CN116809975B

CN116809975B - Device and method for undistorted online monitoring system of selective laser melting pool

Info

Publication number: CN116809975B
Application number: CN202311092228.3A
Authority: CN
Inventors: 王迪; 黄焱明; 蒋仁武; 王岩; 于佳明; 吕佳乐
Original assignee: South China University of Technology SCUT
Current assignee: Guangdong Leijia Additive Technology Co ltd
Priority date: 2023-08-29
Filing date: 2023-08-29
Publication date: 2023-12-05
Anticipated expiration: 2043-08-29
Also published as: CN116809975A

Abstract

The invention provides a device and a method for an undistorted online monitoring system of a selective laser melting pool, and relates to the technical field of selective laser melting forming detection. The device comprises an image acquisition mechanism, a radiation signal acquisition mechanism, a dichroic mirror mechanism, a laser, a scanning galvanometer and a processor which are arranged on a selective melting chamber platform; the dichroic mirror mechanism is respectively connected with the image acquisition mechanism and the radiation signal acquisition mechanism; the image acquisition mechanism is vertical to the selective melting chamber platform; the radiation collection line of the radiation signal collection mechanism is parallel to the selective melting chamber platform, and the light inlet hole of the radiation signal collection mechanism is communicated with the second light outlet hole of the dichroic mirror; the laser and the scanning galvanometer are oppositely arranged at two sides of the dichroic mirror mechanism; the processor is respectively connected with the image acquisition mechanism and the radiation signal acquisition mechanism. The defects are accurately judged through the comparison and verification of the data acquired by the photodiode and the infrared imager, and the situation of data distortion is avoided.

Description

Device and method for undistorted online monitoring system of selective laser melting pool

Technical Field

The invention relates to the technical field of selective laser melting forming detection, in particular to a device and a method for a selective laser melting pool undistorted on-line monitoring system.

Background

The forming process of the selective laser melting technology is influenced by a plurality of factors such as materials, laser and light path systems, scanning strategies, external environments, machinery, geometric characteristics and the like, and defects such as cracks, holes, unfused and the like are often generated in the forming process under the influence of the factors. These defects affect the mechanical properties of the selective laser melting molded part, so that the repeatability of the selective laser melting molding quality is greatly reduced. The detection of defects caused by the factors is an important means for improving the quality of selective laser melting forming, and the on-line detection and feedback control can reduce the post-treatment procedures, accelerate the production efficiency and save the production cost of parts.

At present, nondestructive detection and destructive detection are mainly used for detecting selective laser melting defects, and the nondestructive detection is mainly used in the current academic research and industrial engineering because the nondestructive detection has small influence on the part forming process and the final quality. The common detection method is to collect by using a CCD camera or a thermal infrared imager and a photodiode optical splitting path, and although the method can detect a certain defect, the optical path is divided into two parts, so that the condition of light signal energy reduction can be caused, and the detection result can possibly cause distortion.

Therefore, the device and the method for designing the non-distortion online monitoring system of the selective laser melting pool accurately judge the defects by comparing and verifying the data acquired by the photodiode and the infrared imager, avoid the situation of data distortion and solve the problem to be solved urgently at present.

Disclosure of Invention

The invention aims to provide a device of a non-distorted online monitoring system of a selective laser melting molten pool, which is used for acquiring thermal image data of the molten pool in a printing area by arranging an image acquisition mechanism, acquiring radiation signal data of the printing area by arranging a radiation signal acquisition mechanism, completing alternate data acquisition by angle adjustment of a dichroic mirror mechanism by the two data acquisition mechanisms, further acquiring two different data acquisition results when each layer of printing data acquisition is carried out, and finally completing data processing and comparison analysis of the two data acquisition results in a processor to judge a defect area. The whole device provides a structural basis for analyzing and comparing thermal imaging data and diode photoelectric data, ensures that analysis and processing of defect data can be performed stably, further realizes accurate judgment of defects in the printing process, and effectively avoids the condition that the judgment accuracy is reduced due to data distortion.

The invention also aims to provide a method for the non-distorted online monitoring system of the selective laser melting pool, which is characterized in that the acquisition of photoelectric data is carried out before each layer of printing to analyze defect data once, then the acquisition and analysis of thermal imaging data are carried out in the printing process to acquire the acquisition and analysis results of different types of data in each layer of printing process, and then the mutual comparison and verification of the analysis results are carried out to finish the accurate judgment of defects, thereby avoiding the situation of error judgment of defects caused by data distortion due to objective and subjective reasons of single type of data acquisition and greatly improving the accuracy and instantaneity of online detection of SLM defects.

In a first aspect, the invention provides a device of an online monitoring system for undistorted molten pool of selective laser melting, which comprises an image acquisition mechanism, a radiation signal acquisition mechanism, a dichroic mirror mechanism, a laser, a scanning galvanometer and a processor, wherein the image acquisition mechanism, the radiation signal acquisition mechanism, the dichroic mirror mechanism, the laser, the scanning galvanometer and the processor are arranged on a selective melting chamber platform; the dichroic mirror mechanism is respectively connected with the image acquisition mechanism and the radiation signal acquisition mechanism; the light ray acquisition line of the image acquisition mechanism is vertical to the selective melting chamber platform, and the light inlet hole of the image acquisition mechanism is communicated with the first light outlet hole of the dichroic mirror; the radiation collection line of the radiation signal collection mechanism is parallel to the selective melting chamber platform, and the light inlet hole of the radiation signal collection mechanism is communicated with the second light outlet hole of the dichroic mirror; the laser and the scanning galvanometer are oppositely arranged at two sides of the dichroic mirror mechanism; the processor is respectively connected with the image acquisition mechanism and the radiation signal acquisition mechanism; the dichroic mirror mechanism reflects a signal reflected by laser light emitted by the scanning laser of the scanning galvanometer through angle adjustment, so that the reflected signal is sent to the image acquisition mechanism through the first light outlet hole or is sent to the radiation signal acquisition mechanism through the second light outlet hole.

