CN107175329B - A 3D printing device and method for layer-by-layer detection and reverse part model and defect location - Google Patents

Abstract

The invention discloses a device and a method for detecting a reverse part model and positioning defects layer by layer in 3D printing; the device comprises a scanning galvanometer, a semi-transparent semi-reflecting mirror, an optical filter, a laser head, a high-speed camera, a controller and the like. The scanning galvanometer is used for controlling laser beams to selectively melt metal powder, and the semi-transparent semi-reflecting mirror reflects the molten pool radiation into the high-speed camera and converts the molten pool radiation into image information which is transmitted to the controller. The device monitors powder melting in the SLM processing process, feeds the powder melting back to a computer software interface, reflects molten pool characteristics at different positions in real time, accurately measures the profile of each melting layer, obtains a part model in a reverse solving mode, and compares and analyzes the model and an original three-dimensional model to obtain the error of the metal 3D printed part and the original model data in the aspect of precision and size. The position and the three-dimensional shape of the internal defect in the 3D printing process can be accurately acquired, and the destructive test of the printing part in the later stage on the part is avoided.

Description

A 3D printing device and method for layer-by-layer detection and reverse part model and defect location

technical field

The invention relates to metal 3D printing process monitoring and precise quality control, in particular to a 3D printing layer-by-layer detection reverse part model and a device and method for locating defects.

Background technique

Selective Laser Melting (SLM) technology is a rapid prototyping 3D printing technology that can directly form high-density, high-precision metal parts, but more than 50 different factors are at play in the melting process, such as size and Shape errors, voids in the fused layer, high residual stress in the final part, and the interrelationship of various variables such as material properties make the printing process difficult to quantitatively control.

The development of quality monitoring has led to a significant improvement in the surface roughness and performance of the molded parts in additive manufacturing technology, reducing the deformation of the internal structure. A series of key parameters need to be monitored in the laser melting system, including oxygen content, laser output power, powder spreading and powder quality, etc. However, it is not enough to comprehensively evaluate the quality of parts simply based on equipment process, the printing process itself must be monitored. The real-time monitoring system can make an effective contribution to the early and effective detection of printing defects and the avoidance of defects.

Concept laser's QM melt pool 3D system monitors the entire printing process through photodiodes and CMOS cameras, and uses coaxial sensors to monitor the thermal radiation of the melt pool; One point, one layer, or one component. In this way, it allows for automated monitoring of the melt pool, while it also enables insight into the interior of the part during the build process.

At present, the difficulty of quality monitoring lies in the accuracy of information collection and processing, and whether it can accurately reflect the processing status; as well as the correction of the processing process, printing defects or self-organization defects caused by a large number of influencing factors, and the whole process is highly dynamic. characteristics, developing a self-correcting control loop is a major difficulty.

Contents of the invention

The purpose of the present invention is to overcome the shortcomings and deficiencies of the above-mentioned prior art, and provide a device and method for 3D printing layer-by-layer detection, reverse part model and defect location. The present invention monitors the location and characteristics of the molten pool, and reversely calculates the three-dimensional model through the contour data of each layer. Such signals can be visualized and analyzed on the three-dimensional model immediately after the printing process is completed. Users can trace the printing process of each part based on location. Influences generated within the part during the printing process can be better detected and analyzed. By analyzing the printing defects of parts, find out the causes and propose solutions.

The present invention realizes through following technical scheme:

A 3D printing device for layer-by-layer detection of reverse part models and positioning defects, including a laser head 3, a scanning galvanometer 9, a computer 1, a half-transparent mirror 16, a high-speed camera 20, and a controller 2; the high-speed camera 20 passes through The controller 2 is connected to the computer 1 by telecommunication;

The laser light path 17 of the laser head 3 is reflected into the scanning vibrating mirror 9 through the semi-transparent mirror 16, and the laser beam is controlled by the scanning vibrating mirror 9 to selectively melt the metal powder laid on the work platform 15; The mirror 9 collects the molten pool radiation, and transmits it to the high-speed camera 20 through the half-transparent mirror 16, and the high-speed camera 20 processes the molten pool radiation data, and converts the image information into the controller 2, and the controller 2 uses It is used to process the image data to determine the position of the molten pool and generate the contour of each molten layer.

