A method and system for quality inspection of 3D printing engineering materials

Abstract

The present invention discloses a quality detection method and system for 3D printing engineering materials. The quality detection method for 3D printing engineering materials includes: collecting data of the 3D printing process and pre-processing the 3D printed samples; using a high-resolution spectrometer to comprehensively scan the pre-processed samples, recording the spectral feature data of different regions, and establishing a spectral feature data matrix; analyzing the spectral feature data, calculating the defect values ​​of different regions, and updating the defect values ​​of different regions to the spectral feature data matrix; calculating the total error value of the printing process according to the data of the 3D printing process; if the total error value does not exceed the set threshold, the quality of the 3D printed sample is determined to be qualified, and if the total error value exceeds the set threshold, the region with the highest defect value is retrieved from the spectral feature data matrix to provide feedback information. The present application provides a comprehensive evaluation of multiple factors for the quality detection of 3D printing engineering materials, which is conducive to quality control and fixed-point detection.

Description

Technical Field

[0001] The present invention relates to spectral analysis technology, and in particular to a method and system for quality detection of 3D printing engineering materials. Background Art

[0002] Despite the rapid development of 3D printing technology, it still faces challenges in terms of material performance, mechanical strength, precision and cost. The introduction of spectral analysis technology has shown unique value in addressing these limitations. As a flexible and non-destructive detection method, it is suitable for rapid analysis of rare materials and reduces sample loss. Spectral analysis can greatly improve detection efficiency with its high-speed multi-channel characteristics, and high-precision analysis capabilities, especially the advantages of near-infrared microspectroscopy in fine structure and composition analysis, are crucial to the quality control of 3D printing materials. Summary of the invention

[0003] The present invention provides a method for quality inspection of 3D printing engineering materials, comprising:

[0004] S110, collecting data of the 3D printing process and pre-processing the 3D printing samples;

[0005] S120, using a high-resolution spectrometer to comprehensively scan the pre-treated sample, record spectral feature data of different regions, and establish a spectral feature data matrix;

[0006] S130, analyzing the spectral feature data, calculating the defect values ​​of different regions, and updating the defect values ​​of different regions into the spectral feature data matrix;

[0007] S140, calculating a total error value of the printing process according to data of the 3D printing process;

[0008] S150. If the total error value does not exceed the set threshold, the quality of the 3D printed sample is determined to be qualified. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided.

[0009] A method for quality inspection of 3D printing engineering materials as described above, wherein the 3D printing process includes three-dimensional modeling, slicing processing, material loading and printing, and the data of the 3D printing process includes three-dimensional model data, slicing data, material property data and printing configuration data.

[0010] A method for quality inspection of 3D printing engineering materials as described above, wherein the spectral characteristic data includes an intensity characteristic value, a baseline value and a signal-to-noise ratio, wherein the intensity characteristic value represents the light intensity detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to light of a specific wavelength; the baseline value is the lowest continuous background of the spectrum graph, and ideally should be a horizontal line; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise.

[0011] In the above-mentioned method for quality inspection of 3D printing engineering materials, the spectral feature data analysis and defect value calculation are specifically divided into the following sub-steps:

[0012] Perform filtering operations on spectral feature data;

[0013] Perform feature extraction on the spectral feature data after filtering to determine the key feature peak position and intensity;

[0014] Calculate defect values ​​in different areas based on the position and intensity of feature peaks extracted from features;

[0015] Update the defect values ​​of different regions into the spectral feature data matrix;

[0016] In the above-mentioned method for quality inspection of 3D printing engineering materials, the process of calculating the process error value specifically includes the following sub-steps:

[0017] Calculate geometric error values ​​based on three-dimensional model data;

[0018] Calculate the material shrinkage compensation error value according to the material property data;

[0019] Calculate the interlayer bonding error value based on the slice data;

[0020] Calculate the printing path deviation error value according to the printing configuration data;

[0021] Calculate the total error value according to the geometric error value, the material shrinkage compensation error value, the interlayer bonding error value and the printing path deviation error value;

[0022] The present invention also provides a system for quality inspection of 3D printing engineering materials, including: a data collection and preprocessing module, a spectral feature data matrix module, a defect value calculation module, an error value calculation module, and a feedback module.

