A digital processing method for thermally deformed microstructure of titanium alloy

Abstract

The present invention relates to a method for digitally processing the microstructure of a hot-deformed titanium alloy, belonging to the field of titanium alloy forging processing. The method comprises the following steps: step 1: grayscale processing of the microstructure; step 2: binarization of the grayscale image; step 3: two-point statistics; step 4: image dimensionality reduction; and step 5: image reconstruction. The method uses the main components of the tissue data obtained in step 4 and the average coordinate change matrix as input, calculates the original two-point statistical information of the tissue, and uses a phase recovery algorithm to restore the tissue image from the two-point statistical information. The method outputs a binary image with the same characteristics as the original hot-deformed titanium alloy microstructure image. The present invention develops a method for digitally processing the microstructure of a hot-deformed titanium alloy. The method processes the microstructure morphology of the hot-deformed titanium alloy, captures the microstructure morphological features, and then uses image processing technology to significantly reduce the data volume while still retaining the microstructure features, converting the data into computer-readable digital information.

Description

Technical Field

[0001] The invention belongs to the field of titanium alloy forging processing, and in particular relates to a digital processing method for the thermal deformation microstructure of a titanium alloy. Background Art

[0002] Titanium alloys undergo significant microstructural changes during the forging process, and their mechanical properties are closely related to their microstructure. Therefore, studying the microstructural evolution of titanium alloys during hot deformation is essential. Currently, research on the microstructure of titanium alloys during hot deformation mainly relies on characterization and statistics of microstructural images obtained by corrosion. This is achieved by manually observing image morphology differences and comparing image features such as phase composition, phase content, and grain size. This significantly limits the rapid development of titanium alloy material research and forging process applications. The continuous advancement of computer hardware and software development and the continuous improvement of computing power have greatly promoted the development of material research and forging processing. However, the difficulty in digitally processing the microstructure of titanium alloys during hot deformation seriously limits the computational power and accuracy of computer processing in titanium alloy forging process research. This is mainly due to the difficulty in digitally representing the different phase compositions, phase content, and size characteristics within the titanium alloy microstructure. Existing computer processing technology can only display microstructural images, but cannot perform computational predictions and dynamic display of microstructural images during the forging process, which directly affects the accuracy of forging performance predictions. Therefore, digital processing of the microstructure morphology of titanium alloy thermal deformation has important social benefits and economic value for accelerating the development and forging application of titanium alloy materials.

[0003] Digital processing of the microstructure morphology of titanium alloys during thermal deformation is a difficult problem faced in computer calculations for titanium alloy material development and forging applications. By digitally processing the microstructure morphology of titanium alloys, the connection between forging process and mechanical properties can be established quickly and accurately, but how to digitize microstructure images through computer technology is an extremely important link in promoting the calculation of titanium alloy material development and forging process applications. In order to improve the efficiency of computer digitization of microstructure images of titanium alloys during thermal deformation and reduce the loss of microstructure morphological features of titanium alloys, the present invention proposes a method for digitizing the microstructure of titanium alloys during thermal deformation, which can effectively capture the microstructure morphological features, thereby converting them into digital information that can be recognized by computers, meeting the needs of fast and accurate calculation of titanium alloy material development and forging processes. Summary of the Invention

[0004] Technical issues to be solved:

[0005] In order to avoid the shortcomings of the existing technology, the present invention performs computer recognition on the microstructure morphology of thermal deformation of titanium alloys. A digital processing method for the microstructure of thermal deformation of titanium alloys has been developed. The method processes the microstructure morphology of thermal deformation of titanium alloys, captures the microstructure morphology characteristics, and then uses image processing technology to greatly reduce the amount of data while still retaining the microstructure characteristics and converting it into digital information that can be recognized by computers.

[0006] The technical solution of the present invention is: a method for digital processing of thermally deformed microstructure of titanium alloy, characterized by the following specific steps:

[0007] Step 1: Microstructure grayscale processing: Use programming software to read the thermal deformation microstructure images of different titanium alloys and assign grayscale values ​​to the images; eliminate the color information of the microstructure images;

[0008] Step 2: Grayscale image binarization: Binarize the grayscale image obtained in step 1 to make the α and β phases in the grayscale image in step 1 into a high-contrast black and white binary image.