As one possible implementation, the image acquisition mechanism includes a support connection box, a mirror support frame, a sealing cover, an infrared imager, a mirror frame, and a mirror; the support connecting box, the reflector supporting frame and the sealing cover are sequentially connected, one end, far away from the reflector supporting frame, of the support connecting box is arranged on a dichroic mirror frame of the dichroic mirror mechanism, and the reflector frame is positioned in the reflector supporting frame and is respectively connected with the reflector supporting frame and the sealing cover; the reflector is arranged on the reflector frame; the infrared imaging instrument is arranged on one side of the reflector support frame, and the position of the light inlet of the infrared imaging instrument is matched with the position of the reflector; the reflector is positioned on a light collecting line communicated with the first light outlet; the infrared imaging instrument is connected with the processor; the position of the first light exit aperture of the dichroic mirror mechanism is adapted to the position of the mirror.

In the invention, a concrete composition of an image acquisition mechanism is provided, a supporting connection box, a reflector support frame and a sealing cover are mutually connected to form a sealed cavity structure, and a thermal imaging signal reflected by a dichroic mirror linearly propagates in the cavity and enters an infrared imager to be acquired through reflection of the reflector, so that the collection of thermal imaging data is completed.

As one possible implementation manner, the radiation signal acquisition mechanism comprises an acquisition card, a spectroscope frame, a first optical filter, a first convex lens, a first photodiode, a second optical filter, a second convex lens and a second photodiode, wherein the spectroscope, the spectroscope frame, the first optical filter, the first convex lens, the first photodiode, the second optical filter, the second convex lens and the second photodiode are arranged in the photodiode frame; a diode light inlet hole is formed in the photodiode mirror bracket; the position of the beam-splitting light inlet of the beam splitter mirror frame is matched with the position of the diode light inlet; the first convex lens, the first optical filter and the first photodiode are sequentially arranged at intervals, and the first convex lens is close to the spectroscope frame; the position of the first light splitting and emitting hole of the spectroscope frame is matched with the position of the first convex lens; the second convex lens, the second optical filter and the second photodiode are sequentially arranged at intervals, and the second convex lens is close to the spectroscope frame; the position of the second light splitting and emitting hole of the spectroscope frame is matched with the position of the second convex lens; the acquisition card is respectively connected with the first photodiode, the second photodiode and the processor; the position of the diode light inlet hole is communicated with the second light outlet hole of the dichroic mirror mechanism.

The invention provides a specific composition of a radiation signal acquisition mechanism, which is characterized in that a radiation signal reflected by a dichroic mirror is split by a spectroscope, the split signals are respectively acquired by different photodiodes through different filtering and focusing processes, and finally the acquired signals are gathered by an acquisition card and then are transmitted to a processor for unified analysis and processing. In the invention, a first photodiode and a first optical filter mainly collect optical radiation signals of a wave band of about 800nm, a second photodiode and a second optical filter mainly collect optical radiation signals of a wave band of about 900nm, and finally, the two optical radiation signals are converted into current change signals in the photodiode, and then, the change data of the current is transmitted to a processor in the form of a voltage value by an acquisition card for processing analysis.

In a second aspect, the invention provides a method for an online monitoring system of undistorted selective laser melting pool, which is applied to the device of the online monitoring system of undistorted selective laser melting pool, and comprises the steps of setting a layer number threshold value n, acquiring printing layer number information, and judging the printing layer number m: if m is less than n, continuing to acquire the number of printing layers to judge; if m is more than or equal to n, forming an online detection instruction, and finishing the following detection steps: acquiring optical radiation signals of m layers, and completing first signal data processing to form first processing data; stopping acquiring an optical radiation signal when the powder of the m+1 layer is paved, and starting to acquire an infrared thermal imaging signal until the printing of the m+1 layer is completed; performing second signal data processing on the infrared thermal imaging signals to form second processing data; performing defect detection analysis on the first processing data and the second processing data to form defect detection result data; repeating the detecting step until printing is completed.

As a possible implementation manner, acquiring an optical radiation signal of m layers, and completing first signal data processing to form first processing data, including: first dividing the optical radiation signal Frame processing to form first frame processing data; acquiring short-time average amplitude of each frame in the first framing data to form a first short-time average amplitude setI is the total frame number in the first frame processing data; detecting abnormal outliers of the short-time average amplitude set A to form a first abnormal outlier data set; and combining the first abnormal outlier data set and index position coordinates corresponding to the first abnormal outlier in the first abnormal outlier data set to form first processing data B.

In the invention, the radiation signals collected by the photodiodes are stored in a transient state after being transmitted into a computer to form a data set. Then, a first framing process is performed, wherein the framing process uses a framing function enframe to perform framing pretreatment, and a rectangular function is selected as a window function. Specifically, let the displacement of the previous frame function relative to the next frame function be δx, the overlap be overlap, the frame rate be fs, and the length of the data set x be L. The overlap is introduced to make the characteristic parameter change smoother. Then the frame rate calculation formula is:

. Setting the frame rate of an infrared camera of the infrared imager as f, taking the framing frame rate fs=f and overlap=2×δx for better correspondence with a thermal image shot by the infrared imager, substituting the framing frame rate fs=f and the overlap=2×δx into a frame rate calculation formula to obtain δx, and further completing framing based on displacement δx. And measuring the data fluctuation condition of each frame by utilizing the short-time average amplitude for the first frame processing data after framing, namely obtaining a first short-time average amplitude set A corresponding to each frame. And judging the abnormal interest group point analysis by adopting an anomaly detection algorithm LOF, searching frame data with abnormal short-time average amplitude value change, temporarily marking the frame data as abnormal data, and determining the corresponding index coordinate position to form first processing data.