The controller 2 includes: an image acquisition module, an image contour extraction module, and an image triangulation module;

The image acquisition module is used to control the high-speed camera 20 to collect the real-time image data of the melting pool during the forming process of each layer of the workpiece, and store it in its memory;

The image contour extraction module displays the color image fed back to the high-speed camera 20 as a grayscale image, and establishes its coordinate system; uses the median filter template to filter the grayscale image to smooth the image and remove noise; utilizes the grayscale histogram , select the threshold of the histogram as the minimum value, perform binarization on the image according to the threshold, divide it into molten pool pixels and non-melted pool pixels, and extract the molten pool outline;

The image triangulation module approximates the fault contour obtained by image processing with a polygon, and then connects the adjacent fault polygon vertices to form a triangle, and then triangulates the upper and lower end faces of the object, and outputs the STL file;

At the beginning of the working cycle, the image information is collected by the image acquisition module, which is transmitted to the image contour extraction module to extract the contour information of the molten pool, and the process file is established according to the information, and the processing status is fed back on the computer interface. After the processing of this layer is completed, according to the process The contour of the layer is extracted from the file; the image triangulation module obtains the complete three-dimensional model of the workpiece according to the multi-layer contour of the workpiece, and outputs the STL file.

An optical filter 19 is added on the optical path between the high-speed camera 20 and the half-mirror 16 to filter out the melting pool collection band.

The optical filter 19 adopts a narrow-band optical filter with a center wavelength in the range of 600-650 nm to ensure the spectral sensitivity of the high-speed camera 20 .

The high-speed camera 20 is a COMS high-speed camera with a pixel resolution of not less than 1024×1024 and a frame rate of 7000 frames per second; the shortest exposure time of the overall shutter is 1us; the dynamic range is 120dB; the spectral range is 400nm-950nm, and 8-bit sampling resolution.

A 3D printing layer-by-layer detection reverse part model and a method for locating defects, which comprises the following steps:

Step 1: The scanning galvanometer 9, the half-transparent mirror 16, and the optical filter 19 form a coaxial optical path, and the molten pool radiated light is reflected and filtered to the high-speed camera 20 through the coaxial optical path;

Step 2: Establish a coordinate system with the center of the workpiece forming plane as the origin, and the high-speed camera 20 captures the position of the molten pool on the forming plane according to the plane forming trajectory, and simultaneously records the shape of the molten pool at this position;

Step 3: After the image processing of the controller 2, the size of the molten pool is obtained. When the size of the molten pool deviates from the deviation range of the standard value, it is recorded as an abnormal position, otherwise it is a normal position; the controller 2 feeds back the position information to the computer 1 in real time. On the monitoring interface, the molten pool information is reflected in the corresponding position of the monitoring interface. If it is a normal position, it will display green, and if it is an abnormal position, it will display red;

Step 4: After the data processing of this layer is completed, the high-speed camera 20 collects the forming plane data of this layer, and the controller 2 extracts and saves the contour data of this layer of the workpiece; after the overall processing of the part is completed, generate a three-dimensional model, compare and analyze the currently generated 3D model with the original 3D model pre-built in the computer 1, and obtain the error in the accuracy and size of the metal 3D printing parts and the original model data; at the same time, the abnormal position in the model is red high lights up, and displays the number of layers available for viewing.

The deviation range of the molten pool size in step three from the standard value is 5%-15%.

Compared with the prior art, the present invention has the following advantages and effects:

The invention monitors the powder melting in the SLM processing process, and feeds back to the computer to reflect the characteristics of the molten pool at different positions in real time, and accurately measures the contour (including the internal closed contour) of each melting layer, and obtains the part model through the reverse method. The model is compared and analyzed with the original 3D model to obtain the error in the accuracy and size of the metal 3D printed parts and the original model data. At the same time, combined with the data analysis of the characteristics of the molten pool at different positions (the width of the molten pool after solidification), the position and three-dimensional shape of the internal defects during the 3D printing process can be accurately obtained, avoiding the destructive test of the printed parts in the later stage.

Description of drawings

Figure 1 is a schematic diagram of the structure of the 3D printing layer-by-layer detection reverse part model and defect location device of the present invention.

Figure 2 is a schematic diagram of the computer interface; A in the figure represents the molding layer of each layer of the part; B represents the abnormality.

Fig. 3 is the working flow chart of the present invention.

In Figure 1: computer 1, controller 2, laser head 3, auxiliary structure beam expander 4, three-dimensional dynamic focusing system 5, control board 6, control board 7, vibrating mirror control card 8, scanning vibrating mirror 9, Y scanning motor And its lens 10, X scanning motor and its lens 11, control panel (12,13,14), working platform 15, half-transparent mirror 16, laser light path 17, melting pool radiation light path 18, optical filter 19, high-speed camera 20.

Detailed ways

The present invention will be described in further detail below in conjunction with specific embodiments.