[0009] Step 3: Two-point statistics: The two-point statistical calculation of the tissue image is performed by randomly placing vectors in the binarized image in step 2. The α and β phases in the binarized image are captured by the two-point statistical function to obtain random vector two-point statistical information of different microstructural morphological characteristics including phase content, phase morphology, phase distribution, and phase size;

[0010] Step 4: Image dimensionality reduction: The two-point statistical information of the microstructure in step 3 is transformed into coordinates through principal component analysis. The coordinate axes of the new coordinate system are arranged in sequence according to the average variance of the two-point statistics of different tissues. The first few coordinate component values ​​reflecting the main differences in different microstructure characteristics and the average coordinate change matrix are retained to achieve digital dimensionality reduction and storage of tissue images.

[0011] Step 5: Image reconstruction: Using the main components of the tissue data and the average coordinate change matrix obtained in step 4 as input, calculate the original two-point statistical information of the tissue, and restore the tissue image from the two-point statistical information using the phase recovery algorithm, and output a binary image with the same features as the original titanium alloy thermal deformation microstructure image.

[0012] A further technical solution of the present invention is: in step 1,

[0013] The grayscale of the image is assigned using the objective function P = (0.28-0.31)R + (0.59-0.6)G + (0.1-0.15)B, where R, G, and B are the color components of the image pixels, and the sum of the coefficients is 1.

[0014] A further technical solution of the present invention is: in step 2, a threshold processing function is used Perform binarization processing, where O(P) is the pixel value of the binary image and T is the threshold value for setting image clarity.

[0015] A further technical solution of the present invention is: in step 4, the values ​​of the first 2 to 4 coordinate components reflecting the main differences of different microscopic morphological features are retained.

[0016] Beneficial effects

[0017] The beneficial effects of the present invention are as follows: the digital processing method of the thermal deformation microstructure of titanium alloy of the present invention adopts the two-point statistical method and the principal component analysis method to process the thermal deformation microstructure image of titanium alloy, which effectively solves the problems of large deviation, low efficiency, and difficulty in feature recognition existing in the existing processing technology.

[0018] Compared to existing microstructure morphology processing methods, this method uses programming software and an objective function to read the titanium alloy thermal deformation microstructure image and perform grayscale assignment. This effectively eliminates color information in the microstructure image and eliminates morphology differences caused by different microstructure preparation methods, providing a high-quality image for step 2 and avoiding poor digital processing results caused by improper input of the original image. Step 2 uses a threshold processing function to present the grayscale image as a high-contrast black and white binary image, making the α and β phase content, morphology, and distribution characteristics in the microstructure more obvious, ensuring efficient capture of microstructure features in step 3. Step 3 uses a random vector placement method to perform two-point statistical calculations on the microstructure image, effectively capturing different microstructure morphology characteristics such as the α and β phase content, morphology, distribution, and size, and representing them as computer-readable digital information data, providing the necessary basic data for microstructure digital processing. In step 4, the statistical average variances of two points of different tissues are arranged in sequence using the principal component analysis method to extract the values ​​of the main coordinate components that affect the differences in tissue morphology. The main coordinate component values ​​that can change the tissue morphology and the average coordinate change matrix are obtained, achieving free transformation of the tissue image and data dimensionality reduction storage. This provides the necessary microstructure digital data for efficient computer calculation of the forging process, while avoiding the complex and disordered accumulation of digital information data captured in step 3. In step 5, the stored main coordinate components and average coordinate change matrix are used as input to restore the tissue image by calculating two-point statistical information and using a phase recovery algorithm. At the same time, based on the different input values ​​of the coordinate components and change matrices caused by different forging deformation parameters, binary images with different tissue morphological characteristics are output, realizing dynamic image display of forging process microstructure calculation and prediction.

[0019] This method is completely different from the traditional microstructure computer processing method. It can digitize the microstructure image of titanium alloy forging thermal deformation, and has practical application value for the prediction and calculation of forging deformation process parameters, microstructure, mechanical properties, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 (a) Grayscale image, (b) black and white binary image;

[0021] Figure 2 (a) Random vector two-point statistical information, (b) principal component analysis coordinate change diagram;

[0022] Figure 3 (a) Two-point statistical average variance arrangement, (b) digital dimensionality reduction storage;

[0023] Figure 4 (a) Original two-point statistical information, (b) Output binary image;

[0024] Figure 5 (a) TC6 titanium alloy grayscale image, (b) black and white binary image;

[0025] Figure 6 (a) Two-point statistical information of random vector of TC6 titanium alloy, (b) principal component analysis coordinate change diagram;

[0026] Figure 7 (a) Two-point statistical average variance arrangement of TC6 titanium alloy, (b) digital dimensionality reduction storage;

[0027] Figure 8 (a) Original two-point statistical information of TC6 titanium alloy, (b) output binary image;

[0028] Figure 9 (a) TC29 titanium alloy grayscale image, (b) black and white binary image;