[0023] Data collection and preprocessing module: collects data from the 3D printing process and preprocesses 3D printed samples;

[0024] Spectral feature data matrix module: Use a high-resolution spectrometer to comprehensively scan the pre-treated samples, record the spectral feature data of different areas, and establish a spectral feature data matrix;

[0025] Defect value calculation module: analyzes the spectral feature data, calculates the defect values ​​of different areas, and updates the defect values ​​of different areas to the spectral feature data matrix;

[0026] Error value calculation module: calculates the total error value of the printing process based on the data of the 3D printing process;

[0027] Feedback module: If the total error value does not exceed the set threshold, the 3D printed sample is judged to be of qualified quality. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided;

[0028] A system for quality inspection of 3D printing engineering materials as described above, wherein the 3D printing process includes three-dimensional modeling, slicing processing, material loading and printing, and the data of the 3D printing process includes three-dimensional model data, slicing data, material property data and printing configuration data.

[0029] A system for quality inspection of 3D printing engineering materials as described above, wherein the spectral characteristic data includes an intensity characteristic value, a baseline value and a signal-to-noise ratio, wherein the intensity characteristic value represents the light intensity detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to light of a specific wavelength; the baseline value is the lowest continuous background of the spectrum graph, and ideally should be a horizontal line; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise.

[0030] In the above-mentioned system for quality inspection of 3D printing engineering materials, the spectral feature data analysis and defect value calculation are specifically divided into the following sub-steps:

[0031] Perform filtering operations on spectral feature data;

[0032] Perform feature extraction on the spectral feature data after filtering to determine the key feature peak position and intensity;

[0033] Calculate defect values ​​in different areas based on the position and intensity of feature peaks extracted from features;

[0034] Update the defect values ​​of different regions into the spectral feature data matrix;

[0035] In the system for quality inspection of 3D printing engineering materials as described above, the process of calculating the process error value specifically includes the following sub-steps:

[0036] Calculate geometric error values ​​based on three-dimensional model data;

[0037] Calculate the material shrinkage compensation error value according to the material property data;

[0038] Calculate the interlayer bonding error value based on the slice data;

[0039] Calculate the printing path deviation error value according to the printing configuration data;

[0040] Calculate the total error value according to the geometric error value, the material shrinkage compensation error value, the interlayer bonding error value and the printing path deviation error value;

[0041] The beneficial effects achieved by the present invention are as follows: a comprehensive evaluation of multiple factors is provided for the quality inspection of 3D printing engineering materials, which is beneficial to quality control and fixed-point inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0043] Figure 1 This is a flow chart of a method for quality inspection of 3D printing engineering materials provided in Example 1 of the present application;

[0044] Figure 2 It is a schematic diagram of a system for quality inspection of 3D printing engineering materials provided in the second embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0046] Embodiment 1

[0047] like Figure 1 As shown, Embodiment 1 of the present application provides a quality inspection method for 3D printing engineering materials.

[0048] S110: Collect data of the 3D printing process and pre-process the 3D printed samples;

[0049] The 3D printing process includes 3D modeling, slicing, loading materials and printing. The data of the 3D printing process includes 3D model data, slicing data, material property data and printing configuration data. The 3D model data is usually generated by CAD (computer-aided design) software, and the geometric information of the 3D model is described by triangular facets. The slicing data contains the instruction information such as the printing path, filling mode, speed, temperature, etc. of each layer. The material property data includes the physical and chemical properties of the materials used, such as melting temperature, viscosity, curing rate, strength, etc. The printer configuration data involves the hardware settings of the printer, such as nozzle diameter, platform temperature, upper limit of printing speed, etc.

[0050] After 3D printing is completed, the sample is pre-processed to remove the support structure and ensure the flatness and cleanliness of the test surface.

[0051] S120: Use a high-resolution spectrometer to comprehensively scan the pre-treated sample, record the spectral characteristic data of different regions, and establish a spectral characteristic data matrix;

[0052] A high-precision spectrometer is used to scan the surface and near-surface of 3D printed parts, and the material composition, microstructure and potential defects (such as pores, cracks, etc.) are identified through the absorption and reflection characteristics of light of different wavelengths.

[0053] Spectral characteristic data include intensity characteristic value, baseline value and signal-to-noise ratio. The intensity characteristic value represents the light intensity detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to the specific wavelength of light; the baseline value is the lowest continuous background of the spectrum graph, which should be a horizontal line ideally; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise. A high signal-to-noise ratio means that the data is more reliable and the analysis results are more accurate.

[0054] The spectral characteristic data matrix is ​​used to record the spectral data of different regions. Specifically, the spectral characteristic data matrix is ​​expressed as: Where A represents the spectral feature data matrix, a ij represents the spectral feature data subset of the i-th row and j-th column area, m represents the maximum number of rows in the area, n represents the maximum number of columns in the area, and I ij represents the intensity characteristic value of the region in the i-th row and j-th column, B ij represents the baseline value of the region in the i-th row and j-th column, SNR ij represents the signal-to-noise ratio of the region in the i-th row and j-th column, DF ij Indicates the defect value, and the initial value is 0.

