Online Monitoring System for Laser Additive Manufacturing and Method for Constructing a Digital Twin of a Component

Abstract

The present invention discloses an online monitoring system for laser additive manufacturing and a method for constructing a digital twin of a component. The method generates a geometric model in real time according to the size and morphology of the molten pool and the size and morphology of the deposition layer measured in real time, and synchronously inputs the measurement information of the stress state, temperature, microstructure, and element content into the geometric model, thereby forming digital twins of the molten pool and the deposition layer online. The present invention has the characteristics of fast calculation speed and one-to-one correspondence between the internal state of the component and the actual processed component, aiming to achieve comprehensive monitoring and control of the manufacturing process, facilitate the study of physical phenomena involved in the laser additive manufacturing process, and also facilitate the real-time adjustment and optimization of process parameters.

Description

Technical Field

[0001] The present invention belongs to the field of laser additive manufacturing, and particularly refers to a method for constructing a digital twin including deposition layer size, stress state, element content, and microstructure, and specifically relates to an online monitoring system for laser additive manufacturing and a method for constructing a digital twin of a component. Background Art

[0002] During the laser additive manufacturing process, factors such as deformation and cracking caused by stress, element burn-off caused by high-energy lasers, uneven distribution of microstructure caused by alternating thermal gradients, and unstable forming caused by poor processes have hindered the development of this technology. Although the process parameters of laser additive manufacturing can be optimized through process exploration and simulation calculations to obtain good forming quality and component performance, it is difficult to predict the state of actual production parts. For example, the residual stress, element content, and microstructure distribution of additive manufacturing components need to be obtained through destructive tests. With the increasing demand for intelligent manufacturing, real-time detection of the internal state (such as stress, element content, microstructure, etc.) of the processed object and obtaining the digital twin of the component after manufacturing are of great significance for the intelligent upgrade of the entire manufacturing industry. By constructing a one-to-one digital twin relationship, not only can the manufacturing quality of the component be controlled during the production process, but also key parameters such as deformation, fatigue characteristics, and service life of the component can be traced and predicted during subsequent assembly and service processes, enabling effective tracking and control of the entire life cycle of the component.

[0003] In response to this demand, the present invention proposes an online monitoring system for laser additive manufacturing and a method for constructing a digital twin of a component, aiming to achieve comprehensive monitoring and control of the manufacturing process, improve the quality of additive manufacturing components, and interface with other digital twin systems. Summary of the Invention

[0004] To solve the above problems, the present invention discloses an online monitoring system for laser additive manufacturing and a method for constructing a digital twin of a component, aiming to improve the manufacturing quality of laser additive manufacturing components. By real-time observation of size, stress, temperature, element content, and microstructure, a digital twin of the laser additive manufacturing component is established, and then real-time monitoring and feedback control of the manufacturing process are realized, which is convenient for studying the physical phenomena involved in the laser additive manufacturing process and also facilitates the real-time adjustment and optimization of process parameters.

[0005] To achieve the above object, the technical solution of the present invention is as follows:

[0006] An online monitoring system for laser additive manufacturing includes a substrate and a laser head for emitting laser. The laser head emits laser to melt the powder conveyed coaxially, forming a molten pool on the substrate, and a deposition layer is formed after the molten pool cools. Above the nozzle of the laser head, a hollow shaft motor is installed. The inner ring of the hollow shaft motor is matched with the laser head, and an outer ring of the hollow shaft motor is installed with a detection system for real-time measuring the size and morphology of the molten pool and the size and morphology of the deposition layer, as well as a temperature sensor and a spectrometer. On the back of the substrate, a remelting molten pool morphology monitoring sensor is provided. The system also includes a computer. By subtracting the height of the molten pool from the height of the deposition layer after cooling, the displacement difference generated during the cooling and solidification of the molten pool is obtained and used as a surface displacement load to be loaded onto the three-dimensional model of the molten pool surface and the three-dimensional model of the remelting molten pool, and the stress generated by the cooling shrinkage of the molten pool is calculated.

[0007] Further, the detection system for real-time measuring the size and morphology of the molten pool and the size and morphology of the deposition layer includes a side monitoring CCD, a coaxial monitoring CCD, a line laser generator, and a line laser monitoring CCD. The line laser generator and the line laser monitoring CCD are symmetrically installed, and their axes are in the same plane. The side monitoring CCD is arranged at a certain angle with the line laser generator and the line laser monitoring CCD.