Example

as the picture shows. The invention discloses a 3D printing layer-by-layer detection and reverse parts model and a device for locating defects, including a laser head 3, a scanning galvanometer 9, a computer 1, a half-transparent mirror 16, a high-speed camera 20, and a controller 2; The high-speed camera 20 is telecommunications connected with the computer 1 through the controller 2;

The laser light path 17 of the laser head 3 is reflected into the scanning vibrating mirror 9 through the semi-transparent mirror 16, and the laser beam is controlled by the scanning vibrating mirror 9 to selectively melt the metal powder laid on the work platform 15; The mirror 9 collects the molten pool radiation, and transmits it to the high-speed camera 20 through the half-transparent mirror 16, and the high-speed camera 20 processes the molten pool radiation data, and converts the image information into the controller 2, and the controller 2 uses It is used to process the image data to determine the position of the molten pool and generate the contour of each molten layer.

An optical filter 19 is added on the optical path between the high-speed camera 20 and the half-mirror 16 to filter out the melting pool collection band.

The optical filter 19 adopts a narrow-band optical filter with a center wavelength in the range of 600-650 nm to ensure the spectral sensitivity of the high-speed camera 20 .

The high-speed camera 20 is a COMS high-speed camera with a pixel resolution of not less than 1024×1024 and a frame rate of 7000 frames per second; the shortest exposure time of the overall shutter is 1us; the dynamic range is 120dB; the spectral range is 400nm-950nm, and 8-bit sampling resolution.

The half-mirror 16 is used for 100% reflection of the 1064nm laser wavelength, and allows 100% transmission of visible light and near-infrared light to the high-speed camera 20 .

The controller 2 includes: an image acquisition module, an image contour extraction module, and an image triangulation module.

The image acquisition module is used to control the high-speed camera 20 to collect the real-time image data of the melting pool during the forming process of each layer of the workpiece, and store it in its memory;

The image contour extraction module displays the color image fed back to the high-speed camera 20 as a grayscale image, and establishes its coordinate system; uses the median filter template to filter the grayscale image to smooth the image and remove noise; utilizes the grayscale histogram , select the threshold of the histogram as the minimum value, perform binarization on the image according to the threshold, divide it into molten pool pixels and non-melted pool pixels, and extract the molten pool outline;

The image triangulation module approximates the fault contour obtained by image processing with a polygon, and then connects the adjacent fault polygon vertices to form a triangle, and then triangulates the upper and lower end faces of the object, and outputs the STL file;

At the beginning of the working cycle, the image information is collected by the image acquisition module, which is transmitted to the image contour extraction module to extract the contour information of the molten pool, and the process file is established according to the information, and the processing status is fed back on the computer interface. After the processing of this layer is completed, according to the process The contour of the layer is extracted from the file; the image triangulation module obtains the complete three-dimensional model of the workpiece according to the multi-layer contour of the workpiece, and outputs the STL file.

The 3D printing method of the present invention detects the reverse part model and locates the defect layer by layer, which can be realized through the following steps:

Step 1: The scanning galvanometer 9, the half mirror 16 and the optical filter 19 form a coaxial optical path, and the radiated light of the melting pool is reflected and filtered to the high-speed camera 20 through the coaxial optical path;

Step 2: Establish a coordinate system with the center of the workpiece forming plane as the origin, and the high-speed camera 20 captures the position of the molten pool on the forming plane according to the plane forming trajectory, and simultaneously records the shape of the molten pool at this position;

Step 3: After the image processing of the controller 2, the size of the molten pool is obtained. When the size of the molten pool deviates from the deviation range of the standard value, it is recorded as an abnormal position, otherwise it is a normal position; the controller 2 feeds back the position information to the computer 1 in real time. On the monitoring interface, the molten pool information is reflected in the corresponding position of the monitoring interface. If it is a normal position, it will display green, and if it is an abnormal position, it will display red;

Step 4: After the data processing of this layer is completed, the high-speed camera 20 collects the forming plane data of this layer, and the controller 2 extracts and saves the contour data of this layer of the workpiece; after the overall processing of the part is completed, generate a three-dimensional model, compare and analyze the currently generated 3D model with the original 3D model pre-built in the computer 1, and obtain the error in the accuracy and size of the metal 3D printing parts and the original model data; at the same time, the abnormal position in the model is red high lights up, and displays the number of layers available for viewing.

The deviation range of the molten pool size in step three from the standard value is 5%-15%.

The standard value of molten pool size needs to be determined according to powder material, laser energy density, and scanning speed.

As described above, the present invention can be preferably carried out.

The implementation of the present invention is not limited by the above examples, and any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods, and are all included in within the protection scope of the present invention.