[0055] S130: Analyze the spectral feature data, calculate the defect values ​​of different regions, and update the defect values ​​of different regions into the spectral feature data matrix;

[0056] Spectral feature data analysis and defect value calculation are specifically divided into the following sub-steps:

[0057] S131: filtering the spectral feature data;

[0058] The filtering technology is applied to remove the random noise in the spectral feature data, and the iterative correction method is used to remove the non-uniform background baseline of the spectrum. The corrected intensity characteristic value is I corrected , the spectral data is normalized to eliminate the intensity difference and facilitate comparison. Specifically, the formula is used: Calculate the normalized intensity characteristic value, where I ref represents the normalized intensity eigenvalue, I corrected represents the corrected intensity characteristic value, max(I corrected ) indicates the maximum value of the corrected intensity.

[0059] S132: extracting features from the filtered spectral feature data to determine key feature peak positions and intensities;

[0060] After filtering, the spectral feature data is extracted to determine the key characteristic peak position and intensity. The characteristic peak is represented by λ peak .

[0061] S133: Calculating defect values ​​of different regions according to the characteristic peak positions and intensities extracted from the features;

[0062] The defect value is divided into two defect indicators. Defect indicator one is the intensity characteristic value deviation score, and defect indicator two is the signal-to-noise ratio difference score. This application obtains a comprehensive defect score by weighted calculation of the two defect indicator scores. The comprehensive defect score is the defect value. The defect value is used to evaluate the quality of different areas of the 3D printed sample. It is specifically divided into the following sub-steps:

[0063] 1. Calculate the intensity eigenvalue deviation score;

[0064] Assume that the reference region is at the characteristic peak λ peak The intensity at I ref (λ peak ), the corresponding intensity of the area to be tested is I ij (λ peak ), define the intensity eigenvalue deviation fraction DF intensity To quantify the intensity difference, DF intensity Close to 0 indicates a good intensity match, and the greater the deviation, the higher the possibility of defects. Specifically, the formula is used: Calculate the intensity characteristic value deviation score of each region, where (DF intensity ) ij represents the intensity eigenvalue deviation score of the i-th row and j-th column area, Iref (λ peak ) indicates that the reference area is at the characteristic peak λ peak The intensity at I ij (λ peak ) indicates that the area to be tested is at the characteristic peak λ peak The strength of the place.

[0065] 2. Calculate the signal-to-noise ratio difference score;

[0066] Let the signal-to-noise ratio in the reference area be SNR ref , the signal-to-noise ratio of the area to be measured is SNR ij , define the signal-to-noise ratio difference score DF SNR , DF SNR It reflects the degree of decrease in the signal-to-noise ratio. The smaller the value, the closer the data quality is to the standard or reference state. Specifically, the formula is: Calculate the signal-to-noise ratio difference score of each region, where (DF SNR ) ij Represents the signal-to-noise ratio difference score of the i-th row and j-th column area, SNR ref Indicates the signal-to-noise ratio in the reference area, SNR ij Indicates the signal-to-noise ratio of the area to be measured.

[0067] 3. Calculate the comprehensive defect score based on the intensity characteristic value deviation score and the signal-to-noise ratio difference score;

[0068] Taking into account the intensity deviation and signal-to-noise ratio difference, the comprehensive defect score DF is defined total , specifically, using the formula: Calculate the comprehensive defect score, where (DF total ) ij represents the comprehensive defect score of the region in row i and column j, (DF intensity ) ij represents the intensity eigenvalue deviation score of the i-th row and j-th column region, (DF SNR ) ij represents the signal-to-noise ratio difference score of the i-th row and j-th column area, represents the weight of the intensity eigenvalue deviation score, and β represents the weight of the signal-to-noise ratio difference score.

[0069] S134: updating defect values ​​of different regions into the spectral feature data matrix;

[0070] The comprehensive defect score DF of the area calculated in the above steps is total Assign the defect value DF to the corresponding area in the spectral feature data matrix A ij , specifically, using the formula:

[0071] DF ij =DFtotal ,i=1,2,3,...,m,j=1,2,3,...,n are assigned, where DF ij represents the defect value of the area in the i-th row and j-th column, DF total represents the comprehensive defect score of the region in the i-th row and j-th column, m represents the maximum number of rows in the region, and n represents the maximum number of columns in the region, until the defect values ​​of all regions in the spectral feature data matrix are assigned.

