Method for predicting contact resistance between contact surfaces based on finite element
The contact resistance is predicted using the finite element method, which solves the problem of inaccurate contact resistance prediction caused by noise data interference, optimizes product design, improves system reliability and reduces costs.
Patent Information
- Application Number
- CN202510907629.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-02
- Publication Date
- 2025-10-24
AI Technical Summary
Existing contact resistance prediction technology is affected by noise data, resulting in data distortion and affecting prediction accuracy. Commonly used algorithms have limitations, making it difficult to optimize materials and geometries during the product design stage to reduce poor contact failures.
A finite element-based contact surface contact resistance prediction method was adopted. Gaussian distribution white noise sequences were generated by MATLAB. COMSOL and ABAQUS software were used for parametric modeling and finite element calculation to predict the contact area and resistance.
Optimize materials and geometries during the product design phase to reduce poor contact failures, improve system reliability, reduce costs, and achieve stability and reliability of contact resistance.
Smart Images
Figure CN120832791A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of electrical contact, more particularly to a contact resistance prediction method between contact surfaces based on finite elements. BACKGROUND
[0002] The prediction technology of contact resistance plays a crucial role in the field of electrical engineering and materials science. This technology can timely detect structural defects and damage in electrical connections, assess the potential risks to the safe operation of electrical systems, and monitor and manage the entire life cycle of electrical connections. Currently, in terms of hardware equipment, such as contact resistance measuring instruments and sensors, has been relatively mature. However, the contact resistance prediction analysis technology is still a short board in the practical application of this field, because in the process of data collection, it will be affected by many factors, leading to data distortion.
[0003] In 2018, the relevant standard introduced by the Engineering Construction Association classifies several categories of abnormal data types for contact resistance, including strong noise interference, long-period interference, no change in data, data beyond the reasonable range, strong and weak noise interference, drift, data loss, and data jump points. During the use of monitoring equipment, it is inevitable to be affected by circuit, system electromagnetic radiation, external environment and load, etc., which makes the monitoring signal data appear noise interference, and then leads to the deviation or error of the measured value relative to the true value. This kind of abnormal data is called noise data.
[0004] When evaluating the accuracy of contact resistance prediction, two indicators, signal-to-noise ratio (SNR) and root mean square error (RMSE), are usually used. The larger the SNR value, the smaller the RMSE, and the higher the prediction accuracy. To handle noise data anomalies, data cleaning methods can be used.
[0005] Currently, the commonly used contact resistance prediction algorithms include physical model-based prediction, statistical model-based prediction, and machine learning methods. However, these methods have some shortcomings and certain limitations in their scope of application. SUMMARY
[0006] Therefore, the present application provides a contact resistance prediction method between contact surfaces based on finite elements.
[0007] To achieve the above purpose, the present application adopts the following technical solutions:
[0008] A contact resistance prediction method between contact surfaces based on finite elements, comprising the following steps:
[0009] Confirm the initial parameters for generating the contact resistance prediction three-dimensional model, and generate a Gaussian distribution white noise two-dimensional random sequence in MATLAB, calculate the Fourier transform, and obtain the first function of the surface height distribution through filtering processing;
[0010] Given the second function that conforms to the law of the real surface, and the first function to get the final height distribution function;
[0011] The final height distribution function is derived as a three-dimensional point cloud data, imported into COMSOL software for parametric modeling, generating roughness model;
[0012] In ABAQUS, finite element modeling and calculation, roughness model is imported into ABAQUS, assembly, set interaction, material properties and analysis step, mesh division;
[0013] The contact area of the contact surface to be predicted is calculated, and the contact resistance between the contact surfaces is calculated according to the contact area.