[0008] Further, the temperature sensor and the spectrometer are installed on the outer ring of the hollow shaft motor and aligned with the test points on the deposition layer.

[0009] Further, the remelting molten pool morphology monitoring sensor uses an ultrasonic detector, which is located on the back of the substrate and is used to detect the three-dimensional morphology of the remelting molten pool.

[0010] A method for constructing a component digital twin using the above online monitoring system for laser additive manufacturing. The method is as follows: According to the size and morphology of the molten pool and the size and morphology of the deposition layer measured in real time, a geometric model is generated in real time, and the measurement information of the stress state, temperature, microstructure, and element content is synchronously input into the geometric model, and then a digital twin of the molten pool and the deposition layer is formed online.

[0011] Further, the specific process of generating the geometric model in real time according to the size and morphology of the molten pool and the size and morphology of the deposition layer measured in real time is as follows: The side monitoring CCD takes pictures from the side of the molten pool to obtain a side view of the molten pool; the coaxial monitoring CCD takes coaxial pictures from above the molten pool to obtain a top view of the molten pool; the size of the molten pool in different views is obtained from the surface images of the molten pool taken, and the three-dimensional geometric size of the surface of the molten pool is synthesized to generate a three-dimensional model of the molten pool; the line laser emitted by the line laser generator irradiates on the surface of the deposited and solidified deposition layer, and the line laser monitoring CCD collects the contour of the line laser to obtain the geometric morphology of the deposition layer and generate a three-dimensional model of the deposition layer.

[0012] Furthermore, the specific process of the stress state measurement is as follows: During the laser additive manufacturing process, a digital twin calculation domain (finite element geometric model) of the molten pool is formed based on the three-dimensional model of the molten pool, as well as the height of the molten pool and the height of the deposited layer after cooling. By comparing the geometric dimensions of the molten pool morphology and the deposited layer morphology, the surface displacement and volume shrinkage values generated when the material shrinks from the molten state to the solidified state are obtained. The surface displacement and volume shrinkage values are used as loads and applied to the real-time generated digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling shrinkage of the molten pool is calculated; The volume change generated by the shrinkage of the remelted molten pool is measured by ultrasound and used as a load and applied to the digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling shrinkage of the remelted molten pool is calculated; Further, the stress generated by the shrinkage of the molten pool and the stress generated by the shrinkage of the remelted molten pool are superimposed to obtain the stress state of the deposited layer. During the manufacturing process, as the laser heat source advances, the above calculation process is repeated for the formation of each new molten pool, and the calculation results of each step are iterated, thereby reflecting the stress change inside the deposited layer in real time.

[0013] Furthermore, the monitoring of the microstructure and element content is realized by a laser-induced spectrometer. First, before online detection, plasma-excited spectra are generated by laser scanning pure metal and the spectra are measured using a spectrometer to calibrate the spectral lines of the elements and establish a database of plasma spectral maps for different elements. During the laser additive manufacturing process, the plasma-excited spectra induced by the high-energy laser beam are synchronously collected by the spectrometer. The element types at the test points are determined by analyzing the peak wavelengths of the spectral lines, and the element distribution inside the deposited layer is generated with time as the unit and in combination with the scanning speed and imported into the three-dimensional model of the deposited layer; The method for microstructure monitoring is as follows: Before the test, alloy specimens with different microstructures are scanned by laser, and the peak values of the plasma-excited spectra, the wavelengths of the element spectral lines, the full width at half maximum, the peak profile area, and the overall pattern characteristics of the spectral lines are analyzed. The relationship between the plasma-excited spectral signals and the microstructures is established through a deep learning network. During the laser additive manufacturing process, the plasma-excited spectra induced by the high-energy laser beam are synchronously collected by the spectrometer. The spectral line characteristics are analyzed through the deep learning network to determine the microstructure at the test points, and the microstructure types inside the deposited layer are generated with time as the unit and in combination with the scanning speed and imported into the three-dimensional model of the deposited layer.

[0014] The beneficial effects of the present invention are as follows:

[0015] 1. The digital twin of the present invention is generated based on the real-time collected morphologies of the molten pool and the deposited layer, and corresponds one-to-one with the actual deposition process.