[0072] S140: Calculating a total error value of the printing process according to data of the 3D printing process;

[0073] In the 3D printing process, due to the deviations in the chemical properties and working conditions of the printing materials and 3D printers under different temperature environments, there will be certain errors in the finished products. This application calculates the error value in the 3D printing process by establishing a multi-factor integrated error evaluation model. The data collected during the 3D printing process, including three-dimensional model data, slice data, material property data, and printing configuration data, are used as input data of the multi-factor integrated error evaluation model. The process of calculating the process error value specifically includes the following sub-steps:

[0074] S141: Calculating geometric error values ​​according to the three-dimensional model data;

[0075] There is a certain error between the geometric values ​​of the expected sample and the actual sample during 3D printing. Specifically, the formula is used:

[0076] Calculate the geometric error value, where E geom represents the geometric error value, represents the expected sample geometry fixed point coordinates, Represents the actual sample geometric fixed point coordinates, i represents the coordinate subscript, and its value is 1...n.

[0077] S142: Calculating a material shrinkage compensation error value according to material property data;

[0078] The printing materials used in the 3D printing process will shrink due to actual conditions such as temperature, and there will be a certain error value. Specifically, the formula is used:

[0079] E shrink =V model ×(η shrink +ε temp ×(ΔT cooling )) Calculate the material shrinkage compensation error value, where E shrink Indicates the material shrinkage compensation error value, V model represents the sample volume, η shrink represents the basic shrinkage rate, ε temp Denotes the coefficient of thermal expansion, ΔTcooling Indicates the change in the print temperature when it cools down to room temperature.

[0080] S143: Calculating the interlayer bonding error value according to the slice data;

[0081] Due to uneven light intensity and temperature distribution during 3D printing, the curing degree of each layer will be different, which will affect the adhesion between layers and there will be an interlayer adhesion error value. Specifically, the formula is used: Calculate the interlayer adhesion error value, where E layer Indicates the interlayer bonding error value, σ j represents the actual interlayer adhesion, σ ideal represents the expected interlayer adhesion, j represents the number of layers, and takes values ​​of 1...m.

[0082] S144: Calculating a printing path deviation error value according to the printing configuration data;

[0083] During the 3D printing process, since the actual motion trajectory of the 3D printer does not completely match the designed or expected path, there is a printing path deviation error value. Specifically, the formula is used:

[0084] Calculate the printing path deviation error value, where E path Indicates the print path deviation error value. represents the expected path point coordinates, represents the actual path point coordinates, k represents the number of path points, and its value is 1...l.

[0085] S145: calculating a total error value according to the geometric error value, the material shrinkage compensation error value, the interlayer bonding error value and the printing path deviation error value;

[0086] The total error value is calculated according to the geometric error value, the material shrinkage compensation error value, the interlayer adhesion error value and the printing path deviation error value. Specifically, the formula is used:

[0087] E total =γ1E path +γ2E layer +γ3E shrink +γ4E geom Calculate the total error value, where E total Represents the total error value, E path Indicates the printing path deviation error value, E layer Indicates the interlayer bonding error value, E shrink Indicates the material shrinkage compensation error value, E geom Represents the geometric error value, γ1 represents the weight of the printing path deviation error value, γ2 represents the interlayer adhesion error value, γ3 represents the weight of the material shrinkage compensation error value, and γ4 represents the weight of the geometric error value.

[0088] S150: If the total error value does not exceed the set threshold, the quality of the 3D printed sample is determined to be qualified. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided;

[0089] If the total error value does not exceed the set threshold, the 3D printed sample is judged to be qualified. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided. The feedback information points out the area with the biggest problem in the 3D sample, which is conducive to the operator to make fixed-point modifications and improve work efficiency. Specifically, the formula is used: E total >Ω? true:max(A[a ij (DF ij )]) for comparison, where E total represents the total error value, Ω represents the set threshold, max(A[a ij (DF ij )] represents the defect value DF in the spectral feature data matrix A ij The highest subset a ij .

[0090] Embodiment 2

[0091] like Figure 2 As shown, the second embodiment of the present application provides a quality inspection system for 3D printing engineering materials, including:

[0092] Data collection and preprocessing module 21: collects data of the 3D printing process and preprocesses the 3D printed samples;

[0093] The 3D printing process includes 3D modeling, slicing, loading materials and printing, etc. The data of the 3D printing process includes 3D model data, slicing data, material property data and printing configuration data, etc. The 3D model data is usually generated by CAD (computer-aided design) software, and the geometric information of the 3D model is described by triangular facets. The slicing data contains the instruction information such as the printing path, filling mode, speed, temperature, etc. of each layer. The material property data includes the physical and chemical properties of the materials used, such as melting temperature, viscosity, curing rate, strength, etc. The printer configuration data involves the hardware settings of the printer, such as nozzle diameter, platform temperature, upper limit of printing speed, etc.