[0014] Optionally, the root mean square roughness RMS and the average roughness RA in the roughness model are represented by the following formulas, respectively:
[0015]
[0016] L represents the sampling length, x represents the sampling position, y (x) is the vertical distance of the profile from the centerline. Optionally, the general form of the autocovariance function in the roughness model is represented as:
[0017] C(h)=E[(z(x)-μ)(z(x+h)-μ)];
[0018] Where h is the displacement vector between the points in space; z(x) represents the height of the surface at position x; μ is the mean of the surface height; E[·] represents the expectation;
[0019] On a three-dimensional surface, h contains a vector with three components, and the calculation of the autocovariance function involves summing the displacements in three directions, that is:
[0020]
[0021] N is the number of sampling points on the surface, x i and x j are the positions of the two surface sampling points.
[0022] Given the autocovariance function of a three-dimensional surface, the calculation formula is as follows:
[0023]
[0024] Where L x and L y are the sizes of the surface in the x and y directions, respectively.
[0025] Optionally, the roughness model uses a power spectral density function to describe the frequency domain characteristics of the rough signal, and the calculation formula is as follows:
[0026]
[0027] S(ω x ,ω y ) is a two-dimensional power spectral density function, which describes the energy distribution of the rough signal at different frequency components. ω x and ω y represent the spatial frequency components in the x and y directions, respectively. ρ(σ x ,σ y ) is the autocorrelation function of the rough signal, which describes the correlation between different positions of the signal. cos(ω x σ x +ω y σ y ) is a cosine function, which represents the phase relationship of the signal at different frequency components.
[0028] Optionally, in ABAQUS, the current flowing through the two contact surfaces is determined by the following formula:
[0029]
[0030] In the formula, J is the current density from point A on one surface to point B on the other surface through the cross section, and are the potentials of the corresponding points on the contact surface, σ g is the gap conductivity.
[0031] Optionally, the gap conductivity is determined by the following formula:
[0032]
[0033] where, is the average temperature at points A and B, d is the gap between points A and B, p is the contact pressure transmitted through the interface between points A and B, is the average value of the predefined field between points A and B,
[0034] Optionally, the contact resistance between the contact surfaces is the inverse of the conductivity, where the conductivity is calculated using the following formula:
[0035]
[0036] where A c is the contact area, A is the total area; σ1 and σ2 represent the conductivities of the two different materials, respectively.
[0037] Compared with the prior art, the contact resistance prediction method between contact surfaces based on finite elements provided by the application can help engineers to select more suitable materials, optimize geometric shapes and sizes, and determine the optimal pressure range by using the finite element method to predict the contact resistance with the rich materials provided by the finite element software ABAQUS in the product design stage (such as switches, connectors, electrical contacts, etc.). Secondly, ABAQUS can realize the calculation of the contact area under single or complex working conditions, which helps to design products with stable, low and reliable contact resistance under expected working conditions, thereby reducing the failure rate caused by poor contact and improving the reliability of the entire system. In summary, the contact resistance prediction method between contact surfaces based on finite elements is a powerful engineering tool that combines theory, model and practical application, and can bring multiple benefits such as design optimization, cost savings, performance improvement and reliability enhancement. BRIEF DESCRIPTION OF DRAWINGS
[0038] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only embodiments of the application, and those skilled in the art can obtain other drawings according to the provided drawings without creative labor.
[0039] Figure 1 The experimental process and result graph for extracting the real spinning core die 42CrMo steel outer surface periodic amplitude of the application;
[0040] Figure 2 The experimental process and result graph for extracting the real spinning core die 42CrMo steel outer surface periodic amplitude of the application;
[0041] Figure 3 The comparison result graph of the experimental value and the simulated value of the roughness of the real spinning blank 2219 aluminum alloy inner surface of the application;
[0042] Figure 4 The comparison result graph of the experimental value and the simulated value of the roughness of the spinning core die 42CrMo steel outer surface of the application;
[0043] Figure 5 The simulated surface data distribution topographic map of the application;
[0044] Figure 6 The comparison graph of the Gaussian distribution and the amplitude frequency distribution of the simulated data conforming to the real result of the application;
[0045] Figure 7 The simulation spinning blank 2219 aluminum alloy and core die 42CrMo steel model preprocessing process graph of the application;
[0046] Figure 8 Equivalent plastic strain map under different loads and different temperatures for 0 degree loading of the present application;
[0047] Figure 9 Equivalent plastic strain map under different loads and different temperatures for 90 degree loading of the present application;
[0048] Figure 10 Calculation result map for 0 degree loading of the present application;
[0049] Figure 11 Calculation result map for 90 degree loading of the present application. DETAILED DESCRIPTION
[0050] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.