As one possible implementation manner, performing second signal data processing on the infrared thermal imaging signal to form second processing data, including: calculating the molten pool area according to the frame rate on the second signal data to form a frame rate molten pool area data set; detecting abnormal outliers of the frame rate molten pool area data set to form a second abnormal outlier data set; and combining the second abnormal outlier data set and index coordinates corresponding to the second abnormal outlier in the second abnormal outlier data set to form second processing data C.

In the invention, the infrared thermal imaging signal is processed by framing according to the frame rate of an infrared camera, and then the molten pool area of each frame is calculated to form a frame rate molten pool area data set. And judging the abnormal data in the frame rate molten pool area data set by using an abnormal detection algorithm LOF, and determining index coordinates determined by frames corresponding to the abnormal data after determining the abnormal data to form second processing data.

As one possible implementation manner, performing defect detection analysis on the first processing data and the second processing data to form defect detection result data, including: determining an index position reference point, and performing grouping analysis on the first processing data B based on the index position reference point to form a first processing grouping data set D; performing grouping analysis on the second processing data C based on the index position reference points to form a second processing grouping data set E; and setting a position matching threshold gamma, and carrying out matching judgment on the data group in the first processing packet data set D and the data group in the second processing packet data set E to form defect detection result data.

The invention discloses a method for analyzing the first processing data and the second processing data in a multiple comparison mode, which comprises the steps of carrying out grouping analysis based on distance by establishing index position reference points with relatively fixed index coordinates, completing grouping of possible abnormal data at different positions, and accurately determining defects by comparing and judging the abnormal data groups obtained in two ways according to the grouping. It should be noted that, in order to avoid that the selected index position reference point is located at an abnormal position, the index position reference point may be determined to be outside the printing area. Meanwhile, the index position reference point is a common reference point for grouping abnormal data obtained in two modes, and the uniformity of data analysis can be ensured.

As a possible implementation manner, determining the index position reference point, performing packet analysis based on the index position reference point on the first processing data B to form a first processing packet data set D, including: determining index position reference pointsAcquiring index coordinates corresponding to each first abnormal outlier in the first processing data B, and arranging the index coordinates according to a framing sequence to form a first index coordinate sequence set ∈ ->U is the number of the first abnormal outlier determined according to the framing sequence; determining the first index coordinate order set >Is +.>First index distance between->Forming a first index distance set; according to the arrangement sequence of the first index coordinates in the first index coordinate sequence set, sequentially extracting the first 1+v first index distances in the first index distance set according to the rule that one first index distance is obtained every time the first index distances are increased, and carrying out first average distance->Is calculated by (1): />Wherein v is more than or equal to 2 and less than or equal to u; setting a first grouping distance threshold alpha, when +.>When the first abnormal index coordinates corresponding to the first v first index distances meeting the condition are determined as a first abnormal index coordinate set, and the first abnormal index coordinates are determined from 1+v th first abnormal index coordinatesStarting the index distance, continuing to calculate a first average distance, and performing grouping judgment according to a first grouping distance threshold alpha; for each first abnormality index coordinate set, determining a relative index position reference point +.>Is used as a reference distance to determine a reference point of relative index position>Establishing a first range area of the first abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each first index coordinate in the first abnormal index coordinate set as the radius of the first index range; combining all the first abnormal index coordinate sets and the corresponding first range areas to form a first processing grouping data set +. >W represents the total number of the first abnormality index coordinate group, +.>Representing a corresponding first range region.

In the invention, the first processing data is subjected to grouping analysis, and firstly grouping is divided based on the framing sequence, wherein the main basis of the division is to sequentially expand abnormal index coordinates to judge the average distance of relative index position reference points, and the threshold value of the first grouping distance can be obtained based on big data analysis so as to improve the accuracy of threshold value judgment. It will be appreciated that if the abnormal index coordinates are two points of different defect locations, the average distance of the reference points to the index locations will be significantly different from the value when the index coordinates of the same defect location are obtained. Meanwhile, because the analysis is carried out according to the framing sequence, the positions of the index coordinates are continuously adjacent, namely when index coordinates of different defect positions exist, the group formed by the previous index coordinates is necessarily the index coordinates positioned under the same defect, and the grouping accuracy is fully ensured. After the group is determined, the defect area defined by the index coordinates in the group needs to be expressed, so that a clear range area can be defined in the subsequent comparison analysis conveniently, an equivalent index coordinate is determined through an average distance and is taken as an ideal center of the defect, a circular area is defined by taking the center as a reference and taking the radius of the distance of the center relative to the farthest index coordinate in the group, the circular area can fully contain the defect area, and the index coordinates outside in the whole circumferential direction can be connected with the center as a reference to form an irregular shape so as to define the defect area more accurately.

As a possible implementation manner, performing packet analysis based on the index position reference point on the second processing data C to form a second processing packet data set E includes: acquiring index coordinates corresponding to each second abnormal outlier in the second processing data C, and arranging according to the framing sequence to form a second index coordinate sequence setP is the number of the second abnormal outlier determined according to the framing sequence; determining a second index coordinate order set respectively +.>Is added with index position reference point +.>Second index distance between->Forming a second index distance set; sequentially extracting the first 1+q second index distances in the second index distance set according to the arrangement sequence of the second index coordinates in the second index coordinate sequence set and taking one second index distance obtained every time as a rule, and carrying out second average distanceIs calculated by (1): />Wherein q is more than or equal to 2 and less than or equal to p; setting a second packet distance threshold beta, when +.>Determining the second abnormal index coordinates corresponding to the first q second index distances meeting the conditions as a second abnormal index coordinate set, starting from the 1+q second index distance, continuing to calculate a second average distance, and performing grouping judgment according to a second grouping distance threshold beta; for each second abnormal index coordinate set, determining relative index position reference point +. >Is used as a reference distance to determine a reference point of the relative index position>Establishing a second range region of the second abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each second index coordinate in the second abnormal index coordinate set as the radius of the second index range; combining all the second abnormal index coordinate sets and the corresponding second range areas to form a second processing grouping data setR represents the total number of first abnormality index coordinate groups, +.>Representing a corresponding second range region.