[0094] After 3D printing is completed, the sample is pre-processed to remove the support structure and ensure the flatness and cleanliness of the test surface.

[0095] Spectral characteristic data matrix module 22: using a high-resolution spectrometer to comprehensively scan the pre-treated sample, record the spectral characteristic data of different regions, and establish a spectral characteristic data matrix;

[0096] A high-precision spectrometer is used to scan the surface and near-surface of 3D printed parts, and the material composition, microstructure and potential defects (such as pores, cracks, etc.) are identified through the absorption and reflection characteristics of light of different wavelengths.

[0097] Spectral characteristic data include intensity characteristic value, baseline value and signal-to-noise ratio. The intensity characteristic value represents the light intensity detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to the specific wavelength of light; the baseline value is the lowest continuous background of the spectrum graph, which should be a horizontal line ideally; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise. A high signal-to-noise ratio means that the data is more reliable and the analysis results are more accurate.

[0098] The spectral characteristic data matrix is ​​used to record the spectral data of different regions. Specifically, the spectral characteristic data matrix is ​​expressed as: Where A represents the spectral feature data matrix, a ij represents the spectral feature data subset of the i-th row and j-th column area, m represents the maximum number of rows in the area, n represents the maximum number of columns in the area, and I ij represents the intensity characteristic value of the region in the i-th row and j-th column, B ij represents the baseline value of the region in the i-th row and j-th column, SNR ij represents the signal-to-noise ratio of the region in the i-th row and j-th column, DF ij Indicates the defect value, and the initial value is 0.

[0099] Defect value calculation module 23: analyzes the spectral feature data, calculates the defect values ​​of different regions, and updates the defect values ​​of different regions into the spectral feature data matrix;

[0100] Spectral feature data analysis and defect value calculation are specifically divided into the following sub-steps:

[0101] Data filtering module: filter the spectral feature data;

[0102] The filtering technology is applied to remove the random noise in the spectral feature data, and the iterative correction method is used to remove the non-uniform background baseline of the spectrum. The corrected intensity characteristic value is I corrected , the spectral data is normalized to eliminate the intensity difference and facilitate comparison. Specifically, the formula is used: Calculate the normalized intensity characteristic value, where I ref represents the normalized intensity eigenvalue, I corrected represents the corrected intensity characteristic value, max(I corrected ) indicates the maximum value of the corrected intensity.

[0103] Feature extraction module: extracts features from the spectral feature data after filtering to determine the key feature peak position and intensity;

[0104] After filtering, the spectral feature data is extracted to determine the key characteristic peak position and intensity. The characteristic peak is represented by λ peak .

[0105] Calculation module: Calculate the defect values ​​of different areas according to the position and intensity of the characteristic peaks extracted by features;

[0106] The defect value is divided into two defect indicators. Defect indicator one is the intensity characteristic value deviation score, and defect indicator two is the signal-to-noise ratio difference score. This application obtains a comprehensive defect score by weighted calculation of the two defect indicator scores. The comprehensive defect score is the defect value. The defect value is used to evaluate the quality of different areas of the 3D printed sample. It is specifically divided into the following sub-steps:

[0107] 1. Calculate the intensity eigenvalue deviation score;

[0108] Assume that the reference region is at the characteristic peak λ peak The intensity at I ref (λ peak ), the corresponding intensity of the area to be tested is I ij (λ peak ), define the intensity eigenvalue deviation fraction DF intensity To quantify the intensity difference, DF intensity Close to 0 indicates a good intensity match, and the greater the deviation, the higher the possibility of defects. Specifically, the formula is used: Calculate the intensity characteristic value deviation score of each region, where (DF intensity ) ij represents the intensity eigenvalue deviation score of the i-th row and j-th column area, I ref (λ peak ) indicates that the reference area is at the characteristic peak λ peak The intensity at I ij (λ peak ) indicates that the area to be tested is at the characteristic peak λ peak The strength of the place.