[0051] Embodiment 1
[0052] The embodiment of the present application discloses a contact resistance prediction method between contact surfaces based on finite elements, comprising the following steps:
[0053] Confirm the initial parameters for generating the three-dimensional model for predicting the contact resistance, and generate a Gaussian distribution white noise two-dimensional random sequence in MATLAB, calculate the Fourier transform, and obtain the first function of the surface height distribution through filtering processing;
[0054] A second function conforming to the real surface law is given, and the first function is fused to obtain the final height distribution function;
[0055] The final height distribution function is derived into three-dimensional point cloud data, imported into COMSOL software for parameterized modeling, and a roughness model is generated;
[0056] Finite element modeling and calculation are performed in ABAQUS, the roughness model is imported into ABAQUS, assembly, setting interaction, material properties and analysis steps, and mesh division are performed;
[0057] The contact area of the contact surface to be predicted is calculated, and the contact resistance between the contact surfaces is calculated according to the contact area.
[0058] GTMRS at contact interfaces algorithm:
[0059] Important parameters of surface roughness are root mean square roughness (RMS) and average roughness (RA), which are represented by the following equations, respectively:
[0060]
[0061] Both methods are used to represent surface roughness, but RMS is more sensitive to larger peaks and valleys.
[0062] The autocovariance function of a three-dimensional surface describes the relationship between the height variations of the surface at different locations. Specifically, the autocovariance function can be generally represented as:
[0063] C(h) = E[(z(x) - μ)(z(x + h) - μ)]
[0064] where h is the displacement vector between spatial points; z(x) represents the height of the surface at location x; μ is the mean of the surface height; E[·] represents the expectation.
[0065] On a three-dimensional surface, h is a vector containing three components, and the calculation of the autocovariance function involves summing the displacements in three directions, i.e.:
[0066]
[0067] N is the number of sampling points on the surface, x i and x j are the positions of two surface sampling points.
[0068] Given the autocovariance function of a three-dimensional surface:
[0069]
[0070] where L x and L y are the sizes of the surface in the x and y directions, respectively. The form of this function indicates that the autocovariance function is determined by the distance from the position to the boundary of the surface, i.e. the farther the boundary distance, the smaller the autocovariance, indicating that the surface has smaller variations at positions far away. It can be used to describe the correlation of the height variations of the surface, specifically the degree of association between the surface at different locations. In practical applications, such an autocovariance function can be used to model and analyze the characteristics of three-dimensional surfaces.
[0071] Given the autocovariance function of the surface height, it describes the correlation of the height variations of the surface at different locations.
[0072]
[0073] represents the double integral over the surface region, z(x, y) represents the height of the surface at position (x, y); z(x + σx y + σ y ) represents the height after displacement. This integral reflects the degree of surface height variation at a given displacement.
[0074] The autocorrelation function is defined as:
[0075]
[0076] Substituting the autocovariance function into the definition of the autocorrelation function, we get:
[0077]
[0078] where C(0,0) represents the autocovariance at zero displacement.
[0079] According to the above formula, we get:
[0080]
[0081] The power spectral density function represents the energy distribution of the signal in the frequency domain, while the autocorrelation function describes the correlation of the signal in the time domain. The power spectral density function is usually defined as the Fourier transform of the autocorrelation function, i.e.:
[0082]
[0083] R(τ) is the autocorrelation function; exp(-j2πfτ) is the complex exponential term, which represents the change of the sine wave of frequency f in the time domain.