[0029] Figure 10 (a) Two-point statistical information of random vector of TC29 titanium alloy, (b) coordinate change diagram of principal component analysis;

[0030] Figure 11 (a) Two-point statistical mean variance arrangement of TC29 titanium alloy, (b) digital dimensionality reduction storage;

[0031] Figure 12 (a) Original two-point statistical information of TC29 titanium alloy, (b) output binary image. DETAILED DESCRIPTION

[0032] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

[0033] This embodiment provides a method for digitally processing the microstructure of a titanium alloy subjected to thermal deformation, the method comprising the following steps:

[0034] (1) Image grayscale processing: Python, Matlab and other programming software are used to read the thermal deformation microstructure image of titanium alloy, and the grayscale processing function P = (0.28 ~ 0.31) R + (0.59 ~ 0.6) G + (0.1 ~ 0.15) B is used to select the appropriate coefficient value and compilation degree to assign grayscale to the microstructure image, eliminate the image color information, and obtain the grayscale image ( Figure 1 (a)). R, G, and B are the color components of the image pixels, and their coefficients sum to 1.

[0035] (2) Image binarization: For the grayscale image ( Figure 1 In (a), the threshold processing function is used to perform binarization processing in programming software such as python and matlab, so that the α and β phases are high-contrast black and white binary images ( Figure 1 (b));

[0036] (3) Two-point statistics: Take the binarized image ( Figure 1 The random vector placement method in (b) is used to perform two-point statistical calculations on the tissue image. The two phases α and β in the binary image are captured by the two-point statistical function to obtain random vector two-point statistical information containing different microstructure morphological characteristics such as phase content, phase morphology, phase distribution, and phase size ( Figure 2 (a)).

[0037] (4) Image dimensionality reduction: Two statistical information of step 3 ( Figure 2 In (a), the coordinate changes were performed by principal component analysis ( Figure 2 (b)), so that the coordinate axes of the new coordinate system are arranged in order according to the statistical average variance of two points of different organizations ( Figure 3 In (a), the first 2 to 4 coordinate components reflecting the main differences in different microscopic morphological features and the average coordinate change matrix are retained to achieve digital dimensionality reduction storage of tissue images ( Figure 3 (b)).

[0038] (5) Image reconstruction: Using the main components of the tissue data obtained in step 4 and the average coordinate change matrix as input, calculate the original two-point statistical information of the tissue ( Figure 4 In (a), the tissue image is restored from the two-point statistical information by the phase recovery algorithm, and a binary image with the same characteristics as the original titanium alloy thermal deformation microstructure image can be output ( Figure 4 (b)).

[0039] This method realizes the digital processing and storage of the microstructure of titanium alloy thermal deformation, solves the problem of computer recognition of microstructure images in the study of thermal deformation of titanium alloy forging, and provides an efficient computing means for the computer processing technology of titanium alloy material research and development and engineering application of forging technology.

[0040] Example 1:

[0041] This embodiment provides a method for digitally processing the microstructure of a titanium alloy subjected to thermal deformation, including the following steps:

[0042] Step 1: Image grayscale processing: Use Python to read the thermal deformation microstructure image of TC6 titanium alloy, and use the grayscale processing function P = 0.29R + 0.58G + 0.13B to assign grayscale to the microstructure image, such as Figure 5 As shown in (a);

[0043] Step 2: Binarize the image, use the threshold processing function, and write a program in Python to binarize the image ( Figure 5 (a) is binarized, and the α and β phases are high-contrast black and white binary images ( Figure 5 In (b), it can be seen that the microstructure features of the image are highlighted;

[0044] Step 3: Two-point statistics, use the two-point statistics function to write a program in Python to binarize the image ( Figure 5 The α and β phases in (b) are captured to obtain two-point statistical information of random vectors containing different microstructural morphological characteristics such as phase content, phase morphology, phase distribution, and phase size ( Figure 6 (a));

[0045] Step 4: Image dimensionality reduction, using principal component analysis, write a program in Python to analyze the two-point statistical information graph in step 3 ( Figure 6 (a)) to change the coordinates ( Figure 6 (b)), so that the coordinate axes of the new coordinate system are arranged in order according to the statistical average variance of two points of different organizations ( Figure 7 In (a), the first 2 to 4 coordinate components reflecting the main differences in different microscopic morphological features and the average coordinate change matrix are retained to achieve digital dimensionality reduction storage of tissue images ( Figure 7 (b));

[0046] Step 5: Image reconstruction, using the main components of the tissue data obtained in step 4 and the average coordinate change matrix as input, calculate the original two-point statistical information of the tissue ( Figure 8 In (a), the tissue image is restored from the two-point statistical information using the phase recovery algorithm, and a binary image with the same characteristics as the original TC6 titanium alloy thermal deformation microstructure image is output ( Figure 8 (b)). It can be seen that this method can effectively digitize the microstructure image of titanium alloy thermal deformation.