In the present invention, the packet analysis method for the second processed data is the same as the packet analysis method for the first processed data. The determination is also made in frame order based on a distance determination of the second packet distance threshold. After grouping, the center position determined according to the average distance defines a reasonable range to form a investigation range of defect analysis.

As a possible implementation manner, setting a position matching threshold γ, performing matching judgment on the data set in the first processing packet data set D and the data set in the second processing packet data set E, to form defect detection result data, including: extracting each second range area from the second processing packet data set E in turn, and calculating and judging the overlapping area of each second range area and all the first range areas in the first processing packet data set D: if the overlapping area is smaller than the position matching threshold gamma, determining that the corresponding first range area and the corresponding second range area are not matched; and if the overlapping area is not smaller than the position matching threshold gamma, determining that the corresponding first range area is matched with the second range area, and determining the position defined by the first range area and the second range area as a defect area.

In the invention, because the corresponding defect area ranges are formed in the analysis of the first processing data and the second processing data, when the defect area is confirmed, the judgment is directly carried out according to the coincidence degree of the defect areas acquired by the two data, after all, the higher the coincidence degree is, the higher the identity of the defect positions measured by the two data is, the inaccurate defect judgment caused by the distortion of a single data processing mode is eliminated, the accuracy of determining the defect positions can be improved, and after all, the possibility that the positions of the coincident areas are the true defect areas is higher.

The device and the method for the non-distortion on-line monitoring system of the selective laser melting pool have the beneficial effects that:

the device realizes the acquisition of the thermal image data of the molten pool in the printing area by arranging an image acquisition mechanism, realizes the acquisition of the radiation signal data in the printing area by arranging a radiation signal acquisition mechanism, completes alternate data acquisition by carrying out angle adjustment on a dichroic mirror of a dichroic mirror mechanism by two data acquisition mechanisms, further acquires two different data acquisition results when carrying out the printing data acquisition of each layer, and finally completes the processing and the comparison analysis of the data by the two data acquisition results in a processor so as to realize the judgment of the defect area. The whole device provides a structural basis for analyzing and comparing thermal imaging data and diode photoelectric data, ensures that analysis and processing of defect data can be performed stably, further realizes accurate judgment of defects in the printing process, and effectively avoids the condition that the judgment accuracy is reduced due to data distortion.

According to the method, the photoelectric data is collected before each layer of printing to analyze the defect data once, then the thermal imaging data is collected and analyzed in the printing process, the collection and analysis results of different types of data in each layer of printing process are obtained, further the mutual comparison and verification of the analysis results are carried out, the accurate judgment of the defects is completed, the defect judgment error caused by data distortion due to objective and subjective reasons in single type of data collection is avoided, and the accuracy and the instantaneity of SLM defect online detection are greatly improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments of the present invention will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and should not be considered as limiting the scope, and other related drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram of a device of an online monitoring system for undistorted molten pool of selective laser melting according to an embodiment of the invention;

FIG. 2 is a schematic structural diagram of an image acquisition mechanism of a device of an online monitoring system for undistorted molten pool laser melting in a selected area, which is provided by an embodiment of the invention;

FIG. 3 is a schematic structural diagram of a radiation signal acquisition mechanism of a device of an online monitoring system for undistorted molten pool laser melting in a selected area, which is provided by the embodiment of the invention;

FIG. 4 is a signal acquisition light path diagram of an image acquisition mechanism of a device of an online monitoring system for undistorted molten pool of selective laser melting, provided by an embodiment of the invention;

FIG. 5 is a signal acquisition light path diagram of a radiation signal acquisition mechanism of a device of an on-line monitoring system for undistorted molten pool of selective laser melting, provided by an embodiment of the invention;

FIG. 6 is a step diagram of a method for a selective laser melting pool undistorted on-line monitoring system, according to an embodiment of the invention.

Reference numerals:

01. an image acquisition mechanism; 11. supporting the connection box; 12. a mirror support; 13. sealing cover; 14. an infrared imager; 15. a mirror mount; 16. a reflecting mirror; 02. a radiation signal acquisition mechanism; 21. a collection card; 22. photodiode holders; 23. a beam splitter; 24. a spectroscope frame; 252. a first optical filter; 251. a first convex lens; 253. a first photodiode; 262. a second optical filter; 261. a second convex lens; 263. a second photodiode; 03. a dichroic mirror mechanism; 31. a dichroic mirror; 32. a dichroic mirror frame; 04. a laser; 05. scanning a vibrating mirror; 06. a processor; 07. a zone selection melting chamber platform.

Detailed Description

The technical solutions in the embodiments of the present invention will be described below with reference to the accompanying drawings in the embodiments of the present invention.

