A method for constructing a representative volume element of a workpiece surface layer

Abstract

The present application relates to the research of structural fatigue life, in particular to a method for predicting the fatigue life of a complex structural part under multi-axial load; the present application firstly carries out tensile and fatigue experiments of the structural material to obtain its macroscopic material properties; then measures the polycrystalline indentation modulus of the structural material to obtain the indentation modulus of different grain orientations and the single-crystal elastic coefficient and other microscopic material properties; finally, a cross-scale finite element analysis model of the structural part is established to predict the multi-axial fatigue life of the structural part; the present application considers the actual structural characteristics of the bearing part, and the description of its fatigue behavior is more accurate; by simulating the actual loading mode of the structural part, the fatigue life of the structural part under complex load can be predicted, and the error caused by simplifying the load can be reduced; by using the submodel nesting method, the macrostructure and microstructure of the part to be predicted are comprehensively considered, which can greatly improve the prediction accuracy.

Description

Technical Field

[0001] This invention relates to the study of the microstructure of the surface layer of a workpiece, and specifically to a method for constructing representative volume elements of the surface layer. Background Technology

[0002] With the development of finite element method (FEM) technology, simulating representative volume elements based on crystal plasticity has become the preferred method for predicting workpiece fatigue life. However, current representative volume element models only include information such as grain size and orientation. The surface integrity parameters of the sample, such as residual stress, surface morphology, and microstructure, significantly affect the fatigue life of the workpiece, leading to a large discrepancy between the predicted and actual fatigue life. Therefore, establishing representative volume elements that incorporate surface integrity information remains a key focus of current research. Patent CN201710607748.1 discloses a finite element modeling method based on SEM-EBSD images of real microstructure, which establishes representative volume elements of the material by processing SEM and EBSD images of the material; Patent CN202111108049.5 discloses a crystal plasticity finite element modeling and simulation method, which establishes representative volume elements containing average geometric orientation information of grains by using Dream.3D software and Python scripts; Patent CN201710089897.3 discloses a method for constructing a representative volume element model of Ni3Al-based alloy, which generates representative volume elements by statistically analyzing the distribution law of micro-geometric feature parameters of the alloy microstructure. However, these representative volume element methods have the following shortcomings: ① The constructed representative volume element model differs significantly from the actual crystal structure of the material, resulting in low accuracy in finite element simulation studies; ② The established representative volume element model only includes microscopic information about the material's interior, while fatigue cracks often initiate on the surface of the sample, making it less targeted; ③ The influence of the integrity parameters of the sample's machined surface is not considered, leading to a large discrepancy between the predicted fatigue life and the actual fatigue life. These shortcomings greatly complicate fatigue life prediction. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method for constructing a representative volume element of a workpiece surface layer. This method obtains the surface properties of a workpiece by detecting the surface morphology, residual stress, and surface integrity data of the processed sample, in order to establish a representative volume element of the surface layer.

[0004] To solve the above technical problems, the present invention adopts the following technical solution, and the present invention adopts the following steps:

[0005] ① Construct a workpiece coordinate system to obtain surface integrity data such as workpiece surface morphology, residual stress, and microstructure;

[0006] ② Generate representative volume elements containing information on surface grain gradient distribution and surface roughness based on the workpiece surface morphology and surface microstructure;

[0007] 201 Extract grain size distribution and grain boundary angle distribution data of the workpiece surface strengthening layer and matrix layer to generate representative volume elements with gradient grain structure;

[0008] 202. Based on the extracted height distribution information of the machined surface, calculate the root mean square roughness value.

[0009]

[0010]

[0011] In the formula, z(x,y) is the actual measured height of the sample surface. S is the zero average surface height value. q This represents the root mean square roughness value of the surface.

[0012] The profile spacing and aspect ratio of the workpiece's machined surface texture are characterized using the autocorrelation function and autocorrelation length:

[0013]

[0014] l x,ACF (τ x )=∫f ACF (τ x ,τ y )dτ y (4)

[0015] l y,ACF (τ y )=∫f ACF (τ x ,τ y )dτ x (5)

[0016]

[0017]

[0018]

[0019] In the formula, τ x and τ y For planar translation coordinates; l x,ACF (τ x ) and l y,ACF (τ y ) are the autocorrelation coefficients in the x and y directions, respectively; β x and β y λ represents the autocorrelation lengths in the x and y directions, respectively;s R is the filter length; ACL The ratio of autocorrelation lengths;

[0020] 203. Using the surface roughness data obtained in 202, a two-dimensional exponential autocorrelation function is established to simulate the surface morphology of the machined surface:

[0021]

[0022]

[0023] Where, β x ′ and β y ′ are the corrected autocorrelation lengths in the x and y directions, respectively, and R(x,y) is the autocorrelation function of the generated surface; 203 The power spectral density function is obtained by performing a Fourier transform on the obtained autocorrelation function;

[0024] ③ Import the residual stress data of the workpiece surface into the generated representative volume element to establish the final representative volume element model of the surface.