[0047] Example 2:

[0048] Step 1: Image grayscale processing: Use Python to read the thermal deformation microstructure image of TC29 titanium alloy, and use the grayscale processing function P = 0.3R + 0.59G + 0.11B to assign grayscale to the microstructure image, such as Figure 9 As shown in (a);

[0049] Step 2: Binarize the image, use the threshold processing function, and write a program in Python to binarize the image ( Figure 9 (a) is binarized, and the α and β phases are high-contrast black and white binary images ( Figure 9 In (b), it can be seen that the microstructure features of the image are highlighted;

[0050] Step 3: Two-point statistics, use the two-point statistics function to write a program in Python to binarize the image ( Figure 9 The α and β phases in (b) are captured to obtain two-point statistical information of random vectors containing different microstructural morphological characteristics such as phase content, phase morphology, phase distribution, and phase size ( Figure 10 (a));

[0051] Step 4: Image dimensionality reduction, using principal component analysis, write a program in Python to analyze the two-point statistical information graph in step 3 ( Figure 10 (a)) to change the coordinates ( Figure 10 (b)), so that the coordinate axes of the new coordinate system are arranged in order according to the statistical average variance of two points of different organizations ( Figure 11 In (a), the first 2 to 4 coordinate components reflecting the main differences in different microscopic morphological features and the average coordinate change matrix are retained to achieve digital dimensionality reduction storage of tissue images ( Figure 11 (b));

[0052] Step 5: Image reconstruction, using the main components of the tissue data obtained in step 4 and the average coordinate change matrix as input, calculate the original two-point statistical information of the tissue ( Figure 12 In (a), the tissue image is restored from the two-point statistical information using the phase recovery algorithm, and the output is a binary image with the same characteristics as the original TC29 titanium alloy thermal deformation microstructure image ( Figure 12 (b)). It can be seen that this method can effectively digitize the microstructure image of titanium alloy thermal deformation.

[0053] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention without departing from the principles and purpose of the present invention.

Claims

1. A method for digital processing of thermally deformed microstructure of titanium alloy, characterized in that The specific steps are as follows: Step 1: Microstructure grayscale processing: Use programming software to read the thermal deformation microstructure images of different titanium alloys and assign grayscale values ​​to the images; eliminate the color information of the microstructure images; Step 2: Grayscale image binarization: Binarize the grayscale image obtained in step 1 to make the α and β phases in the grayscale image in step 1 into a high-contrast black and white binary image. Step 3: Two-point statistics: The two-point statistical calculation of the tissue image is performed by randomly placing vectors in the binarized image in step 2. The α and β phases in the binarized image are captured by the two-point statistical function to obtain random vector two-point statistical information of different microstructural morphological characteristics including phase content, phase morphology, phase distribution, and phase size; Step 4: Image dimensionality reduction: The two-point statistical information of the random vector of the microscopic tissue morphology characteristics in step 3 is transformed into coordinates through principal component analysis. The coordinate axes of the new coordinate system are arranged in sequence according to the average variance of the two-point statistics of different tissues. The first few coordinate component values ​​reflecting the main differences in different microscopic morphology characteristics and the average coordinate change matrix are retained to achieve digital dimensionality reduction and storage of tissue images. Step 5: Image reconstruction: Using the digital principal components of the tissue image and the average coordinate change matrix obtained in step 4 as input, calculate the original two-point statistical information of the tissue, and restore the tissue image from the two-point statistical information using the phase recovery algorithm, and output a binary image with the same features as the original titanium alloy thermal deformation microstructure image.

2. The method for digital processing of hot-deformed microstructure of titanium alloy according to claim 1, characterized in that: In step 1, the grayscale of the image is assigned using the objective function P=(0.28-0.31)R+(0.59-0.6)G+(0.1-0.15)B, where R, G, and B are the color components of the image pixels, and the sum of the coefficients is 1.

3. The method for digital processing of hot-deformed microstructure of titanium alloy according to claim 1, characterized in that: In step 2, the threshold processing function is used , perform binarization processing, where O(P) is the pixel value of the binary image and T is the threshold value for setting image clarity.

4. The method for digital processing of hot-deformed microstructure of titanium alloy according to claim 1, characterized in that: In step 4, the values ​​of the first 2 to 4 coordinate components reflecting the main differences of different microscopic morphological features are retained.

Images