[0109] 2. Calculate the signal-to-noise ratio difference score;

[0110] Let the signal-to-noise ratio in the reference area be SNR ref , the signal-to-noise ratio of the area to be measured is SNR ij , define the signal-to-noise ratio difference score DF SNR , DF SNR It reflects the degree of decrease in the signal-to-noise ratio. The smaller the value, the closer the data quality is to the standard or reference state. Specifically, the formula is: Calculate the signal-to-noise ratio difference score of each region, where (DF SNR ) ijRepresents the signal-to-noise ratio difference score of the i-th row and j-th column area, SNR ref Indicates the signal-to-noise ratio in the reference area, SNR ij Indicates the signal-to-noise ratio of the area to be measured.

[0111] 3. Calculate the comprehensive defect score based on the intensity characteristic value deviation score and the signal-to-noise ratio difference score;

[0112] Taking into account the intensity deviation and signal-to-noise ratio difference, the comprehensive defect score DF is defined total , specifically, using the formula: Calculate the comprehensive defect score, where (DF total ) ij represents the comprehensive defect score of the region in row i and column j, (DF intensity ) ij represents the intensity eigenvalue deviation score of the i-th row and j-th column region, (DF SNR ) ij represents the signal-to-noise ratio difference score of the i-th row and j-th column area, represents the weight of the intensity eigenvalue deviation score, and β represents the weight of the signal-to-noise ratio difference score.

[0113] Matrix update module: updates the defect values ​​of different regions into the spectral feature data matrix;

[0114] The comprehensive defect score DF of the area calculated in the above steps is total Assign the defect value DF to the corresponding area in the spectral feature data matrix A ij , specifically, using the formula:

[0115] DF ij =DF total ,i=1,2,3,...,m,j=1,2,3,...,n are assigned, where DF ij represents the defect value of the area in the i-th row and j-th column, DF total represents the comprehensive defect score of the region in the i-th row and j-th column, m represents the maximum number of rows in the region, and n represents the maximum number of columns in the region, until the defect values ​​of all regions in the spectral feature data matrix are assigned.

[0116] Error value calculation module 24: calculates the total error value of the printing process according to the data of the 3D printing process;

[0117] In the 3D printing process, due to the deviations in the chemical properties and working conditions of the printing materials and 3D printers under different temperature environments, there will be certain errors in the finished products. This application calculates the error value in the 3D printing process by establishing a multi-factor integrated error evaluation model. The data collected during the 3D printing process, including three-dimensional model data, slice data, material property data, and printing configuration data, are used as input data of the multi-factor integrated error evaluation model. The process of calculating the process error value specifically includes the following sub-steps:

[0118] 1. Calculate the geometric error value based on the 3D model data;

[0119] There is a certain error between the geometric values ​​of the expected sample and the actual sample during 3D printing. Specifically, the formula is used:

[0120] Calculate the geometric error value, where E geom represents the geometric error value, represents the expected sample geometry fixed point coordinates, Represents the actual sample geometric fixed point coordinates, i represents the coordinate subscript, and its value is 1...n.

[0121] 2. Calculate the material shrinkage compensation error value based on the material property data;

[0122] The printing materials used in the 3D printing process will shrink due to actual conditions such as temperature, and there will be a certain error value. Specifically, the formula is used:

[0123] E shrink =V model ×(η shrink +ε temp ×(ΔT cooling )) Calculate the material shrinkage compensation error value, where E shrink Indicates the material shrinkage compensation error value, V model represents the sample volume, η shrink represents the basic shrinkage rate, ε temp Indicates the special expansion coefficient, ΔT cooling Indicates the change in the print temperature when it cools down to room temperature.

[0124] 3. Calculate the interlayer bonding error value based on the slice data;

[0125] Due to uneven light intensity and temperature distribution during 3D printing, the curing degree of each layer will be different, which will affect the adhesion between layers and there will be an interlayer adhesion error value. Specifically, the formula is used: Calculate the interlayer adhesion error value, where E layer Indicates the interlayer bonding error value, σ j represents the actual interlayer adhesion, σideal represents the expected interlayer adhesion, j represents the number of layers, and takes values ​​of 1...m.

[0126] 4. Calculate the printing path deviation error value according to the printing configuration data;

[0127] During the 3D printing process, since the actual motion trajectory of the 3D printer does not completely match the designed or expected path, there is a printing path deviation error value. Specifically, the formula is used:

[0128] Calculate the printing path deviation error value, where E path Indicates the print path deviation error value. represents the expected path point coordinates, represents the actual path point coordinates, k represents the number of path points, and its value is 1...l.