[0016] 2. The digital twin of laser additive manufacturing proposed by the present invention calculates stress through the surface displacement and volume change generated by the cooling and shrinkage of the molten pool, which is different from the traditional method of first calculating the temperature field and then calculating the stress field. The method proposed by the present invention greatly improves the calculation speed and can directly reflect the stress change inside the deposited layer.

[0017] 3. The monitoring method proposed by the present invention can not only capture the morphology of the surface molten pool, but also capture the morphology of the remelting molten pool, making the twin model more accurate.

[0018] 4. The method for constructing the digital twin of laser additive manufacturing proposed by the present invention monitors the element composition and microstructure through the plasma emission spectrum induced by the heat source laser, without the need for an additional plasma excitation source, with low cost and convenient operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is the isometric main view of the online monitoring system for laser additive manufacturing of the device of the present invention;

[0020] Figure 2 is the side roll view of the online monitoring system for laser additive manufacturing of the present invention;

[0021] Figure 3 is the schematic diagram of molten pool size reconstruction, Figure 3 in which (a) shows the side view of the surface molten pool on the remelting molten pool, (b) shows the top view of the surface molten pool, and (c) shows the three-dimensional model of the molten pool surface and the three-dimensional model of the remelting molten pool;

[0022] Figure 4 is the schematic diagram of the stress calculation method, Figure 4 in which (a) shows the height of the molten pool and the height of the deposited layer collected in real time, and the difference between the two can obtain the surface displacement generated by the cooling and shrinkage of the molten pool; (b) shows the three-dimensional model of the molten pool surface and the three-dimensional model of the remelting molten pool.

[0023] DESCRIPTION OF THE REFERENCE NUMERALS

[0024] 100 - laser head, 200 - hollow shaft motor, 201 - side monitoring CCD, 202 - line laser monitoring CCD, 203 - temperature sensor, 204 - high-precision spectrometer, 205 - coaxial monitoring CCD, 210 - line laser generator, 211 - line laser, 300 - deposited layer, 310 - molten pool, 311 - side view of the surface molten pool, 313 - top view of the surface molten pool, 314 - remelting molten pool, 315 - three-dimensional model of the molten pool surface, 316 - three-dimensional model of the remelting molten pool, 320 - height of the molten pool, 321 - height of the deposited layer after cooling, 400 - substrate, 500 - ultrasonic measuring instrument. DETAILED DESCRIPTION OF THE INVENTION

[0025] The present invention will be further illustrated below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and not to limit the scope of the present invention.

[0026] As Figure 1-2 shown, an online monitoring system for laser additive manufacturing in this embodiment includes a laser head 100, a side monitoring CCD 201, a coaxial monitoring CCD 205, a line laser generator 210, a line laser monitoring CCD 202, a temperature sensor 203, a high-precision spectrometer 204, an ultrasonic detector 500, a hollow shaft motor 200, and a high-performance computer. It is characterized in that: the hollow shaft motor 200 is installed above the nozzle of the laser head 100, its inner ring is matched with the laser head 100, and the outer ring can rotate circumferentially. The side monitoring CCD, coaxial monitoring CCD, line laser generator 210, line laser monitoring CCD, temperature sensor, and high-precision spectrometer are all installed on the outer ring of the hollow shaft motor 200. Among them, the line laser generator 210 and the line laser monitoring CCD are symmetrically installed, and the axes of the two are in the same plane, so that the line laser monitoring CCD can conveniently capture the laser generated by the line laser generator 210. The side monitoring CCD is installed on the outer ring of the hollow shaft motor 200 and is arranged at 90 degrees to the line laser generator 210 and the line laser monitoring CCD. The temperature sensor and the high-precision spectrometer are installed on the outer ring of the hollow shaft motor 200 and are aligned with the test points on the deposition layer 300. The ultrasonic detector is located on the back of the substrate and is used to detect the three-dimensional morphology of the remelting molten pool 314.

[0027] During the manufacturing process, the hollow shaft motor 200 can drive devices such as the side monitoring CCD 201, coaxial monitoring CCD 205, line laser generator 210, line laser monitoring CCD 202, temperature sensor 203, high-precision spectrometer 204, and ultrasonic detector 500 to rotate around the laser head 100, adjust the monitoring angle of the sensor according to different scanning directions, and ensure that the line laser 211 is always perpendicular to the deposition layer 300. The high-performance computer is used to receive and process the signals monitored by various sensors, perform calculations, analysis, and integration, and generate a digital twin of the deposition layer 300 in real time.