Referring to fig. 1-5, an embodiment of the present invention provides a device of an online monitoring system for undistorted molten pool of selective laser melting, which comprises an image acquisition mechanism 01, a radiation signal acquisition mechanism 02, a dichroic mirror mechanism 03, a laser 04, a scanning galvanometer 05 and a processor 06, which are arranged on a platform 07 of a selective melting chamber; the dichroic mirror mechanism 03 is respectively connected with the image acquisition mechanism 01 and the radiation signal acquisition mechanism 02; the light ray collection line of the image collection mechanism 01 is vertical to the selective melting chamber platform 07, and the light inlet hole of the image collection mechanism 01 is communicated with the first light outlet hole of the dichroic mirror 31; the radiation collection line of the radiation signal collection mechanism 02 is parallel to the selective melting chamber platform 07, and the light inlet hole of the radiation signal collection mechanism 02 is communicated with the second light outlet hole of the dichroic mirror 31; the laser 04 and the scanning galvanometer 05 are arranged on two sides of the dichroic mirror mechanism 03; the processor 06 is respectively connected with the image acquisition mechanism 01 and the radiation signal acquisition mechanism 02; the dichroic mirror mechanism 03 reflects a signal reflected by the laser light emitted by the scanning laser 04 by the scanning galvanometer 05 through angle adjustment, and transmits the reflected signal to the image pickup mechanism 01 through the first light exit hole or transmits the reflected signal to the radiation signal pickup mechanism 02 through the second light exit hole.

The device realizes the acquisition of the thermal image data of the molten pool in the printing area by arranging an image acquisition mechanism 01, realizes the acquisition of the radiation signal data in the printing area by arranging a radiation signal acquisition mechanism 02, completes alternate data acquisition by carrying out angle adjustment on a dichroic mirror 31 of a dichroic mirror mechanism 03 by two data acquisition mechanisms, further acquires two different data acquisition results when carrying out the printing data acquisition of each layer, and finally completes the processing and the comparison analysis of the data by the two data acquisition results in a processor 06, thereby realizing the judgment of the defect area. The whole device provides a structural basis for analyzing and comparing thermal imaging data and diode photoelectric data, ensures that analysis and processing of defect data can be performed stably, further realizes accurate judgment of defects in the printing process, and effectively avoids the condition that the judgment accuracy is reduced due to data distortion.

For the image acquisition mechanism 01, it includes a support connection box 11, a mirror support frame 12, a sealing cover 13, an infrared imager 14, a mirror frame 15, and a mirror 16; the supporting connection box 11, the reflector support frame 12 and the sealing cover 13 are sequentially connected, one end, far away from the reflector support frame 12, of the supporting connection box 11 is arranged on a dichroic mirror frame 32 of the dichroic mirror mechanism 03, and the reflector frame 15 is positioned in the reflector support frame 12 and is respectively connected with the reflector support frame 12 and the sealing cover 13; the mirror 16 is provided on the mirror frame 15; the infrared imager 14 is disposed on one side of the reflector support 12, and the position of the light inlet of the infrared imager 14 is adapted to the position of the reflector 16; the position of the reflecting mirror 16 is positioned on a light collecting line communicated with the first light outlet hole; infrared imager 14 is connected to processor 06; the position of the first light exit hole of the dichroic mirror mechanism 03 is adapted to the position of the reflecting mirror 16.

The supporting connection box 11, the reflector support frame 12 and the sealing cover 13 are mutually connected to form a sealed cavity structure, and the thermal imaging signals reflected by the dichroic mirror 31 linearly propagate in the cavity and enter the infrared imager 14 to be collected through reflection of the reflector 16, so that the collection of thermal imaging data is completed.

For the radiation signal collection mechanism 02, it includes a collection card 21 and a beam splitter 23, a beam splitter bracket 24, a first optical filter 252, a first convex lens 251, a first photodiode 253, a second optical filter 262, a second convex lens 261 and a second photodiode 263 disposed in a photodiode bracket 22; the photodiode mirror holder 22 is provided with a diode light inlet; the position of the beam splitting light inlet of the beam splitter bracket 24 is matched with the position of the diode light inlet; the first convex lens 251, the first optical filter 252 and the first photodiode 253 are sequentially arranged at intervals, and the first convex lens 251 is close to the spectroscope frame 24; the position of the first light-splitting exit hole of the spectroscope frame 24 is adapted to the position of the first convex lens 251; the second convex lens 261, the second optical filter 262 and the second photodiode 263 are sequentially arranged at intervals, and the second convex lens 261 is close to the spectroscope frame 24; the position of the second light-splitting exit aperture of the spectroscope frame 24 is adapted to the position of the second convex lens 261; the acquisition card 21 is respectively connected with the first photodiode 253, the second photodiode 263 and the processor 06; the position of the diode light inlet is communicated with the second light outlet of the dichroic mirror mechanism 03.

The radiation signal reflected by the dichroic mirror 31 is split by the spectroscope 23, the split signals are collected by different photodiodes through different filtering and focusing processes respectively, and finally the collected signals are collected by the collecting card 21 and then are sent to the processor 06 for unified analysis and processing. In the present invention, the first photodiode 253 and the first filter 252 mainly collect optical radiation signals of about 800nm wavelength band, the second photodiode 263 and the second filter 262 mainly collect optical radiation signals of about 900nm wavelength band, and finally both optical radiation signals are converted into current change signals in the photodiodes, and then the change data of the current is transmitted to the processor 06 in the form of voltage value by the acquisition card 21 for processing analysis.

Referring to fig. 6, an embodiment of the present invention provides a method for a selective laser melting bath undistorted on-line monitoring system. The invention provides a device applied to an online monitoring system for undistorted selective laser melting pool. According to the method, the photoelectric data is collected before each layer of printing to analyze the defect data once, then the thermal imaging data is collected and analyzed in the printing process, the collection and analysis results of different types of data in each layer of printing process are obtained, further the mutual comparison and verification of the analysis results are carried out, the accurate judgment of the defects is completed, the defect judgment error caused by data distortion due to objective and subjective reasons in single type of data collection is avoided, and the accuracy and the instantaneity of SLM defect online detection are greatly improved.

