensional molten pool intensity cloud map, and use bilateral filtering to remove noise;

[0082] Use a three-dimensional interpolation method to fill the data missing regions in the denoised two-dimensional molten pool intensity cloud map to obtain a standard map of the molten pool intensity array;

[0083] According to the layer-by-layer standard map of the molten pool intensity array and based on the three-dimensional reconstruction algorithm, construct a three-dimensional molten pool intensity distribution model to obtain a three-dimensional molten pool intensity cloud map. The three-dimensional molten pool intensity distribution expression is:

[0084]

[0085] Wherein, I(x, y, z) is the melt pool intensity value at the position (x, y, z) in the three-dimensional melt pool intensity cloud map, (x, y, z) is a three-dimensional coordinate system, x and y are the two-dimensional plane coordinates of the printing area, z is the number of printing layers, and I k (x, y) is the melt pool intensity data of each layer, N is the total number of printing layers, d is the layer thickness, and δ(z - k·d) is the mapping of the melt pool intensity data of the k-th layer in the z direction;

[0086] The volume rendering Marching Cubes algorithm is used to combine the melt pool intensity array maps into a continuous three-dimensional model, and the visual effect is enhanced through pseudo-color mapping.

[0087] As Figure 6 shown, it should be noted that the reconstruction of the three-dimensional melt pool intensity cloud map is based on the two-dimensional melt pool intensity data accumulated layer by layer, and a complete three-dimensional melt pool intensity model is generated by combining a three-dimensional reconstruction algorithm. This model can intuitively display the change trend of the melt pool intensity in the height direction during the printing process, providing an important basis for printing quality evaluation and defect identification.

[0088] The whole process mainly includes data preprocessing, three-dimensional reconstruction, feature extraction and defect identification; the data preprocessing stage includes standardizing, denoising and smoothing the two-dimensional melt pool intensity cloud maps collected layer by layer, using the bilateral filtering method to remove noise while maintaining the clarity of the melt pool boundary; in order to ensure the consistency of the melt pool intensity distribution between different layers, histogram matching is required to reduce the intensity deviation between layers; the system uses a three-dimensional interpolation method to fill the possible data missing areas.

[0089] In this embodiment, through data processing and visualization technology, the two-dimensional melt pool intensity cloud map of each layer can be generated, and the three-dimensional melt pool intensity cloud map of the printing process can be generated by using three-dimensional reconstruction technology; the three-dimensional cloud map intuitively shows the overall distribution of the melt pool intensity during the printing process, providing comprehensive support for defect identification and quality evaluation.

[0090] S4. Perform defect identification on the two-dimensional melt pool intensity cloud map and the three-dimensional melt pool intensity cloud map respectively, extract the defect area and mark it to issue a warning.

[0091] Step S4 includes defect detection on the two-dimensional melt pool intensity cloud map, which includes the following steps:

[0092] Calculate the average intensity μ and standard deviation σ of the two-dimensional melt pool intensity cloud map, and set the abnormal area threshold according to the average intensity μ and standard deviation σ;

[0093] Compare the abnormal area threshold with the intensity of the two-dimensional melt pool intensity cloud map. If it is not within the range of the abnormal area threshold, highlight and mark the defect area on the two-dimensional melt pool intensity cloud map and issue a warning message. The expression is:

[0094] T low ≤I k (x, y) ≤ T high ;

[0095] In the formula, T low is the lower limit of the abnormal area threshold range, and T low = μ - 3σ; I k (x, y) is the melt pool strength data of each layer, and T high is the upper limit of the abnormal area threshold range, and T high = μ + 3σ.

[0096] As Figure 7 and Figure 8 shown, it should be noted that if the strength value of a certain area is lower than T low , it means that the material is not fully melted, resulting in pore defects; while if the strength value is higher than T high , it may cause overburning or spattering of the material. After detecting an abnormality, the system will mark the defective area on the cloud map and pop up a warning message on the user interface to remind the operator to adjust the laser power or scanning speed in time.