[0028] A method for constructing a component digital twin using the above online monitoring system for laser additive manufacturing includes the following steps:

[0029] (1) The laser is emitted from the laser head 100, melts the powder conveyed coaxially, and forms a molten pool 310 on the substrate;

[0030] (2) The side monitoring CCD 201 captures the molten pool 310 from the side to obtain a side view 311 of the surface molten pool. The coaxial monitoring CCD 205 captures the molten pool 310 from above to obtain a top view 313 of the surface molten pool. The ultrasonic measuring instrument 500 detects the morphology of the remelting molten pool 314 from the back of the substrate 400 to obtain the morphology of the remelting molten pool 314 below the surface of the substrate 400;

[0031] (3) The high-performance computer constructs a three-dimensional model of the molten pool 310 based on the data measured by the side monitoring CCD 201, the coaxial monitoring CCD 205, and the ultrasonic measuring instrument 500, including a three-dimensional model 315 of the molten pool surface and a three-dimensional model 316 of the remelting molten pool;

[0032] (4) Simultaneously with step (3), the line laser generator 210 emits a line laser 211, which irradiates behind the molten pool 310. The line laser monitoring CCD collects the contour of the line laser 211 and transmits the data to the high-performance computer to construct a three-dimensional model of the deposition layer 300;

[0033] (5) The high-performance computer calculates the volume and surface displacement changes generated during the process of the deposited material cooling and solidifying and shrinking from the molten pool 310 to the deposition layer 300 based on the size differences between the three-dimensional model of the molten pool 310 and the three-dimensional model of the deposition layer 300, and loads it as a load onto the generated three-dimensional model of the molten pool, calculates the stress generated by the volume shrinkage of the material, and records it inside the model.

[0034] The monitoring of the size and shape of the deposition layer 300 includes the monitoring of the morphology of the molten pool 310 and the monitoring of the morphology of the deposition layer 300.

[0035] The monitoring of the morphology of the molten pool is divided into the monitoring of the morphology of the surface of the molten pool 310 and the monitoring of the morphology of the remelting molten pool 314. The monitoring of the surface morphology of the molten pool 310 measures the size of the molten pool 3B10 through 2 CCDs. The side monitoring CCD 201 captures the molten pool 310 from the side to obtain a side view 311 of the surface molten pool. The coaxial monitoring CCD 205 captures the molten pool 310 from above to obtain a top view 313 of the surface molten pool; the surface molten pool sizes in different views are obtained from the captured surface images of the molten pool 310, and the three-dimensional geometric size of the surface of the molten pool 310 is synthesized to generate a three-dimensional model 315.

[0036] The monitoring of the morphology of the remelting molten pool 314 uses the ultrasonic detection method. The morphology of the remelting molten pool 314 refers to the molten pool part formed below the surface of the material during the laser additive manufacturing process when the surface of the substrate 400 or the surface of the previous deposition layer 300 is melted by the laser. The three-dimensional geometric size of the laser remelting molten pool 314 is detected by the ultrasonic probe set at the bottom of the additive manufacturing substrate 400 to obtain the three-dimensional geometric size of the laser remelting molten pool 314 and generate a three-dimensional model.

[0037] The morphology of the deposited layer 300 can be monitored by the line laser method. The line laser 211 is irradiated on the surface of the already deposited and solidified deposited layer 300, preferably at a distance of about 3 mm from the molten pool. The contour of the line laser 211 is collected by the line laser monitoring CCD 202 to obtain the geometric morphology of the deposited layer 300 and generate a three-dimensional model.

[0038] The measurement of the stress state is achieved by calculating the displacement difference between the molten pool 310 and the surface of the deposited layer 300. The specific method is as follows: During the laser additive manufacturing process, a three-dimensional geometric model of the molten pool 310 is established in real time based on the surface morphology of the molten pool 310 and the morphology of the remelted molten pool 314 to form a digital twin calculation domain. Since the deposited layer 300 is formed by the cooling and shrinkage of the molten pool 310, by comparing the geometric dimensions of the morphology of the molten pool 310 and the deposited layer 300, the surface displacement and volume shrinkage values generated when the material shrinks from the molten state to the solidified state can be obtained. The surface displacement and volume shrinkage values are used as loads and loaded onto the real-time generated digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling and shrinkage of the molten pool 310 is calculated. Further, the volume change generated by the shrinkage of the remelted molten pool 314 is measured by ultrasonic and used as a load and loaded onto the digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling and shrinkage of the remelted molten pool 314 is calculated. The stress generated by the shrinkage of the surface molten pool and the stress generated by the shrinkage of the remelted molten pool 314 are superimposed to obtain the stress state of the deposited layer. During the manufacturing process, as the laser heat source advances, the above calculation process is repeated for the formation of each new molten pool, and the calculation results of each step are iterated, thereby reflecting the stress change inside the deposited layer 300 in real time.

