A method and related apparatus for online monitoring of laser powder bed fusion porosity defects

Abstract

This invention discloses an online monitoring method and related equipment for porosity defects in laser powder bed fusion. During processing, airborne acoustic emission signals are collected layer by layer. The signals are segmented according to an intra-layer scanning strategy, decomposed into several scanning vectors, and the duration of each vector is obtained to form an intra-layer unit signal group. Variational mode decomposition is performed on each signal segment to obtain multiple modal components, and the energy proportion of each modal component is calculated. High-frequency modal components are divided according to a preset high-frequency threshold, and the sum of their energy proportions is calculated as the high-frequency energy proportion of the signal segment. The average value is calculated layer by layer to obtain the high-frequency energy proportion change curve, which is used to identify abnormal processing layers. For abnormal layers, the center frequencies of modal components with energy proportions exceeding the preset threshold are statistically analyzed to determine the main frequency distribution range of the high-frequency band of the signal. Based on the matching relationship between this frequency distribution range and the preset porosity defect characteristic frequency range, the porosity defect type is determined. This method improves the accuracy and traceability of frequency band analysis by segmenting the scanning vectors and identifies porosity characteristics in real time. The purpose of this invention is to stably extract key frequency ranges that reflect differences in pore type, segment the signal for a finer-grained processing unit such as the scanning vector, avoid mixing of acoustic information from different spatial locations, and improve the accuracy and traceability of frequency band analysis.

Description

Technical Field

[0001] This invention belongs to the field of laser powder bed melting technology, specifically relating to an online monitoring method and related equipment for porosity defects in laser powder bed melting. Background Technology

[0002] With the widespread application of Laser Powder Bed Fusion (LPBF) technology in aerospace, energy equipment, and mold manufacturing, the impact of internal porosity defects on forming quality and service reliability has gradually attracted attention. During LPBF, factors such as material absorptivity, energy density, molten pool dynamics, and gas flow can easily lead to the formation of various types of defects, including unfused pores, gas pores, and keyhole pores. These pores differ significantly in size, morphology, and formation mechanism, directly affecting the density, mechanical properties, and service life of the components. Therefore, real-time identification of pore characteristics and analysis of their formation patterns during processing have become crucial technical requirements for increasing forming stability.

[0003] To monitor the LPBF process, current research often employs methods such as thermal imagers, coaxial cameras, photodetectors, and acoustic emission sensors to acquire the molten pool temperature field, optical radiation signals, or mechanical vibration signals. Among these, airborne acoustic emission signals can reflect dynamic processes such as energy fluctuations, porosity collapse, and metal vapor flow within the molten pool in a non-contact manner, making them an important method for identifying process anomalies. However, due to the influence of equipment structure, scanning strategy, powder bed materials, and environmental noise, acoustic signals exhibit complex frequency components and strong non-stationarity, making it difficult to directly use the raw acoustic signals to reflect the porosity formation process.

[0004] Existing technologies often employ time-frequency analysis methods such as short-time Fourier transform, wavelet transform, and empirical mode decomposition (EMD) to process LPBF acoustic signals. While these methods can characterize the frequency composition of the signal to some extent, they generally suffer from modal aliasing, insufficient decomposition accuracy, and sensitivity to noise, making it difficult to stably extract key frequency ranges reflecting differences in pore type. Furthermore, existing signal processing workflows typically use the entire layer or the entire scan region as the analysis object, without segmenting the signal at the finer-grained processing unit of the scan vector. This results in the mixing of acoustic information from different spatial locations, reducing the accuracy and traceability of frequency band analysis. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention provides an online monitoring method and related equipment for pore defects in laser powder bed fusion. The purpose is to stably extract key frequency ranges that reflect differences in pore types, segment the signal for a finer-grained processing unit such as the scanning vector, avoid mixing of acoustic information from different spatial locations, improve the accuracy and traceability of frequency band analysis, and further identify pore characteristics and analyze their generation patterns in real time during processing, thereby increasing forming stability.

[0006] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0007] According to a first aspect of the present invention, a method for online monitoring of molten porosity defects in laser powder beds is provided, comprising:

[0008] During laser powder bed melting process, airborne acoustic emission signals are collected in units of layer number;

[0009] The airborne acoustic emission signal is segmented according to the intra-layer scanning strategy, the scanning strategy is decomposed into several scanning vectors, and the duration of each scanning vector is obtained to obtain an intra-layer unit signal group in terms of scanning vector.