[0097] Among them, step S4 includes defect detection of the three-dimensional melt pool strength cloud map, including the following steps:

[0098] S41. Obtain the three-dimensional melt pool strength cloud map data, calculate the spatial gradient of the melt pool strength by using the numerical differentiation method, and perform edge detection by using the Laplace operator to obtain the melt pool edge image. The expression is:

[0099]

[0100] In the formula, is the result of the Laplace operator acting on the melt pool strength value at the position (x, y, z) in the three-dimensional melt pool strength cloud map, is the change rate of the three-dimensional melt pool strength distribution model in the x direction, is the change rate of the three-dimensional melt pool strength distribution model in the y direction; is the change rate of the three-dimensional melt pool strength distribution model in the z direction;

[0101] S42. Use three-dimensional morphological filtering to extract the abnormal area from the melt pool edge image;

[0102] S43. Perform voxel analysis on the extracted abnormal area, calculate its volume and density. The expression is:

[0103]

[0104] Wherein, v is the volume of the abnormal area, R is the abnormal area, and δI(x, y, z) is the volume of a single voxel in the abnormal area; ρ is the density of the abnormal area, and m is the total mass of the abnormal area;

[0105] S44. Judge the defect according to the volume and density of the abnormal area. If it is a defect area, highlight it on the three-dimensional molten pool intensity cloud map and send out a warning message.

[0106] Such as Figure 6 and Figure 8 As shown, step S44 includes: presetting a low density threshold, a high density threshold, and a volume threshold, comparing the volume and density of the abnormal area with the density threshold and the volume threshold. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is less than the low density threshold, it is an unfused defect, and it is highlighted in blue on the three-dimensional molten pool intensity cloud map and a warning message is sent out. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is greater than the high density threshold, it is an overburned area, and it is highlighted in red on the three-dimensional molten pool intensity cloud map and a warning message is sent out. Otherwise, it is in a normal state.

[0107] It should be noted that a defect recognition algorithm based on statistics and machine learning is constructed to analyze the intensity abnormal areas in the two-dimensional cloud map and the three-dimensional model. According to the empirical data and algorithm thresholds, the areas that may have defects are automatically marked. The system sends out real-time warnings to the operators according to the defect recognition results, indicating potential printing defects and their locations. The warning information includes the defect type, location, and possible impacts, providing a reference for the operators to adjust the process parameters in a timely manner.

[0108] In this embodiment, by integrating the defect recognition algorithm and the warning mechanism, potential defects such as insufficient or excessive molten pool intensity can be quickly detected and a warning is sent to the operator in real time; the operator can adjust the process parameters in a timely manner according to the warning information, thereby reducing the defect rate during the printing process.

[0109] In this embodiment, through the full-frame real-time monitoring and defect warning of the molten pool intensity during the laser powder bed metal additive manufacturing process, the quality control ability of additive manufacturing is significantly improved.

[0110] In addition, in order to meet the real-time requirements of laser powder bed fusion additive manufacturing, the system optimizes the data acquisition, processing, and display processes to ensure that the overall delay does not exceed 100 milliseconds; the real-time optimization mainly involves data acquisition acceleration, parallel computing optimization, and cache management;

[0111] In terms of data acquisition acceleration, the system adopts a multi-threaded asynchronous data acquisition mechanism, and uses DirectMemoryAccess to directly read the molten pool intensity data from the sensor, avoiding CPU overload and reducing data transmission delay at the same time.

[0112] In terms of parallel computing optimization, the system uses GPU accelerated computing and utilizes CUDA or OpenCL technology for two-dimensional molten pool cloud map processing, three-dimensional reconstruction, and defect recognition. GPU parallel computing can significantly improve data processing efficiency. For example, when using the fast Fourier transform for noise removal, its computing speed is more than 10 times higher than that of traditional CPU solutions.

[0113] In terms of cache management, the system adopts the methods of double buffering and data stream batch processing to ensure seamless connection between data acquisition, processing, and display; for the molten pool data of each layer, the system performs pre-computation at the backend and stores it in the cache, and directly calls the cached data when displaying on the interface to reduce the loading time and improve the user interaction experience.