The method for the non-distortion on-line monitoring system of the selective laser melting pool specifically comprises the following steps:

setting a layer number threshold value n, acquiring printing layer number information, and judging the printing layer number m: if m is less than n, continuing to acquire the number of printing layers to judge; if m is more than or equal to n, forming an online detection instruction, and finishing the following detection steps:

the number of layers threshold n is determined according to practical materials, printing parameters and other factors, so that the printing quality is reasonably detected and controlled.

S1: and acquiring optical radiation signals of m layers, and completing first signal data processing to form first processing data.

Acquiring optical radiation signals of m layers, completing first signal data processing, forming first processing data, including: performing first framing processing on the optical radiation signal to form first framing processing data; acquiring short-time average amplitude of each frame in the first framing data to form a first short-time average amplitude setI is the total frame number in the first frame processing data; detecting abnormal outliers of the short-time average amplitude set A to form a first abnormal outlier data set; and combining the first abnormal outlier data set and index position coordinates corresponding to the first abnormal outlier in the first abnormal outlier data set to form first processing data B.

The radiation signals collected by the photodiodes are stored in a transient state after being transmitted into a computer, so that a data set is formed. Then, a first framing process is performed, wherein the framing process uses a framing function enframe to perform framing pretreatment, and a rectangular function is selected as a window function. Specifically, let the displacement of the previous frame function relative to the next frame function be δx, the overlap be overlap, the frame rate be fs, and the length of the data set x be L. The overlap is introduced to make the characteristic parameter change smoother. Then the frame rate calculation formula is:. Setting the frame rate of an infrared camera of an infrared imager as f, taking the framing frame rate fs=f and overlapping=2×δx for substituting the framing frame rate fs=f and overlapping=2×δx for better correspondence with a thermal image shot by the infrared imagerThe frame rate calculation formula can obtain δx, and then frame division processing is completed based on the displacement δx. And measuring the data fluctuation condition of each frame by utilizing the short-time average amplitude for the first frame processing data after framing, namely obtaining a first short-time average amplitude set A corresponding to each frame. And judging the abnormal interest group point analysis by adopting an anomaly detection algorithm LOF, searching frame data with abnormal short-time average amplitude value change, temporarily marking the frame data as abnormal data, and determining the corresponding index coordinate position to form first processing data.

S2: and stopping acquiring an optical radiation signal when the powder of the m+1 layer is paved, and starting to acquire an infrared thermal imaging signal until the printing of the m+1 layer is completed.

The alternation of two different data acquisition modes is completed, so that the two data acquisition modes can accurately and reasonably acquire the completed data information.

S3: and performing second signal data processing on the infrared thermal imaging signals to form second processing data.

Performing a second signal data process on the infrared thermal imaging signal to form second processed data, comprising: calculating the molten pool area according to the frame rate on the second signal data to form a frame rate molten pool area data set; detecting abnormal outliers of the frame rate molten pool area data set to form a second abnormal outlier data set; and combining the second abnormal outlier data set and index coordinates corresponding to the second abnormal outlier in the second abnormal outlier data set to form second processing data C.

The infrared thermal imaging signals are processed by framing according to the frame rate of an infrared camera, and then the molten pool area of each frame is calculated to form a frame rate molten pool area data set. And judging the abnormal data in the frame rate molten pool area data set by using an abnormal detection algorithm LOF, and determining index coordinates determined by frames corresponding to the abnormal data after determining the abnormal data to form second processing data.

S4: and performing defect detection analysis on the first processing data and the second processing data to form defect detection result data.

Performing defect detection analysis on the first processing data and the second processing data to form defect detection result data, including: determining an index position reference point, and performing grouping analysis on the first processing data B based on the index position reference point to form a first processing grouping data set D; performing grouping analysis on the second processing data C based on the index position reference points to form a second processing grouping data set E; and setting a position matching threshold gamma, and carrying out matching judgment on the data group in the first processing packet data set D and the data group in the second processing packet data set E to form defect detection result data.

The invention establishes index position reference points with relatively fixed index coordinates to perform grouping analysis based on distance, so as to complete grouping of possible abnormal data at different positions, and the defect is accurately determined by comparing and judging the abnormal data groups obtained in two ways according to the grouping. It should be noted that, in order to avoid that the selected index position reference point is located at an abnormal position, the index position reference point may be determined to be outside the printing area. Meanwhile, the index position reference point is a common reference point for grouping abnormal data obtained in two modes, and the uniformity of data analysis can be ensured.

Wherein determining an index position reference point, performing packet analysis on the first processing data B based on the index position reference point to form a first processing packet data set D, includes: determining index position reference pointsAcquiring index coordinates corresponding to each first abnormal outlier in the first processing data B, and arranging according to a framing sequence to form a first index coordinate sequence setU is the number of the first abnormal outlier determined according to the framing sequence; determining the first index coordinate order set>Each first anomaly index coordinate of (a)Reference point of reference->First index distance betweenForming a first index distance set; according to the arrangement sequence of the first index coordinates in the first index coordinate sequence set, sequentially extracting the first 1+v first index distances in the first index distance set according to the rule that one first index distance is obtained every time the first index distances are increased, and carrying out first average distance->Is calculated by (1):

wherein v is more than or equal to 2 and less than or equal to u; setting a first grouping distance threshold alpha, when there isWhen the first abnormal index coordinates corresponding to the first v first index distances meeting the conditions are determined to be a first abnormal index coordinate set, the calculation of the first average distance is continued from the 1+v first index distance, and grouping judgment is carried out according to a first grouping distance threshold alpha; for each first abnormality index coordinate set, determining a relative index position reference point +. >Is used as a reference distance to determine a reference point of relative index position>Establishing a first range area of the first abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each first index coordinate in the first abnormal index coordinate set as the radius of the first index range; combining all the first abnormal index coordinate sets and the corresponding first range areas to form a first processing grouping data setW represents the total number of the first abnormality index coordinate group, +.>Representing a corresponding first range region.