[0010] Variational mode decomposition is performed on each signal segment of the unit signal group within the layer to obtain multiple mode components with different frequency ranges;

[0011] Calculate the energy percentage of each modal component in the corresponding signal segment;

[0012] Based on a preset high-frequency threshold, modal components with a center frequency higher than the high-frequency threshold are divided into high-frequency modal components, and the sum of the energy proportions of the high-frequency modal components is calculated as the high-frequency energy proportion of the signal segment.

[0013] For each layer, calculate the average value of the high-frequency energy proportion of the signal segments corresponding to all scan vectors in that layer, and calculate the high-frequency energy proportion change curve in units of layer number layer by layer.

[0014] The abnormal processing layer is identified based on the high-frequency energy ratio change curve.

[0015] For the identified abnormal layer, the center frequencies of the modal components whose energy proportion exceeds a preset ratio threshold are counted in the signal segments corresponding to all scan vectors in that layer, thus obtaining the main frequency distribution range of the high-frequency band of the signal.

[0016] The porosity defect type is determined based on the matching relationship between the main frequency distribution range and the preset porosity defect characteristic frequency range.

[0017] In one possible implementation of the first aspect, the variational mode decomposition is achieved by solving a constrained variational problem, which is:

[0018] To minimize the sum of the bandwidths of all modal components, with the constraint that the sum of all modal components equals the original input signal, the mathematical expression is:

[0019]

[0020] In the formula, For the first One modal component; For the first The central angular frequency of the mode; This represents the total number of modal components. This is a convolution operation; The original input signal; For time variables The first derivative operator; This is the Dirac delta function, used to shift the center frequency to... ; The constraint condition for the modal components is that the sum of all modes is strictly equal to the original input signal.

[0021] In one possible implementation of the first aspect, the solution process for the constrained variational problem is as follows:

[0022] 1) Initialize modal components Center frequency and Lagrange multipliers ;

[0023] 2) Construct an augmented Lagrangian function and use a quadratic penalty term to transform the original constrained optimization problem into an unconstrained variational problem;

[0024] 3) The unconstrained variational problem is solved iteratively using the alternating direction multiplier method framework:

[0025] a) In each iteration In the first step, each modal component is updated based on the frequency domain closed-form solution. ;

[0026] b) Solve for the center frequency using the updated modal components. ;

[0027] c) Last update of the Lagrange multipliers ;

[0028] 4) Repeat the above iterative steps until the convergence condition is met. This yields modal components in multiple frequency ranges;

[0029] The augmented Lagrange multiplier for the unconstrained variational problem is defined as:

[0030]

[0031] in, These are Gram multipliers used to constrain the sum of all modal components to equal the original input signal; It is a secondary penalty factor; To augment the Lagrange function;

[0032] The unconstrained variational problem is solved using the alternating direction multiplier method, with alternating updates. , , Find the saddle point:

[0033]

[0034] According to the parseval theorem, the expression transformed into the frequency domain is:

[0035]

[0036] use By performing variable substitution, we finally obtain:

[0037]

[0038] Similarly, we can conclude that:

[0039]

[0040] In the formula, For the first The frequency domain representation of each modal component, corresponding to the time frequency. Fourier transform; For the first The mode in the th ... Frequency solutions in the next iteration; This is the frequency domain representation of the original input signal; For the frequency domain representation of Lagrange multipliers; It is a continuous frequency variable; This represents the current iteration number; rotating updates. , , This refers to an iterative method that updates variables one by one within the framework of the alternating direction multiplier method.

[0041] In one possible implementation of the first aspect, the calculation of the energy percentage of each modal component in the corresponding signal segment specifically involves:

[0042]

[0043]

[0044] In the formula, For the first The energy percentage of each mode in a signal segment; For the first The energy of each modal component in the corresponding signal segment; This represents the total number of modal components. This is the time index corresponding to the signal frequency band.

[0045] In one possible implementation of the first aspect, the high-frequency energy proportion of the signal segment is calculated using the following formula:

[0046]

[0047] In the formula, This represents the proportion of high-frequency energy in a signal segment. For the first Energy percentage of each mode; For the first The center frequency of each modal component; High-frequency threshold; It is the set of all high-frequency modes with a center frequency greater than the threshold.