[0114] To ensure that the system can operate stably in complex industrial environments, the system has optimized the sensor filtering algorithm, signal anti-interference design, and intelligent anomaly detection.

[0115] In terms of the sensor filtering algorithm, the system adopts adaptive Kalman filtering, which can dynamically adjust the filtering parameters to improve the accuracy of molten pool intensity data. For example, when detecting a high-noise environment such as an increase in equipment vibration, the system will automatically increase the filtering intensity to reduce the error caused by signal fluctuations.

[0116] In terms of signal anti-interference design, for the data transmission of optical sensors, the system has added background light compensation and optical filters to reduce the impact of ambient light on molten pool intensity data.

[0117] In terms of intelligent anomaly detection, the system continuously monitors the stability of sensor data and uses principal component analysis and statistical threshold detection to identify sudden interference signals. For example, if the data of a certain layer shows abnormal high-frequency fluctuations, the system will automatically mark the data as abnormal and prompt the operator to check the sensor status.

[0118] In terms of algorithm expansion, the system provides a machine learning interface that allows users to customize defect recognition models. For example, users can train their own CNN convolutional neural network models and access the system through the API to achieve personalized defect detection.

[0119] In terms of process adaptation, the system supports different models of additive manufacturing equipment, such as SLM120, SLM280A, and their specific printing processes; through adjustable parameter configuration, the system can adapt to different laser power ranges, scanning strategies, and printing materials to ensure efficient monitoring and optimization support under different process conditions.

[0120] Second aspect, the present invention also provides a real-time molten pool monitoring and defect warning system for additive manufacturing, which is implemented by using the real-time molten pool monitoring and defect warning method for additive manufacturing as described above. The system includes:

[0121] An acquisition module, configured to obtain real-time molten pool intensity data of the entire printing layer.

[0122] A two-dimensional cloud map construction module, configured to preprocess the real-time molten pool intensity data, convert the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually display the two-dimensional molten pool intensity cloud map of the printing layer to generate a molten pool intensity array map.

[0123] A three-dimensional cloud map construction module, configured to construct a three-dimensional molten pool intensity distribution model based on the layer-by-layer molten pool intensity cloud map data and based on a three-dimensional reconstruction algorithm, and obtain a three-dimensional molten pool intensity cloud map.

[0124] A defect recognition module, configured to respectively perform defect recognition on the two-dimensional molten pool intensity cloud map and the three-dimensional molten pool intensity cloud map, extract the defect area and mark it to issue a warning.

[0125] It should be noted that this system is a system corresponding to the above real-time molten pool monitoring and defect warning method for additive manufacturing. All implementation manners in the above method embodiments are applicable to the embodiments of this system and can also achieve the same technical effects.

[0126] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0127] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems and modules described above can refer to the corresponding processes in the foregoing method embodiments and will not be repeated here.

[0128] In the embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0129] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0130] In addition, in each embodiment of the present invention, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0131] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs that can store program codes.

[0132] In addition, it should be noted that in the system and method of the present invention, obviously, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations shall be regarded as equivalent solutions of the present invention. Moreover, the steps of performing the above series of processes can naturally be executed in chronological order according to the described order, but it is not necessary to be executed in chronological order. Some steps can be executed in parallel or independently of each other. For those of ordinary skill in the art, it is understandable that all or any steps or components of the method and device of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in the form of hardware, firmware, software, or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0133] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing system. The computing system can be a well-known general system. Therefore, the object of the present invention can also be achieved only by providing a program product containing program codes for implementing the method or device. That is to say, such a program product also constitutes the present invention, and a storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be noted that in the device and method of the present invention, obviously, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations shall be regarded as equivalent solutions of the present invention. Moreover, the steps of performing the above series of processes can naturally be executed in chronological order according to the described order, but it is not necessary to be executed in chronological order. Some steps can be executed in parallel or independently of each other.