[0084] The calculation is:
[0085]
[0086] This expression represents the energy distribution of the random process in the frequency domain, where the power spectral density function S(ω x ,ω y ) is the same at all frequencies, and is obtained by integrating the autocorrelation function multiplied by the cosine wave at each frequency.
[0087] By calculating the formula the value of the autocorrelation function is constrained in the interval [-1, 1], and ρ(0,0) = 1 represents complete correlation (complete match of itself at zero offset); ρ(x,y)→0 represents that the surface texture gradually becomes irrelevant as the offset increases. This standardization facilitates intuitive evaluation of the continuity of the surface texture, allowing direct comparison of the texture correlation characteristics of surfaces with different roughness levels.
[0088] In engineering practice, when the surface roughness is digitized sampling, the discrete autocorrelation function is usually used to describe the discrete random process. The autocorrelation function describes the correlation between the surface height at different positions, and under the digital sampling, the autocorrelation function will be expressed in discrete form.
[0089]
[0090] where N x and N y are the number of discrete sampling points of the surface height data; n x and n y are the discrete displacement of the discrete autocorrelation function.
[0091] The random process χ(x,y) is obtained by a two-dimensional filter to get the random process z(x,y):
[0092]
[0093] where σ x ,σ y ∈(1,2,…,N), n=N / 2, m=M / 2, h(σ x ,σ y ) is the response function of the filter at discrete coordinates.
[0094] The Fourier transform of the above formula is obtained:
[0095] Z(ω x ,ω y ) = H(ω x ,ω y ) A(ω x ,ω y )
[0096] where Z(ω x ,ω y ), H(ω x ,ω y ), A(ω x ,ω y ) are the Fourier transforms of z(x,y), h(σ x ,σ y ), χ(x,y), respectively.
[0097] According to the relevant literature, when the current passes through the interface between different materials, a certain resistance will be generated, which is the contact resistance, which can be expressed as:
[0098]
[0099] RC represents the contact resistance; n represents the total number of contact points; ρ1 and ρ2 represent the resistivity of two different materials, respectively; a iA represents the area of the ith contact point; a represents the radius between the contact points; and σ represents the electrical conductivity of the material.
[0100] When the two metal planes are not in contact: σ = 0 where σ is the electrical conductivity.
[0101] When the two metal planes are in complete contact, the electrical conductivity can be approximated as: σ = σ0 where σ0 is the intrinsic electrical conductivity of the material.
[0102] In the actual contact process, due to the large contact load, the film layer of the material contact surface is destroyed, so it can be considered that the contribution of the film layer resistance to the overall contact resistance is small and can be ignored. In this case, the contact resistance is mainly determined by the intrinsic electrical conductivity of the material, which can be approximated by the following formula:
[0103]
[0104] In the ABAQUS software, when simulating the contact resistance of the contact surface, the cross-sectional electrical conductivity is usually defined to achieve it. The cross-sectional electrical conductivity is the inverse of the contact resistance, which determines the difficulty of the current passing through the contact surface, and the current flowing through the two contact surfaces can be determined by the following formula:
[0105]
[0106] In the formula, J is the current density from point A on one surface to point B on the other surface through the cross section, and are the potentials of the corresponding points on the contact surface, σ g is the gap conductivity. The gap conductivity can be determined by the following formula:
[0107]
[0108] where, is the average temperature at points A and B, d is the gap between points A and B, and p is the contact pressure transmitted through the interface between points A and B, is the average value of the predefined field between points A and B,
[0109] From the above analysis, the contact surface conductivity is affected by temperature, pressure and area. Increasing the pressure can increase the actual contact area, reduce the gap and improve the electrical conductivity. The larger the contact area, the higher the electrical conductivity. Temperature changes affect the resistivity, and appropriate heating can reduce the contact resistance. The conductivity control formula is:
[0110]
[0111] where A cA is the total area; σ1 and σ2 represent the electrical conductivity of two different materials, respectively.