Claims (4)

1. A 3D printing layer-by-layer detection reverse part model and defect location method, comprising a 3D printing layer-by-layer detection reverse part model and defect location device, the reverse part model and defect location device includes: a laser head (3), A scanning vibrating mirror (9) and a computer (1); it is characterized in that it also includes: a half-mirror (16), a high-speed camera (20), and a controller (2); the high-speed camera (20) passes through the controller ( 2) Telecommunications connection with computer (1); The laser light path (17) of the laser head (3) is reflected into the scanning galvanometer (9) through the half-transparent mirror (16), and the laser beam is selectively melted and tiled on the working platform by the scanning galvanometer (9) (15) on the metal powder; at the same time, the scanning galvanometer (9) collects the molten pool radiation, and transmits it to the high-speed camera (20) through the half-transparent mirror (16), and the high-speed camera (20) detects the molten pool radiation. The radiation data is processed, converted into image information and sent to the controller (2), and the controller (2) is used to process the image data to determine the position of the molten pool and generate the outline of each molten layer; The controller (2) includes: The image acquisition module is used to control the high-speed camera (20) to collect the real-time image data of the molten pool during the forming process of each layer of the workpiece, and store it in its memory; The image contour extraction module displays the color image fed back to the high-speed camera (20) as a grayscale image, and establishes its coordinate system; uses the median filter template to filter the grayscale image to smooth the image and remove noise; utilizes the grayscale For the histogram, the threshold of the histogram is selected as the minimum value, and the image is binarized according to the threshold, divided into molten pool pixels and non-melted pool pixels, and the molten pool outline is extracted; The image triangulation module approximates the fault contour obtained by image processing with a polygon, and then connects the adjacent fault polygon vertices to form a triangle, and then triangulates the upper and lower end faces of the object, and outputs the STL file; At the beginning of the working cycle, the image information is collected by the image acquisition module, which is transmitted to the image contour extraction module to extract the contour information of the molten pool, and the process file is established according to the information, and the processing status is fed back on the computer interface. After the processing of this layer is completed, according to the process The contour of the layer is extracted from the file; the image triangulation module obtains the complete three-dimensional model of the workpiece according to the multi-layer contour of the workpiece, and outputs the STL file; An optical filter (19) is added on the optical path between the high-speed camera (20) and the half-mirror (16), for filtering out the melting pool collection band; The 3D printing layer-by-layer detection reverse part model and the method for locating defects are realized through the following steps: Step 1: The scanning galvanometer (9), the half-mirror (16), and the filter (19) form a coaxial optical path, and the molten pool radiated light is reflected and filtered to the high-speed camera (20) through the coaxial optical path; Step 2: Establish a coordinate system with the center of the forming plane of the workpiece as the origin, and the high-speed camera (20) captures the position of the molten pool on the forming plane according to the plane forming trajectory, and simultaneously records the shape of the molten pool at this position; Step 3: After the image processing of the controller (2), the size of the molten pool is obtained. When the size of the molten pool deviates from the deviation range of the standard value, it is recorded as an abnormal position, otherwise it is a normal position; the controller (2) feeds back the position information to On the real-time monitoring interface of the computer (1), the melting pool information is reflected at the corresponding position of the monitoring interface. If it is a normal position, it will display green, and if it is an abnormal position, it will display red; Step 4: After the data processing of this layer is completed, the high-speed camera (20) collects the forming plane data of this layer, and the controller (2) extracts and saves the contour data of this layer of the workpiece; after the overall processing of the part is completed, according to each layer of the workpiece Generate a 3D model from the contour data, compare and analyze the currently generated 3D model with the original 3D model pre-built in the computer (1), and obtain the error in the accuracy and size of the metal 3D printing parts and the original model data; at the same time, in the model The abnormal position within is highlighted in red, and the number of layers it is on is displayed for viewing. 2. According to claim 1, the 3D printing layer-by-layer detection reverse part model and defect location method are characterized in that: the optical filter (19) adopts a narrow-band optical filter with a central wavelength in the range of 600-650 nm, To ensure the spectral sensitivity of the high-speed camera (20). 3. According to claim 1, the 3D printing layer-by-layer detection reverse part model and the method for locating defects are characterized in that: the high-speed camera (20) is a COMS high-speed camera with a pixel resolution of not less than 1024×1024 and a frame number of It can reach 7000 frames per second; the shortest exposure time of the overall shutter is 1μs; the dynamic range is 120dB; the spectral range is 400nm-950nm, and the sampling resolution is 8 bits. 4. According to claim 1, the 3D printing layer-by-layer inspection method for reversing part models and locating defects is characterized in that the deviation range of the melt pool size in step 3 from the standard value is 5%-15%.