The first processing data is subjected to grouping analysis, firstly, grouping is divided based on a framing sequence, the main basis of the division is to sequentially expand abnormal index coordinates to judge the average distance of relative index position reference points, and the first grouping distance threshold value can be obtained based on big data analysis so as to improve the accuracy of threshold value judgment. It will be appreciated that if the abnormal index coordinates are two points of different defect locations, the average distance of the reference points to the index locations will be significantly different from the value when the index coordinates of the same defect location are obtained. Meanwhile, because the analysis is carried out according to the framing sequence, the positions of the index coordinates are continuously adjacent, namely when index coordinates of different defect positions exist, the group formed by the previous index coordinates is necessarily the index coordinates positioned under the same defect, and the grouping accuracy is fully ensured. After the group is determined, the defect area defined by the index coordinates in the group needs to be expressed, so that a clear range area can be defined in the subsequent comparison analysis conveniently, an equivalent index coordinate is determined through an average distance and is taken as an ideal center of the defect, a circular area is defined by taking the center as a reference and taking the radius of the distance of the center relative to the farthest index coordinate in the group, the circular area can fully contain the defect area, and the index coordinates outside in the whole circumferential direction can be connected with the center as a reference to form an irregular shape so as to define the defect area more accurately.

For packet analysis of the second processed data C based on the index position reference point, a second processed packet data set E is formed, comprising: obtaining index coordinates corresponding to each second abnormal outlier in the second processing data CAnd arranged according to the framing order to form a second index coordinate order setP is the number of the second abnormal outlier determined according to the framing sequence; determining a second index coordinate order set respectively +.>Is added with index position reference point +.>Second index distance between->Forming a second index distance set; sequentially extracting the first 1+q second index distances in the second index distance set according to the arrangement sequence of the second index coordinates in the second index coordinate sequence set and taking the rule of obtaining one second index distance every time, and carrying out second average distance->Is calculated by (1):wherein q is more than or equal to 2 and less than or equal to p; setting a second packet distance threshold beta when there isDetermining a second abnormal index coordinate corresponding to the previous second index distance meeting the condition as a second abnormal index coordinate set, starting from the 1+q second index distance, continuing to calculate a second average distance, and performing grouping judgment according to a second grouping distance threshold beta; for each second abnormal index coordinate set, determining a relative index position reference point Is used as a reference distance to determine a reference point of the relative index position>Establishing a second range region of the second abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each second index coordinate in the second abnormal index coordinate set as the radius of the second index range; combining all the second abnormal index coordinate sets and the corresponding second range areas to form a second processing grouping data setR represents the total number of first abnormality index coordinate groups, +.>Representing a corresponding second range region.

The packet analysis mode of the second processing data is the same as the packet analysis mode of the first processing data. The determination is also made in frame order based on a distance determination of the second packet distance threshold. After grouping, the center position determined according to the average distance defines a reasonable range to form a investigation range of defect analysis.

Setting a position matching threshold gamma, performing matching judgment on a data group in the first processing packet data set D and a data group in the second processing packet data set E to form defect detection result data, wherein the method comprises the following steps: extracting each second range area from the second processing packet data set E in turn, and calculating and judging the overlapping area of each second range area and all the first range areas in the first processing packet data set D: if the overlapping area is smaller than the position matching threshold gamma, determining that the corresponding first range area and the corresponding second range area are not matched; and if the overlapping area is not smaller than the position matching threshold gamma, determining that the corresponding first range area is matched with the second range area, and determining the position defined by the first range area and the second range area as a defect area.

Because the corresponding defect area ranges are formed in the analysis of the first processing data and the second processing data, when the defect area is confirmed, the judgment is directly carried out according to the coincidence degree of the defect areas acquired by the two data, after all, the higher the coincidence degree is, the higher the identity of the defect positions measured by the two data is, the accuracy of determining the defect positions can be improved while the inaccuracy of defect judgment caused by the distortion of a single data processing mode is eliminated, and after all, the possibility that the position of the coincident area is a real defect area is higher.

S5: repeating the detecting step until printing is completed.

And detecting each subsequent layer of printing, and ensuring the printing quality in real time.

In summary, the device and the method for the non-distortion online monitoring system of the selective laser melting pool provided by the embodiment of the invention have the beneficial effects that:

In the present invention, "at least one" means one or more, and "a plurality" means two or more. "at least one of" or the like means any combination of these items, including any combination of single item(s) or plural items(s). For example, at least one (one) of a, b, or c may represent: a, b, c, a-b, a-c, b-c, or a-b-c, wherein a, b, c may be single or plural.

It should be understood that, in various embodiments of the present invention, the sequence numbers of the foregoing processes do not mean the order of execution, and the order of execution of the processes should be determined by the functions and internal logic thereof, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Those of ordinary skill in the art will appreciate that the elements and method steps of the examples described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or as a combination of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.

In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.

The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random access memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The foregoing examples illustrate only a few embodiments of the invention, which are described in detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention. Accordingly, the scope of protection of the present invention is to be determined by the appended claims.