[0039] As Figure 4 shown, a three-dimensional model of the molten pool 310, the molten pool height 320, and the height 321 of the deposited layer after cooling can be obtained through the line laser monitoring CCD 202, the side monitoring CCD 201, and the coaxial monitoring CCD 2*5. The difference between the molten pool height 320 and the height 321 of the deposited layer after cooling is obtained, and the displacement difference generated during the cooling and solidification of the molten pool 310 is used as the surface displacement load and loaded onto the three-dimensional model 315 of the molten pool surface and the three-dimensional model 316 of the remelted molten pool, and the stress generated by the cooling and shrinkage of the molten pool *310 is calculated.

[0040] The monitoring of element content and microstructure is realized by laser-induced plasma spectroscopy. First, before online detection, plasma excitation spectra are generated by laser scanning pure metal, and the spectra are measured using a high-precision spectrometer 204. The spectral lines of elements are calibrated, and a database of plasma spectral maps for different elements is established. During the laser additive manufacturing process, the plasma excitation spectra induced by a high-energy laser beam are synchronously collected by the high-resolution spectrometer 204. The element types at the test points are determined by analyzing the peak wavelengths of the spectral lines, and the element distribution inside the deposition layer 300 is generated with time as the unit and in coordination with the scanning speed and imported into the three-dimensional model of the deposition layer 300.

[0041] The method for microstructure monitoring is as follows: Before testing, alloy specimens with different microstructures are scanned by laser, and the characteristics such as the peak values of the plasma excitation spectra, element spectral line wavelengths, full width at half maximum, peak profile area, and overall spectral line patterns are analyzed. The relationship between the plasma excitation spectral signals and the microstructures is established through a deep learning network. During the laser additive manufacturing process, the plasma excitation spectra induced by a high-energy laser beam are synchronously collected by the high-resolution spectrometer. The microstructure at the test points is determined by analyzing the spectral line characteristics through the deep learning network, and the microstructure types inside the deposition layer 300 are generated with time as the unit and in coordination with the scanning speed and imported into the three-dimensional model of the deposition layer 300.

[0042] According to the size and morphology of the molten pool 310 and the size and morphology of the deposition layer 300 measured in real time, a geometric model is generated in real time, and information such as stress, temperature, microstructure, and element content is synchronously input into the geometric model, thereby forming a digital twin of the molten pool 310 and the deposition layer 300 online.

[0043] The above is only the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. An online monitoring system for laser additive manufacturing, comprising a substrate and a laser head for emitting laser. The laser head emits laser to melt the powder conveyed coaxially, forming a molten pool on the substrate, and a deposition layer is formed after the molten pool cools; it is characterized in that, Above the nozzle of the laser head, a hollow shaft motor is installed. The inner ring of the hollow shaft motor is fitted with the laser head. On the outer ring of the hollow shaft motor, a detection system for real-time measuring the size and morphology of the molten pool and the size and morphology of the deposition layer, a temperature sensor, and a spectrometer are installed. On the back of the substrate, a remelting molten pool morphology monitoring sensor is provided. A computer is also included. By taking the difference between the height of the molten pool and the height of the deposition layer after cooling, the displacement difference generated during the cooling and solidification of the molten pool is obtained and used as the surface displacement load to be loaded onto the three-dimensional model of the molten pool surface and the three-dimensional model of the remelting molten pool, and the stress generated by the cooling shrinkage of the molten pool is calculated. The detection system for real-time measuring the size and morphology of the molten pool and the size and morphology of the deposition layer includes a side monitoring CCD, a coaxial monitoring CCD, a line laser generator, and a line laser monitoring CCD. Among them, the line laser generator and the line laser monitoring CCD are symmetrically installed, and their axes are in the same plane. The side monitoring CCD is arranged at 90 degrees to the line laser generator and the line laser monitoring CCD. The specific process of generating a geometric model in real time from the size and morphology of the molten pool and the size and morphology of the deposition layer measured in real time is as follows: The side monitoring CCD takes pictures from the side of the molten pool to obtain a side view of the molten pool; the coaxial monitoring CCD takes coaxial pictures from above the molten pool to obtain a top view of the molten pool; the size of the molten pool in different views is obtained from the surface image of the molten pool taken, and the three-dimensional geometric size of the molten pool surface is synthesized to generate a three-dimensional model of the molten pool; the line laser emitted by the line laser generator irradiates on the surface of the solidified deposition layer that has been deposited, and the line laser monitoring CCD collects the contour of the line laser to obtain the geometric morphology of the deposition layer and generate a three-dimensional model of the deposition layer.