[0048] In one possible implementation of the first aspect, the step of calculating the average high-frequency energy percentage of the signal segments corresponding to all scan vectors within each layer specifically involves:

[0049]

[0050] In the formula, For the first The average high-frequency energy percentage of the layer; For the first Layer The proportion of high-frequency energy in each scan vector signal segment; For the first Number of scan vectors for a layer.

[0051] In one possible implementation of the first aspect, the porosity defect types include unfused porosity and keyhole porosity, wherein the characteristic frequency range corresponding to the unfused porosity is 8kHz to 8.5kHz, and the characteristic frequency range corresponding to the keyhole porosity is 38kHz to 42kHz.

[0052] According to a second aspect of the present invention, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned method for online monitoring of pore defects in laser powder bed fusion.

[0053] According to a third aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned method for online monitoring of porosity defects in laser powder bed fusion.

[0054] According to a fourth aspect of the present invention, a computer program product is provided, which, when executed by a processor, implements the aforementioned method for online monitoring of pore defects in laser powder bed fusion.

[0055] Compared with the prior art, the present invention has at least the following beneficial effects:

[0056] This invention provides an online monitoring method for molten porosity defects in laser powder bed fusion. By finely segmenting the airborne acoustic emission signal according to the scanning vector, it avoids the mixing of acoustic information from different spatial locations, enabling frequency band analysis to accurately correspond to specific processing areas and improving the reliability of defect localization and feature extraction. Variational mode decomposition is used to process the signal segment of each scanning vector, overcoming the mode aliasing and noise sensitivity problems of traditional time-frequency analysis methods (such as short-time Fourier transform or empirical mode decomposition). This allows for more stable separation of key frequency components related to porosity type, thereby enhancing the ability to distinguish defects such as unfused porosity and pores. By calculating the high-frequency energy ratio change curve layer by layer, abnormal processing layers can be quickly identified, and the process status can be fed back in a timely manner, making the forming process more stable and controllable. For abnormal layers, by statistically analyzing the main frequency distribution intervals and matching them with preset porosity defect characteristic frequencies, the porosity type (such as unfused porosity, pores, or keyhole porosity) can be accurately determined. This invention, through online monitoring and defect identification, can reduce internal porosity defects in components, improve density and mechanical properties, thereby extending the service life of components in aerospace, energy equipment and other fields. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the specific embodiments of the present invention, the drawings used in the description of the specific embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0058] Figure 1 This is a flowchart of an online monitoring method for porosity defects in laser powder bed fusion according to the present invention;

[0059] Figure 2 This is a schematic diagram of the airborne acoustic emission signal acquisition system for the laser powder bed melting process in the embodiment.

[0060] Figure 3 The porosity characterization results for different processing parameters in the examples are shown below;

[0061] Figure 4 The embodiment shows the variation of the high-frequency proportion in the airborne acoustic emission signal of each layer under different process parameters.

[0062] Figure 5 The correlation between the number of keyhole pores and the proportion of high frequency in the embodiment is shown;

[0063] Figure 6 In this embodiment, the main frequency distribution under different pore types is statistically analyzed using scanning vectors as the unit. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0065] like Figure 1 As shown, this invention provides an online monitoring method for porosity defects in laser powder bed fusion, and the specific implementation steps are as follows:

[0066] S1. During the laser powder bed melting process, airborne acoustic emission signals are collected in units of layers.

[0067] Specifically, during laser powder bed melting, non-contact airborne acoustic emission sensors are used to collect acoustic signals from the processing area in real time. Signal acquisition is performed on a layer-by-layer basis; that is, after scanning each layer, the complete acoustic signal data corresponding to that layer is saved. The sampling frequency can be set according to actual needs, for example, from 50 kHz to 100 kHz, to ensure coverage of the effective frequency range of the acoustic signals.

[0068] S2. The airborne acoustic emission signal is segmented according to the intra-layer scanning strategy, the scanning strategy is decomposed into several scanning vectors, and the duration of each scanning vector is obtained to obtain an intra-layer unit signal group in terms of scanning vector.