[0025] Step ① includes: 101 Measuring the surface morphology of the workpiece and extracting surface height distribution information;

[0026] 102. Measure the residual stress data of the workpiece surface layer in the x and y directions along the layer depth;

[0027] 103. Wire-cut material on the workpiece surface, prepare a sample, and capture an electron backscatter diffraction image containing the workpiece's machining reinforcement layer.

[0028] Step ③ includes: 301 using the depth z from the surface as the independent variable, fitting the coefficient of thermal expansion used to impart residual stress to the volume element using a polynomial function:

[0029] a(z) = a1z 4 +a2z 3 +a3z 2 +a4z+a5 (11)

[0030] In the formula, a1, a2, a3, a4 and a5 are the polynomial coefficients;

[0031] 302 The representative volume element generated in step ② is numerically iterated and fitted to obtain the coefficient of thermal expansion as the material property of the representative volume element.

[0032] Step 203 generates a filtered surface through two-stage filtering and inverse Fourier transform, and imports the generated filtered surface into a representative volume element to generate a representative volume element containing surface roughness information.

[0033] Compared with existing technologies, the present invention has the following advantages: The representative volume element established by measuring the integrity data of the workpiece's machined surface is more consistent with the actual machined surface state of the workpiece; the surface morphology modeling extracts the surface roughness and autocorrelation lengths in two directions of the sample, which is more consistent with the sample's machined morphology than the previous model established solely based on roughness; the established representative volume element mainly targets the surface performance of the workpiece, making it more accurate and scientific for simulating surface fatigue failure of the workpiece, and more accurate in predicting lifespan compared to traditional representative volume element models. Attached Figure Description

[0034] Figure 1 This is a diagram showing the surface height distribution of the machined part.

[0035] Figure 2 This is a diagram showing the height distribution on the surface of the ground part.

[0036] Figure 3 For axial residual stress in turned and ground parts;

[0037] Figure 4 For circumferential residual stress in machined and ground parts;

[0038] Figure 5 Electron backscatter diffraction (EBSD) image of a machined part;

[0039] Figure 6 Electron backscatter diffraction (EBSD) image of the ground part;

[0040] Figure 7 Modeling the microstructure of the surface layer representing volume elements;

[0041] Figure 8 Model the surface morphology of the surface layer representing the volume element;

[0042] Figure 9 The flowchart shows the generation process of the surface-representing volume element. Detailed Implementation

[0043] The present invention will now be described in detail with reference to the accompanying drawings and specific examples. The main steps are as follows:

[0044] ① Measure surface integrity data such as workpiece surface morphology, residual stress, and microstructure:

[0045] 101 Given AISI 9310 fatigued parts that require surface representative volume elements to be established, which are respectively turned and ground, scan their surface morphology and extract the surface height distribution information after surface correction, as follows: Figure 1 and Figure 2 As shown.

[0046] 102. Take machined and ground samples, perform radial electropolishing, and measure the residual stress data of the axial and axial depth layers on the sample surface. Figure 3 and Figure 4 As shown.

[0047] 103. Take turned and ground samples respectively, wire-cut a portion of the material on their surfaces, prepare samples, and photograph electron backscatter diffraction (EBSD) images including the processing reinforcement layer of the samples. Calculate their grain size distribution curves and grain boundary angle distribution curves, such as... Figure 5 and Figure 6 As shown.

[0048] ② Generate representative volume elements containing information on surface grain gradient distribution and surface roughness based on the surface morphology and surface microstructure of the sample;

[0049] 201. Grain size distribution and grain boundary angle distribution data of the surface strengthening layer and matrix layer of turned and ground samples were extracted respectively. Parameters such as grain type, lattice number, and resolution were set to generate representative volume elements containing gradient grain structures, as detailed below. Figure 7 As shown.

[0050] 202. Calculate the root mean square roughness value based on the extracted height distribution information of the machined surface of the sample.

[0051]

[0052]

[0053] In the formula, z(x,y) is the actual measured height of the sample surface. S is the zero average surface height value. q This represents the root mean square roughness value of the surface.

[0054] Furthermore, to characterize the contour spacing and aspect ratio of the workpiece's machined surface texture, an autocorrelation function and an autocorrelation length are introduced:

[0055]

[0056] l x,ACF (τ x )=∫f ACF (τ x ,τ y )dτ y (4)

[0057] l y,ACF (τ y )=∫f ACF (τ x ,τ y )dτ x (5)

[0058]

[0059]

[0060]

[0061] In the formula, τ x and τ y For planar translation coordinates; l x,ACF (τ x ) and l y,ACF (τ y ) are the autocorrelation coefficients in the x and y directions, respectively; β x and β y λ represents the autocorrelation lengths in the x and y directions, respectively; s R is the filter length; ACL It is the autocorrelation length ratio, used to measure the anisotropy of the profile spacing.

[0062] Furthermore, using the obtained surface roughness data, a two-dimensional exponential autocorrelation function is established to simulate the surface morphology of the machined surface:

[0063]

[0064]

[0065] Where, β x ′ and β y '' represents the corrected autocorrelation lengths in the x and y directions, respectively. R(x,y) is the autocorrelation function of the generated surface. The above parameters for both ground and turned parts are determined by the average values ​​calculated from five measurement regions, as shown in Table 1.