Claims

1. A method of a selective laser melting bath undistorted on-line monitoring system, comprising:

setting a layer number threshold value n, acquiring printing layer number information, and judging the printing layer number m:

if m is less than n, continuing to acquire the number of printing layers to judge;

if m is more than or equal to n, forming an online detection instruction, and finishing the following detection steps:

acquiring optical radiation signals of m layers, and completing first signal data processing to form first processing data;

Stopping acquiring the optical radiation signal when the powder of the m+1 layer is paved, and starting to acquire an infrared thermal imaging signal until the printing of the m+1 layer is completed;

performing second signal data processing on the infrared thermal imaging signals to form second processing data;

performing defect detection analysis on the first processing data and the second processing data to form defect detection result data;

and repeating the detection step until printing is completed.

2. The method of claim 1, wherein the acquiring optical radiation signals of m layers and completing first signal data processing to form first processing data comprises:

performing first framing processing on the optical radiation signal to form first framing processing data;

acquiring a short-time average amplitude of each frame in the first framing data to form a first short-time average amplitude setI is the total frame number in the first frame processing data;

detecting abnormal outliers of the short-time average amplitude set A to form a first abnormal outlier data set;

and combining the first abnormal outlier data set and index position coordinates corresponding to the first abnormal outlier in the first abnormal outlier data set to form the first processing data B.

3. The method of claim 2, wherein said performing a second signal data process on said infrared thermal imaging signal to form second processed data comprises:

calculating the molten pool area according to the frame rate on the second signal data to form a frame rate molten pool area data set;

detecting abnormal outliers of the frame rate molten pool area data set to form a second abnormal outlier data set;

and combining the second abnormal outlier data set and index coordinates corresponding to the second abnormal outlier in the second abnormal outlier data set to form second processing data C.

4. A method of a selective area laser melting bath undistorted on-line monitoring system as recited in claim 3 wherein said performing defect detection analysis on said first process data and said second process data to form defect detection result data comprises:

determining an index position reference point, and performing grouping analysis on the first processing data B based on the index position reference point to form a first processing grouping data set D;

performing packet analysis on the second processing data C based on the index position reference point to form a second processing packet data set E;

And setting a position matching threshold gamma, and carrying out matching judgment on the data group in the first processing packet data set D and the data group in the second processing packet data set E to form the defect detection result data.

5. The method of claim 4, wherein determining an index position reference point, performing a packet analysis on the first processed data B based on the index position reference point, and forming a first processed packet data set D comprises:

determining index position reference pointsAcquiring index coordinates corresponding to each first abnormal outlier in the first processing data B, and arranging the index coordinates according to a framing sequence to form a first index coordinate sequence set +.>U is the number of the first abnormal outlier determined according to the framing sequence;

determining the first index coordinate order setIs associated with the index position reference point>First index distance between->Forming a first index distance set;

sequentially extracting the first 1+v first index distances in the first index distance set according to the arrangement sequence of the first index coordinates in the first index coordinate sequence set and taking the rule that one first index distance is obtained every time, and carrying out first average distance Is calculated by (1):

wherein v is more than or equal to 2 and less than or equal to u;

setting a first grouping distance threshold alpha, when there isDetermining the first abnormal index coordinates corresponding to the first index distances satisfying the condition as a first abnormal index coordinate set, starting from the 1+v first index distance, continuing to calculate the first average distance, and performing grouping judgment according to the first grouping distance threshold alpha;

for each first abnormal index coordinate set, determining a first index average distance relative to the index position reference point, and determining the index position reference point by taking the first index average distance as a reference distanceEstablishing a first range region of the first abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each first index coordinate in the first abnormal index coordinate set as a first index range radius;

combining all of the first anomaly index coordinate sets and the corresponding first range regions to form the first processed packet data setW represents the total number of the first abnormality index coordinate group, +.>Representing the corresponding first range area.

6. The method of claim 5, wherein said performing a packet analysis on said second processed data C based on said index position reference point to form a second processed packet data set E comprises:

acquiring index coordinates corresponding to each second abnormal outlier in the second processing data C, and arranging the index coordinates according to a framing sequence to form a second index coordinate sequence setP is the number of the second abnormal outlier determined according to the framing sequence;

determining the second index coordinate order setIs associated with the index position reference point>Second index distance between->Forming a second index distance set;

sequentially extracting the first 1+q second index distances in the second index distance set according to the arrangement sequence of the second index coordinates in the second index coordinate sequence set and taking the rule that one second index distance is obtained every time, and carrying out second average distanceIs calculated by (1):

wherein q is more than or equal to 2 and less than or equal to p;

setting a second packet distance threshold beta when there isWhen the first q second index distances satisfying the condition correspond to the second abnormal index coordinates, are determined as a second abnormal index coordinate set, and starting from the 1+q th second index distance, Continuing to calculate the second average distance, and performing grouping judgment according to the second grouping distance threshold value beta;

for each of the second anomaly index coordinate sets, determining a reference point relative to the index locationIs used as a reference distance to determine the reference point relative to the index position>Establishing a second range region of the second abnormal index coordinate set by taking the maximum value of the distance between the average index coordinate and each second index coordinate in the second abnormal index coordinate set as a second index range radius;

combining all of the second anomaly index coordinate sets and the corresponding second range regions to form the second process packet data setR represents the total number of the first abnormality index coordinate group,/or->Representing the corresponding second range area.

7. The method of claim 6, wherein the setting a position matching threshold γ, performing matching judgment on the data set in the first processing packet data set D and the data set in the second processing packet data set E, and forming the defect detection result data, includes:

Sequentially extracting each second range area from the second processing packet data set E, and calculating and judging the overlapping area of each second range area and all the first range areas in the first processing packet data set D:

if the overlapping area is smaller than the position matching threshold gamma, determining that the corresponding first range area and the corresponding second range area are not matched;

and if the overlapping area is not smaller than the position matching threshold gamma, determining that the corresponding first range area and the corresponding second range area are matched, and determining the position defined by the first range area and the second range area as a defect area.