[0069] In other words, based on the intra-layer scanning strategy of laser powder bed melting, such as bidirectional scanning or helical scanning, the acoustic signals acquired in step S1 are segmented. The scanning strategy can be decomposed into multiple scanning vectors, each representing a segment of the path continuously scanned by the laser beam on the powder bed. By parsing the scanning strategy file, the start time, end time, and duration of each scanning vector are obtained. Based on this time information, the intra-layer acoustic signal is cut into multiple signal segments with scanning vectors as units, forming intra-layer unit signal groups. This step ensures a one-to-one correspondence between acoustic signals and spatial positions, avoiding the mixing of acoustic information from different scanning vectors.

[0070] Specifically, airborne acoustic emission signals It can be represented as:

[0071]

[0072] in, This refers to the number of processing layers.

[0073] Acoustic emission signal within the layer It is divided into signals in units of scan vectors according to the scanning method. The signal group consists of:

[0074]

[0075] S3. Perform variational mode decomposition on each signal segment of the unit signal group within the layer to obtain multiple mode components with different frequency ranges.

[0076] In other words, variational mode decomposition (VMD) is performed on each scan vector signal segment obtained in step S2. The mode components are distributed from low frequency to high frequency, which can effectively separate complex frequency components in the signal and reduce mode aliasing and noise interference.

[0077] It should be understood that VMD is an adaptive signal decomposition method that can decompose a non-stationary signal into multiple intrinsic mode functions, each mode component having a finite bandwidth and center frequency.

[0078] In one possible implementation, the variational mode decomposition is achieved by solving a constrained variational problem, which is:

[0079] To minimize the sum of the bandwidths of all modal components, with the constraint that the sum of all modal components equals the original input signal, the mathematical expression is:

[0080]

[0081] In the formula, For the first One modal component; For the first The central angular frequency of the mode; The number of modal components; This is a convolution operation; The original input signal; For time variables The first derivative operator; This is the Dirac delta function, used to shift the center frequency to... ; The constraint condition for the modal components is that the sum of all modes is strictly equal to the original input signal.

[0082] In one possible implementation, the solution process for the constrained variational problem is as follows:

[0083] 1) Initialize modal components Center frequency and Lagrange multipliers ;

[0084] 2) Construct an augmented Lagrangian function and use a quadratic penalty term to transform the original constrained optimization problem into an unconstrained variational problem;

[0085] 3) The unconstrained variational problem is solved iteratively using the alternating direction multiplier method framework:

[0086] a) In each iteration In the first step, each modal component is updated based on the frequency domain closed-form solution. ;

[0087] b) Solve for the center frequency using the updated modal components. ;

[0088] c) Last update of the Lagrange multipliers ;

[0089] 4) Repeat the above iterative steps until the convergence condition is met. This yields modal components in multiple frequency ranges;

[0090] Specifically, by introducing a quadratic penalty term and a Lagrange factor, the constrained variational problem is transformed into an unconstrained variational problem. The augmented Lagrange multiplier for the unconstrained variational problem is defined as:

[0091]

[0092] in, These are Gram multipliers used to constrain the sum of all modal components to equal the original input signal; It is a secondary penalty factor; To augment the Lagrange function; For the first One modal component; For the first The central angular frequency of the mode.

[0093] The unconstrained variational problem is solved using the alternating direction multiplier method, with alternating updates. , , Find the saddle point:

[0094]

[0095] According to the parseval theorem, the expression transformed into the frequency domain is:

[0096]

[0097] use By performing variable substitution, we finally obtain:

[0098]

[0099] Similarly, we can conclude that:

[0100]

[0101] In the formula, For the first The frequency domain representation of each modal component, corresponding to the time frequency. Fourier transform; For the first The mode in the th ... Frequency solutions in the next iteration; This is the frequency domain representation of the original input signal. For the frequency domain representation of Lagrange multipliers; It is a continuous frequency variable; This represents the current iteration number; rotating updates. , , This refers to an iterative method that updates variables one by one within the framework of the alternating direction multiplier method.

[0102] S4. Calculate the energy percentage of each modal component in the corresponding signal segment.

[0103] In other words, for each scan vector signal segment, the energy percentage of each modal component obtained in step S3 is calculated, and the energy percentage reflects the relative importance of different frequency components in the signal.

[0104] In one possible implementation, calculating the energy percentage of each modal component in the corresponding signal segment specifically involves:

[0105]

[0106]

[0107] In the formula, For the first The energy percentage of each mode in a signal segment; For the first The energy of each modal component in the corresponding signal segment; For the first One modal component; This represents the total number of modal components. This is the time index corresponding to the signal frequency band.