[0066] Table 1. Surface roughness parameters for turning and grinding.

[0067]

[0068] 203. The power spectral density function is obtained by performing Fourier transform on the autocorrelation function of the obtained turned and ground samples.

[0069] Furthermore, a filter surface is generated through two-stage filtering and inverse Fourier transform;

[0070] Furthermore, the generated filtered surface is imported into the representative volume element to generate a representative volume element containing surface roughness information, specifically as follows: Figure 8 As shown.

[0071] ③ Import the residual stress data of the sample surface into the generated representative volume element to establish the final representative volume element model of the surface. 301 Using the depth from the surface as the independent variable, a strain is fitted in the form of a polynomial function to impart residual stress to the representative volume element of the machined surface of the turned and ground samples:

[0072] a(z) = a1z 4 +a2z 3 +a3z 2 +a4z+a5 (11)

[0073] In the formula, a1, a2, a3, a4 and a5 are the polynomial coefficients;

[0074] 302. The representative volume elements of the previously generated turned and ground samples are brought into the finite element software, and the strain obtained by fitting is used as its load.

[0075] Furthermore, initial boundary conditions are set for the representative volume element to induce stress in the representative volume element due to strain.

[0076] Furthermore, the boundary conditions in the z-direction are removed, and natural springback is simulated to obtain a representative volume element of the surface layer containing residual stress. This element is then compared with the actual residual stress measured on the desired turned and ground parts. If the error is greater than the operational error, the coefficient of thermal expansion is corrected, and the above steps are repeated until the generated residual stress matches the actual residual stress. A specific implementation plan is as follows: Figure 9 As shown, the final result is a surface-representative volumetric element model of the workpiece machining surface integrity data, including residual stress, surface morphology, and microstructure.

[0077] The embodiments described above are merely preferred embodiments of the present invention, and not an exhaustive list of all possible implementations of the present invention. Any obvious modifications made by those skilled in the art without departing from the principles and spirit of the present invention should be considered to be included within the scope of protection of the claims of the present invention.

Claims

1. A method for constructing representative volume elements on the surface of a workpiece, characterized in that... Includes the following steps: ① Construct a workpiece coordinate system to obtain surface integrity data such as workpiece surface morphology, residual stress, and microstructure; ② Generate representative volume elements containing information on surface grain gradient distribution and surface roughness based on the workpiece surface morphology and surface microstructure; 201 Extract grain size distribution and grain boundary angle distribution data of the workpiece surface strengthening layer and matrix layer to generate representative volume elements with gradient grain structure; 202. Based on the extracted height distribution information of the machined surface, calculate the root mean square roughness value. In the formula, z(x,y) is the actual measured height of the sample surface. S is the zero average surface height value. q This represents the root mean square roughness value of the surface. The profile spacing and aspect ratio of the workpiece's machined surface texture are characterized using the autocorrelation function and autocorrelation length: l x,ACF (t x )=∫f ACF (t x ,t y )dτ y (4) l y,ACF (t y )=∫f ACF (t x ,t y )dτ x (5) In the formula, τ x and τ y For planar translation coordinates; l x,ACF (τ x ) and l y,ACF (τ y ) are the autocorrelation coefficients in the x and y directions, respectively; β x and β y λ represents the autocorrelation lengths in the x and y directions, respectively; s R is the filter length; ACL The ratio of autocorrelation lengths; 203. Using the surface roughness data obtained in 202, a two-dimensional exponential autocorrelation function is established to simulate the surface morphology of the machined surface: Where, β x ′ and β y ′ are the corrected autocorrelation lengths in the x and y directions, respectively, and R(x,y) is the autocorrelation function of the generated surface; 203 The power spectral density function is obtained by performing a Fourier transform on the obtained autocorrelation function; ③ Import the residual stress data of the workpiece surface into the generated representative volume element to establish the final representative volume element model of the surface.

2. The method for constructing a representative volume element on the surface of a workpiece according to claim 1, characterized in that... Step ① includes:

101. Measure the surface morphology of the workpiece and extract surface height distribution information; 102. Measure the residual stress data of the workpiece surface layer in the x and y directions along the layer depth; 103. Wire-cut material on the workpiece surface, prepare a sample, and capture an electron backscatter diffraction image containing the workpiece's machining reinforcement layer.

3. A method for constructing representative volume elements on the surface of a workpiece according to claim 1 or 2, characterized in that... Step ③ includes:

301. Using the depth z from the surface as the independent variable, a polynomial function is used to fit the coefficient of thermal expansion used to assign residual stress to the volume element: a(z)=a1z 4 +a2z 3 +a3z 2 +a4z+a5 (11) In the formula, a1, a2, a3, a4 and a5 are the polynomial coefficients; 302 The representative volume element generated in step ② is numerically iterated and fitted to obtain the coefficient of thermal expansion as the material property of the representative volume element.

4. The method for constructing a representative volume element on the surface of a workpiece according to claim 1, characterized in that: Step 203 generates a filtered surface through two-stage filtering and inverse Fourier transform, and imports the generated filtered surface into a representative volume element to generate a representative volume element containing surface roughness information.

Images