[0134] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A real-time molten pool monitoring and defect warning method for additive manufacturing, characterized in that It includes the following steps: S1. Obtain the real-time molten pool intensity data of the entire printing layer; S2. Preprocess the real-time molten pool intensity data, convert the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually display the two-dimensional molten pool intensity cloud map of the printing layer to generate a molten pool intensity array map; S3. According to the layer-by-layer molten pool intensity cloud map data, construct a three-dimensional molten pool intensity distribution model based on a three-dimensional reconstruction algorithm to obtain a three-dimensional molten pool intensity cloud map; S4. Perform defect identification on the two-dimensional molten pool intensity cloud map and the three-dimensional molten pool intensity cloud map respectively, extract the defect areas and issue warnings.

2. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 1, wherein In step S1, obtaining the real-time molten pool intensity data of the entire printing layer includes: using an optical sensor as the data acquisition device, and the optical sensor is installed at a paraxial position of the laser scanning path. The optical sensor moves along with the laser scanning path to collect the molten pool intensity data of the entire printing area, and obtains the full-area intensity distribution of the molten pool during the printing process in real time. The optical sensor is set at an inclination angle of 15 - 20° with the horizontal plane.

3. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 2, wherein, The data acquisition device further includes a narrowband pass filter, and the spectral range of the narrowband pass filter is 880nm - 930nm, so that the optical sensor receives the corresponding band light signal radiated by the molten pool and reduces noise interference.

4. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 2, wherein In step S2, the preprocessing of the real-time molten pool intensity data, converting the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually displaying the two-dimensional molten pool intensity cloud map of the printing layer to generate a molten pool intensity array map includes the following sub-steps: S21. Clean, denoise, and correct the collected real-time molten pool intensity data, and use an image processing algorithm to smooth the collected full-area molten pool intensity distribution data to obtain the standard molten pool intensity data; S22. Convert the standard molten pool intensity data into a grayscale image to obtain the two-dimensional matrix map of the molten pool intensity of each layer; S23. Traverse the elements in the two-dimensional matrix map of the molten pool intensity in sequence using color mapping, match the corresponding color according to the intensity value of each pixel, and fill it into the corresponding grayscale image to obtain the two-dimensional molten pool intensity cloud map; S24. Display the two-dimensional molten pool intensity cloud map of each layer on the interface to show the distribution of the molten pool intensity on the printing area.

5. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 4, wherein In step S3, according to the layer-by-layer molten pool intensity cloud map data, constructing a three-dimensional molten pool intensity distribution model based on a three-dimensional reconstruction algorithm includes the following sub-steps: Perform standardization, denoising, and smoothing processing on the layer-by-layer collected two-dimensional molten pool intensity cloud maps, and use bilateral filtering to remove noise; Use a three-dimensional interpolation method to fill the data missing areas in the denoised two-dimensional molten pool intensity cloud map to obtain the standard molten pool intensity array map; According to the layer-by-layer standard molten pool intensity array maps, construct a three-dimensional molten pool intensity distribution model based on a three-dimensional reconstruction algorithm to obtain a three-dimensional molten pool intensity cloud map. The three-dimensional molten pool intensity distribution expression is: Wherein, I(x, y, z) is the melt pool intensity value at the position (x, y, z) in the three-dimensional melt pool intensity cloud map, (x, y, z) is a three-dimensional coordinate system, x and y are the two-dimensional plane coordinates of the printing area, z is the number of printing layers, and I k (x, y) is the melt pool intensity data of each layer, N is the total number of printing layers, d is the layer thickness, and δ(z - k·d) is the mapping of the melt pool intensity data of the k-th layer in the z direction; Combine the molten pool intensity array maps into a continuous three-dimensional model and enhance the visual effect through pseudo-color mapping.

6. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 5, characterized in that: In step S4, it includes defect detection of the two-dimensional molten pool intensity cloud map, including the following steps: Calculate the average intensity μ and standard deviation σ of the two-dimensional molten pool intensity cloud map, and set the abnormal area threshold according to the average intensity μ and standard deviation σ; Compare the abnormal area threshold with the intensity of the two-dimensional molten pool intensity cloud map. If it is not within the range of the abnormal area threshold, highlight the defect area on the two-dimensional molten pool intensity cloud map and send a warning message. The expression is: T low ≤I k (x, y) ≤ T high ; Wherein, T low is the low value of the abnormal area threshold range, and T low = μ - 3σ; I k (x, y) is the molten pool intensity data of each layer, and T high is the high value of the abnormal area threshold range, and T high = μ + 3σ.

7. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 5, characterized in that: Step S4 includes defect detection of the three-dimensional molten pool intensity cloud map, including the following steps: S41, Obtain the three-dimensional molten pool intensity cloud map data, calculate the spatial gradient of the molten pool intensity using the numerical differentiation method, and perform edge detection using the Laplace operator to obtain the molten pool edge image. The expression is: In the formula, is the result of the Laplace operator acting on the molten pool intensity value at the position (x, y, z) in the three-dimensional molten pool intensity cloud map, is the change rate of the three-dimensional molten pool intensity distribution model in the x direction, is the change rate of the three-dimensional molten pool intensity distribution model in the y direction; is the change rate of the three-dimensional molten pool intensity distribution model in the z direction; S42, Use three-dimensional morphological filtering to extract the abnormal area from the molten pool edge image; S43, Perform voxel analysis on the extracted abnormal area, and calculate its volume and density. The expression is: In the formula, v is the volume of the abnormal area, R is the abnormal area, and δI(x, y, z) is the volume of a single voxel in the abnormal area; ρ is the density of the abnormal area, and m is the total mass of the abnormal area; S44, Judge the defect according to the volume and density of the abnormal area. If it is a defect area, highlight it on the three-dimensional molten pool intensity cloud map and send a warning message.

8. The real-time molten pool monitoring and defect warning method for additive manufacturing according to claim 7, characterized in that: In step S44, the judgment of the defect according to the volume and density of the abnormal area, if it is a defect area, highlight it on the three-dimensional molten pool intensity cloud map and send a warning message, including: Preset a low density threshold, a high density threshold, and a volume threshold. Compare the volume and density of the abnormal area with the density threshold and volume threshold. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is less than the low density threshold, it is an unfused defect, and it is highlighted in blue on the three-dimensional molten pool intensity cloud map and a warning message is sent. If the volume of the abnormal area is greater than the volume threshold and the density of the abnormal area is greater than the high density threshold, it is an overburned area, and it is highlighted in red on the three-dimensional molten pool intensity cloud map and a warning message is sent. Otherwise, it is in a normal state.

9. A real-time molten pool monitoring and defect warning system for additive manufacturing, which is implemented by using a real-time molten pool monitoring and defect warning method for additive manufacturing described in any one of claims 1-8, and is characterized in that, The system includes: A collection module for obtaining real-time molten pool intensity data of the full print layer; A two-dimensional cloud map construction module for preprocessing the real-time molten pool intensity data, converting the processed molten pool intensity data into a two-dimensional molten pool intensity cloud map, and visually displaying the two-dimensional molten pool intensity cloud map of the print layer to generate a molten pool intensity array map; A three-dimensional cloud map construction module for constructing a three-dimensional molten pool intensity distribution model based on the layer-by-layer molten pool intensity cloud map data and obtaining a three-dimensional molten pool intensity cloud map based on a three-dimensional reconstruction algorithm; A defect recognition module for respectively performing defect recognition on the two-dimensional molten pool intensity cloud map and the three-dimensional molten pool intensity cloud map, extracting the defect area and marking it to send a warning.

10. A computer-readable storage medium, characterized in that, A program for a real-time molten pool monitoring and defect warning method for additive manufacturing is stored on the storage medium. When the program for a real-time molten pool monitoring and defect warning method for additive manufacturing is executed, it implements a real-time molten pool monitoring and defect warning method as described in any one of claims 1 to 8.

Images