[0108] S5. Based on a preset high-frequency threshold, the modal components with a center frequency higher than the high-frequency threshold are divided into high-frequency modal components, and the sum of the energy proportions of the high-frequency modal components is calculated as the high-frequency energy proportion of the signal segment.

[0109] In other words, based on a preset high-frequency threshold, the modal components decomposed in step S3 are classified according to their center frequencies. The high-frequency threshold can be set according to the actual processing material and process parameters. For example, for metal powder materials, the high-frequency threshold can be set to 20 kHz to 50 kHz. All modes with center frequencies higher than the high-frequency threshold are classified as high-frequency modes. Then, the sum of the energy proportions of these high-frequency modes is calculated as the high-frequency energy proportion of the scanned vector signal segment. The high-frequency energy proportion can be used to characterize the high-frequency fluctuations related to porosity defects in the dynamics of the molten pool.

[0110] In one possible implementation, the high-frequency energy proportion of the signal segment is calculated using the following formula:

[0111]

[0112] In the formula, This represents the proportion of high-frequency energy in a signal segment. For the first Energy percentage of each mode; For the first The center frequency of each modal component; High-frequency threshold; It is the set of all high-frequency modes with a center frequency greater than the threshold.

[0113] S6. For each layer, calculate the average value of the high-frequency energy ratio of the signal segments corresponding to all scan vectors in that layer, and calculate the high-frequency energy ratio change curve in units of layer number layer by layer.

[0114] Specifically, for each processing layer, the high-frequency energy percentage of all scan vector signal segments within that layer is collected, and their arithmetic mean is calculated to obtain the average high-frequency energy percentage for that layer. As the number of processing layers increases, these average values ​​are recorded layer by layer, and a high-frequency energy percentage variation curve is plotted with the layer number as the unit. The high-frequency energy percentage variation curve reflects the overall trend of high-frequency energy during processing, facilitating the identification of abnormal fluctuations.

[0115] In one possible implementation, the step of calculating the average high-frequency energy percentage of the signal segments corresponding to all scan vectors within each layer specifically involves:

[0116]

[0117] In the formula, For the first The average high-frequency energy percentage of the layer; For the first Layer The proportion of high-frequency energy in each scan vector signal segment; For the first Number of scan vectors for a layer.

[0118] S7. Identify the abnormal processing layer based on the high-frequency energy ratio change curve.

[0119] Specifically, based on the high-frequency energy ratio change curve obtained in step S6, a sliding window is used to identify significant peaks or valleys in the curve in order to locate processing anomalies.

[0120] S8. For the identified abnormal layer, count the center frequencies of the modal components whose energy proportion exceeds the preset ratio threshold in all signal segments corresponding to the scan vectors in that layer, and obtain the main frequency distribution range of the high-frequency band of the signal.

[0121] In other words, for the abnormal layer identified in step S7, the modal frequency distribution of each scan vector signal segment within that layer is further analyzed. Specifically, the center frequencies of modes whose energy percentage exceeds a preset threshold, such as 5% to 10%, are counted among all signal segments in that layer. These center frequency values ​​are collected, and their distribution range is calculated to obtain the main frequency distribution range of the high-frequency band of the signal. This step targets the local frequency characteristics within the abnormal layer, improving the specificity of the analysis.

[0122] For example, for each layer of signal Each unit vector signal The center frequencies of modal components whose energy accounts for more than 10% are statistically analyzed. :

[0123]

[0124] in, For the first The first layer within the layer The unit vector signal of the first unit vector signal The energy percentage of each modal component.

[0125] The overall frequency band distribution during the processing is defined as follows:

[0126]

[0127] in, This means flattening the matrix into a one-dimensional sequence representation.

[0128] S9. Determine the type of pore defect based on the matching relationship between the main frequency distribution range and the preset pore defect characteristic frequency range.

[0129] Specifically, the main frequency distribution range obtained in step S8 is matched with the preset characteristic frequency range of pore defects. It should be noted that the preset characteristic frequency range is based on historical data or experimental calibration; the characteristic frequency range corresponding to unfused pores is 8kHz to 8.5kHz, and the characteristic frequency range corresponding to keyhole pores is 38kHz to 42kHz. By comparing the degree of overlap between the main frequency distribution range and these characteristic ranges, the type of pore defect is determined. For example, if the main frequency distribution range highly matches the keyhole pore characteristic range, then the abnormal layer is determined to have a keyhole pore defect.