[0129] 5. Calculate the total error value based on the geometric error value, material shrinkage compensation error value, interlayer bonding error value and printing path deviation error value;

[0130] The total error value is calculated according to the geometric error value, the material shrinkage compensation error value, the interlayer adhesion error value and the printing path deviation error value. Specifically, the formula is used:

[0131] E total =γ1E path +γ2E layer +γ3E shrink +γ4E geom Calculate the total error value, where E total Represents the total error value, E path Indicates the printing path deviation error value, E layer Indicates the interlayer bonding error value, E shrink Indicates the material shrinkage compensation error value, E geom Represents the geometric error value, γ1 represents the weight of the printing path deviation error value, γ2 represents the interlayer adhesion error value, γ3 represents the weight of the material shrinkage compensation error value, and γ4 represents the weight of the geometric error value.

[0132] Feedback module 25: if the total error value does not exceed the set threshold, the quality of the 3D printed sample is determined to be qualified; if the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided;

[0133] If the total error value does not exceed the set threshold, the 3D printed sample is judged to be qualified. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided. The feedback information points out the area with the biggest problem in the 3D sample, which is conducive to the operator to make fixed-point modifications and improve work efficiency. Specifically, the formula is used: Etotal >Ω? true:max(A[a ij (DF ij )]) for comparison, where E total represents the total error value, Ω represents the set threshold, max(A[a ij (DF ij )] represents the defect value DF in the spectral feature data matrix A ij The highest subset a ij .

[0134] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A quality inspection method for 3D printing engineering materials, characterized in that: include: S110, collecting data of the 3D printing process and pre-processing the 3D printing samples; S120, using a high-resolution spectrometer to comprehensively scan the pre-treated sample, record spectral feature data of different regions, and establish a spectral feature data matrix; S130, analyzing the spectral feature data, calculating the defect values ​​of different regions, and updating the defect values ​​of different regions into the spectral feature data matrix, which is specifically divided into the following sub-steps: S131: filtering the spectral feature data; The filtering technology is applied to remove the random noise in the spectral feature data, and the iterative correction method is used to remove the non-uniform background baseline of the spectrum. The corrected intensity feature value is , the spectral data is normalized to eliminate the intensity difference and facilitate comparison. Specifically, the formula is used: Calculate the normalized intensity eigenvalue, where represents the normalized intensity eigenvalue, represents the corrected intensity eigenvalue, Indicates the maximum value of the intensity after correction; S132: extracting features from the filtered spectral feature data to determine key feature peak positions and intensities; After filtering, the spectral feature data is extracted to determine the key characteristic peak position and intensity. The characteristic peak is expressed as ; S133: Calculating defect values ​​of different regions according to the characteristic peak positions and intensities extracted from the features; The defect value is divided into two defect indicators. Defect indicator one is the intensity characteristic value deviation score, and defect indicator two is the signal-to-noise ratio difference score. The comprehensive defect score is obtained by weighted calculation of the two defect indicator scores. The comprehensive defect score is the defect value. The defect value is used to evaluate the quality of different areas of the 3D printed sample. It is specifically divided into the following sub-steps:

1. Calculate the intensity eigenvalue deviation score; Assume that the reference region is at the characteristic peak The intensity at , the corresponding intensity of the area to be measured is , defining the intensity eigenvalue deviation fraction To quantify the intensity difference, Close to 0 indicates a good intensity match, and the greater the deviation, the higher the possibility of defects. Specifically, the formula is used: Calculate the intensity eigenvalue deviation score of each region, where Indicates Line The intensity eigenvalue deviation fraction of the column region, Indicates that the reference area is in the characteristic peak The strength of Indicates that the area to be tested is at the characteristic peak The strength of the place; 2. Calculate the signal-to-noise ratio difference score; Assume the signal-to-noise ratio in the reference region is , the signal-to-noise ratio of the area to be measured is , defining the signal-to-noise ratio difference score , It reflects the degree of decrease in the signal-to-noise ratio. The smaller the value, the closer the data quality is to the standard or reference state. Specifically, the formula is: Calculate the signal-to-noise ratio difference score for each region, where Indicates Line The signal-to-noise ratio difference score of the column area, represents the signal-to-noise ratio in the reference region, Indicates the signal-to-noise ratio of the area to be measured; 3. Calculate the comprehensive defect score based on the intensity characteristic value deviation score and the signal-to-noise ratio difference score; Comprehensive defect score is defined by considering intensity deviation and signal-to-noise ratio difference , specifically, using the formula: Calculate the composite defect score, where Indicates Line The combined defect score of the column area, Indicates Line The intensity eigenvalue deviation fraction of the column region, Indicates Line The signal-to-noise ratio difference score of the column area, represents the weight of the intensity eigenvalue deviation score, The weight representing the signal-to-noise ratio difference score; S134: updating defect values ​​of different regions into the spectral feature data matrix; The comprehensive defect score of the area calculated in the above steps is Assign the defect value to the corresponding area in the spectral feature data matrix A , specifically, using the formula: Assign values, where Indicates Line Defect value of column area, Indicates Line The combined defect score of the column area, Indicates the maximum number of rows in the area. Indicates the maximum number of regional columns until all regional defect values ​​of the spectral feature data matrix are assigned; S140, calculating the total error value of the printing process according to the data of the 3D printing process, which is specifically divided into the following sub-steps: S141: Calculating geometric error values ​​according to the three-dimensional model data; Using the formula: Calculate the geometric error value, where represents the geometric error value, represents the expected sample geometry fixed point coordinates, Represents the actual sample geometric fixed point coordinates, Indicates the coordinate subscript, with a value of 1...n; S142: Calculating a material shrinkage compensation error value according to material property data; Using the formula: Calculate the material shrinkage compensation error value, where Indicates the material shrinkage compensation error value, represents the sample volume, represents the basic shrinkage rate, is the coefficient of thermal expansion, Indicates the change in the printing temperature when it cools down to room temperature; S143: Calculating the interlayer bonding error value according to the slice data; Using the formula: Calculate the interlayer adhesion error value, where Indicates the interlayer bonding error value, Indicates the actual interlayer adhesion, represents the expected interlayer adhesion, j represents the number of layers, and takes values ​​of 1...m; S144: Calculating a printing path deviation error value according to the printing configuration data; Using the formula: Calculate the printing path deviation error value, where Indicates the print path deviation error value. represents the expected path point coordinates, represents the actual path point coordinates, k represents the number of path points, and its value is 1...l; S145: Calculate the total error value according to the geometric error value, the material shrinkage compensation error value, the interlayer bonding error value and the printing path deviation error value. The calculation formula of the total error value is expressed as: ,in Represents the total error value, Indicates the print path deviation error value. Indicates the interlayer bonding error value, Indicates the material shrinkage compensation error value, represents the geometric error value, Indicates the weight of the printing path deviation error value, Indicates the interlayer adhesion error value, Indicates the weight of the material shrinkage compensation error value, Represents the weight of geometric error value; S150. If the total error value does not exceed the set threshold, the quality of the 3D printed sample is determined to be qualified. If the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided.

2. A quality inspection method for 3D printing engineering materials according to claim 1, characterized in that: The 3D printing process includes three-dimensional modeling, slicing, material loading and printing. The data of the 3D printing process includes three-dimensional model data, slicing data, material property data and printing configuration data.

3. A quality inspection method for 3D printing engineering materials according to claim 1, characterized in that: Spectral characteristic data include intensity characteristic value, baseline value and signal-to-noise ratio. The intensity characteristic value represents the intensity of light detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to light of a specific wavelength; the baseline value is the lowest continuous background of the spectrum graph, which should ideally be a horizontal line; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise.

4. A system for quality inspection of 3D printing engineering materials, used to perform a quality inspection method of 3D printing engineering materials as claimed in any one of claims 1 to 3, characterized in that: include: Data collection and preprocessing module: collects data from the 3D printing process and preprocesses 3D printed samples; Spectral feature data matrix module: Use a high-resolution spectrometer to comprehensively scan the pre-treated samples, record the spectral feature data of different areas, and establish a spectral feature data matrix; Defect value calculation module: analyzes the spectral feature data, calculates the defect values ​​of different areas, and updates the defect values ​​of different areas to the spectral feature data matrix; Error value calculation module: calculates the total error value of the printing process based on the data of the 3D printing process; if the total error value does not exceed the set threshold, the quality of the 3D printed sample is judged to be qualified; if the total error value exceeds the set threshold, the area with the highest defect value is retrieved from the spectral feature data matrix and feedback information is provided.

5. A system for quality inspection of 3D printing engineering materials as claimed in claim 4, characterized in that: The 3D printing process includes three-dimensional modeling, slicing, material loading and printing. The data of the 3D printing process includes three-dimensional model data, slicing data, material property data and printing configuration data.

6. A system for quality inspection of 3D printing engineering materials as claimed in claim 4, characterized in that: Spectral characteristic data include intensity characteristic value, baseline value and signal-to-noise ratio. The intensity characteristic value represents the intensity of light detected at a specific wavelength, and the intensity change reflects the absorption, reflection or transmission characteristics of the sample to light of a specific wavelength; the baseline value is the lowest continuous background of the spectrum graph, which should ideally be a horizontal line; the signal-to-noise ratio describes the ratio of the spectral signal to the background noise.

Images