2. The online monitoring system for laser additive manufacturing according to claim 1, wherein The temperature sensor and the spectrometer are installed on the outer ring of the hollow shaft motor and aligned with the test points on the deposition layer.

3. The on-line monitoring system for laser additive manufacturing according to claim 1, characterized in that, The remelting molten pool morphology monitoring sensor uses an ultrasonic detector and is located on the back of the substrate for detecting the three-dimensional morphology of the remelting molten pool.

4. A method for constructing a digital twin of a component using the online monitoring system for laser additive manufacturing according to any one of claims 1-3, characterized in that, The method is as follows: A geometric model is generated in real time according to the size and morphology of the molten pool and the size and morphology of the deposition layer measured in real time, and the measurement information of the stress state, temperature, microstructure, and element content is synchronously input into the geometric model, thereby forming a digital twin of the molten pool and the deposition layer online.

5. The method for constructing a digital twin of a building component according to claim 4, characterized in that, The specific process of the stress state measurement is as follows: During the laser additive manufacturing process, a digital twin calculation domain of the molten pool is formed based on the three-dimensional model of the molten pool, as well as the height of the molten pool and the height of the deposited layer after cooling. By comparing the geometric dimensions of the molten pool morphology and the deposited layer morphology, the surface displacement and volume shrinkage values generated by the material shrinking from the molten state to the solidified state are obtained. The surface displacement and volume shrinkage values are used as loads and applied to the real-time generated digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling shrinkage of the molten pool is calculated; the volume change generated by the shrinkage of the remelted molten pool is measured by ultrasonic and used as a load and applied to the digital twin calculation domain of the molten pool, and based on this, the stress value generated by the cooling shrinkage of the remelted molten pool is calculated; further, the stress generated by the shrinkage of the molten pool and the stress generated by the shrinkage of the remelted molten pool are superimposed to obtain the stress state of the deposited layer. During the manufacturing process, as the laser heat source advances, the above calculation process is repeated for the formation of each new molten pool, and the calculation results of each step are iterated, thereby reflecting the stress change inside the deposited layer in real time.

6. The method for constructing a digital twin of a building component according to claim 4, characterized in that, The monitoring of the microstructure and element content is realized by a laser-induced spectrometer. First, before the on-line detection, plasma-excited spectra are generated by laser scanning pure metal and the spectra are measured by a spectrometer to calibrate the spectral lines of the elements and establish a database of plasma spectral maps of different elements. During the laser additive manufacturing process, the plasma-excited spectra induced by the high-energy laser beam are synchronously collected by the spectrometer. The element types at the test points are determined by analyzing the peak wavelengths of the spectral lines, and the element distribution inside the deposited layer is generated with time as the unit and in combination with the scanning speed and imported into the three-dimensional model of the deposited layer; the method for microstructure monitoring is as follows: Before the test, alloy specimens with different microstructures are scanned by laser, and the peak values of the plasma-excited spectra, the wavelengths of the element spectral lines, the full width at half maximum, the peak profile area, and the overall pattern characteristics of the spectral lines are analyzed. The relationship between the plasma-excited spectral signals and the microstructures is established through a deep learning network. During the laser additive manufacturing process, the plasma-excited spectra induced by the high-energy laser beam are synchronously collected by the spectrometer. The spectral line characteristics are analyzed through the deep learning network to determine the microstructure at the test points, and the microstructure types inside the deposited layer are generated with time as the unit and in combination with the scanning speed and imported into the three-dimensional model of the deposited layer.

Images