[0130] Example

[0131] Figure 2 The diagram shows a schematic of the airborne acoustic emission signal acquisition system for the laser powder bed melting process in this embodiment. It mainly includes a laser 1, a processing component 2, powder 3, a substrate 4, an acoustic signal sensor 5, an airborne acoustic emission sensor 6, a data acquisition box 7, and a computer 8. During laser processing, the airborne acoustic emission acquisition system collects acoustic signals in real time. In this embodiment, the sampling rate of the airborne acoustic emission sensor 6 is 100Hz.

[0132] In this embodiment, a total of 8 sets of processing parameters were used to produce components with different porosity levels. The processing parameter settings used in this embodiment are shown in Table 1; components with different porosity levels were obtained under the processing parameters corresponding to those in Table 1. Figure 3 The results of porosity characterization under different processing parameters are shown in Table 2. The number of pores and porosity under different processing parameters were statistically analyzed using computer vision algorithms.

[0133] Table 1 Processing Parameter Settings

[0134]

[0135] Table 2. Number of pores and their corresponding categories

[0136]

[0137] This embodiment processes acoustic signals corresponding to different porosity levels, and uses the method of this invention to statistically analyze the high-frequency proportion, analyzing the changes in the high-frequency proportion in the airborne acoustic emission signal of each layer under different processing parameters, such as... Figure 4 As shown, when the pore type is unfused, there are virtually no high-frequency (>20kHz) acoustic components in the airborne acoustic emission signal. As the pore type changes to keyhole pores, high-frequency components begin to appear in the acoustic signal, and the proportion of high frequencies increases with the increase in the number of keyhole pores and porosity.

[0138] Furthermore, the results of keyhole porosity were analyzed separately, and the correlation between the number of keyhole porosity and the proportion of high frequencies was as follows: Figure 5 As shown, the more keyhole pores there are, the higher the proportion of high frequency signals, which is basically proportional. Therefore, keyhole pores will bring high frequency signals, which is consistent with the phenomenon of shock waves that occur during the formation of keyhole pores.

[0139] Finally, according to the method of this invention, the main frequency distribution under different pore types is statistically analyzed using the scanning vector as the unit. Here, only the frequency distribution interval is considered, not the frequency percentage. Figure 6 As shown. Based on the above results, it was found that the low-frequency band (characteristic frequency range) of the unfused pores and keyhole pores is mainly distributed in the range of 8k~8.5kHz, while the high-frequency band (characteristic frequency range) of the keyhole pores is 38k~42kH.

[0140] In another embodiment of the present invention, a computer device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used in the operation of an online monitoring method for molten porosity defects in laser powder bed.

[0141] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory), which is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be Random Access Memory (RAM) or non-volatile memory, such as at least one disk storage device. The processor can load and execute one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the online monitoring method for molten porosity defects in a laser powder bed in the above embodiments.

[0142] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0143] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0144] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0145] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0146] This invention also provides a computer program product, which is used to execute any of the above-described methods for online monitoring of porosity defects in laser powder bed fusion. Since the computer program product provided by this invention belongs to the same inventive concept as the above-described method for online monitoring of porosity defects in laser powder bed fusion, it possesses all the advantages of the above-described method. Therefore, the beneficial effects of the computer program product provided by this invention will not be elaborated upon here.

[0147] In this invention, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0148] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the scope of the technology disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention.

Claims

1. A method for online monitoring of laser powder bed fusion porosity defects, characterized in that, include: During laser powder bed melting process, airborne acoustic emission signals are collected in units of layer number; The airborne acoustic emission signal is segmented according to the intra-layer scanning strategy, the scanning strategy is decomposed into several scanning vectors, and the duration of each scanning vector is obtained to obtain an intra-layer unit signal group in terms of scanning vector. Variational mode decomposition is performed on each signal segment of the unit signal group within the layer to obtain multiple mode components in different frequency ranges. The variational mode decomposition is achieved by solving a constrained variational problem, which is: To minimize the sum of the bandwidths of all modal components, with the constraint that the sum of all modal components equals the original input signal, the mathematical expression is: In the formula, For the first One modal component; For the first The central angular frequency of the mode; This represents the total number of modal components. This is a convolution operation; The original input signal; For time variables The first derivative operator; This is the Dirac delta function, used to shift the center frequency to... ; The constraint condition for the modal components is that the sum of all modes is strictly equal to the original input signal; The energy percentage of each modal component in the corresponding signal segment is calculated as follows: In the formula, For the first The energy percentage of each mode in a signal segment; For the first The energy of each modal component in the corresponding signal segment; This represents the total number of modal components. This is the time index corresponding to the signal frequency band; Based on a preset high-frequency threshold, modal components with a center frequency higher than the high-frequency threshold are divided into high-frequency modal components, and the sum of the energy proportions of the high-frequency modal components is calculated as the high-frequency energy proportion of the signal segment. For each layer, calculate the average value of the high-frequency energy proportion of the signal segments corresponding to all scan vectors in that layer, and calculate the high-frequency energy proportion change curve in units of layer number layer by layer. The abnormal processing layer is identified based on the high-frequency energy ratio change curve. For the identified abnormal layer, the center frequencies of the modal components whose energy proportion exceeds a preset ratio threshold are counted in the signal segments corresponding to all scan vectors in that layer, thus obtaining the main frequency distribution range of the high-frequency band of the signal. Based on the matching relationship between the main frequency distribution range and the preset characteristic frequency range of pore defects, the pore defect type is determined; the pore defect type includes unfused pores and keyhole pores, wherein the characteristic frequency range corresponding to unfused pores is 8kHz to 8.5kHz, and the characteristic frequency range corresponding to keyhole pores is 38kHz to 42kHz.

2. The method for online monitoring of molten porosity defects in laser powder bed according to claim 1, characterized in that, The solution process for the constrained variational problem is as follows: 1) Initialize modal components Center frequency and Lagrange multipliers ; 2) Construct an augmented Lagrangian function and use a quadratic penalty term to transform the original constrained optimization problem into an unconstrained variational problem; 3) The unconstrained variational problem is solved iteratively using the alternating direction multiplier method framework: a) In each iteration In the first step, each modal component is updated based on the frequency domain closed-form solution. ; b) Solve for the center frequency using the updated modal components. ; c) Last update of the Lagrange multipliers ; 4) Repeat the above iterative steps until the convergence condition is met. This yields modal components in multiple frequency ranges; The augmented Lagrange multiplier for the unconstrained variational problem is defined as: in, These are Gram multipliers used to constrain the sum of all modal components to equal the original input signal; It is a secondary penalty factor; To augment the Lagrange function; The unconstrained variational problem is solved using the alternating direction multiplier method, with alternating updates. , , Find the saddle point: According to the parseval theorem, the expression transformed into the frequency domain is: use By performing variable substitution, we finally obtain: Similarly, we can conclude that: In the formula, For the first The frequency domain representation of each modal component, corresponding to the time frequency. Fourier transform; For the first The mode in the th ... Frequency solutions in the next iteration; This is the frequency domain representation of the original input signal; For the frequency domain representation of Lagrange multipliers; It is a continuous frequency variable; This represents the current iteration number; Rotational updates. , , This refers to an iterative method that updates variables one by one within the framework of the alternating direction multiplier method.

3. The method for online monitoring of molten porosity defects in laser powder bed according to claim 1, characterized in that, The high-frequency energy proportion of the signal segment is calculated using the following formula: In the formula, This represents the proportion of high-frequency energy in a signal segment. For the first Energy percentage of each mode; For the first The center frequency of each modal component; High-frequency threshold; It is the set of all high-frequency modes with a center frequency greater than the threshold.

4. The method for online monitoring of molten porosity defects in laser powder bed according to claim 1, characterized in that, For each layer, the average high-frequency energy percentage of the signal segments corresponding to all scan vectors within that layer is calculated, specifically as follows: In the formula, For the first The average high-frequency energy percentage of the layer; For the first Layer The proportion of high-frequency energy in each scan vector signal segment; For the first Number of scan vectors for a layer.

5. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements an online monitoring method for molten porosity defects in laser powder bed as described in any one of claims 1 to 4.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the online monitoring method for molten porosity defects in laser powder bed as described in any one of claims 1 to 4.

7. A computer program product, characterized in that, When the computer program is executed by a processor, it implements an online monitoring method for molten porosity defects in a laser powder bed as described in any one of claims 1 to 4.

Images