[0112] In summary, there are two parameters for simulating Gaussian rough random surface, which are root mean square roughness (RMS) of rough surface (σ) and autocorrelation length (β) in X and Y directions X ,β y . Because the autocorrelation function of most machined surfaces is exponential function, in the process of simulating rough surface, the expression of exponential autocorrelation function is used:
[0113] R = σ 2 exp(-2.3((τ x / β y ) 2 +(τ y / β y ) 2 )) (1)
[0114] In the calculation of the electrical conductivity of the contact surface, the following calculation formula is used:
[0115]
[0116] In the actual detection of workpiece roughness, the surface data is digitized and sampled to become a discrete random dot array, that is, z(x, y) is a discrete numerical value. If the random sequence obeys Gaussian distribution, a Gaussian random rough surface with a set autocorrelation function can be generated by computer simulation using two-dimensional digital low-pass filtering technology.
[0117] Confirm the initial parameters of the generated predicted three-dimensional model: the length in X L , Y L direction, the surface size X L × Y L of the generated three-dimensional model, the root mean square roughness, the number of grid points and the power spectral density;
[0118] Generate a two-dimensional coordinate grid using the meshgrid function in MATLAB, with a grid size of X L × Y L ;
[0119] Generate a Gaussian distribution white noise two-dimensional random sequence η(x, y) using the wgn function of MATLAB;
[0120] Calculate the Fourier transform A(ω x ,ω y ) of the random sequence η(x, y);
[0121] According to the power spectral density G z (ωx ,ω y ), while determining the power spectral density of the input sequence (C is constant because of the Gaussian distribution);
[0122] The transfer function H(ω x ,ω y ) of the filter is calculated by the formula H(ω z ,ω x ) = (G y (ω 1 / 2 ,ω x ) / C) y ;
[0123] The Fourier transform Z(ω x ,ω y ) of the output sequence of the input sequence after passing through the two-dimensional filter is obtained by the formula Z(ω x ,ω y ) = H(ω x ,ω y )A(ω x ,ω y );
[0124] The height distribution function Z1(x,y) of the surface is obtained by inverse Fourier transform of Z(ω x ,ω y );
[0125] The height distribution function Z2(x,y) of the surface is obtained by giving a regular function conforming to the real surface, such as a periodic function and an exponential function, etc.
[0126] The two height distribution functions Z1(x,y) and Z2(x,y) are fused to obtain the final height distribution function Z(x,y);
[0127] The final height distribution function Z(x,y) is exported in the format of three-dimensional point cloud data, that is, the X, Y, Z three-axis data are exported, and then the X, Y, Z three-axis data are imported into the COMSOL software for parametric modeling;
[0128] In the COMSOL software, the imported X, Y, Z three-axis data are selected to fit the scattered points into a surface, that is, a three-dimensional roughness plane is obtained. Then, a roughness model with one surface being the three-dimensional roughness plane and five surfaces being smooth planes is established according to the determined X L , Y L length and the self-determined Z L height;
[0129] The roughness model established in the COMSOL software is exported in the.dat format, and then the roughness model is imported into ABAQUS;
[0130] Assemble the imported roughness model in ABAQUS, set up interactions, define material properties and analysis steps, and divide the mesh;
[0131] Create a job to calculate the contact area A c , the total roughness surface area A is expressed as X L ×Y L Approximate calculations;
[0132] The contact resistance (the inverse of the electrical conductivity) between the contact surfaces is predicted using the calculation in equation (2).
[0133] Example 2
[0134] The 2219 aluminum alloy in this embodiment has a density of ρ = 2730 kg / m 3 , the yield strength σ0.2 at 25°C is 368MPa, the Young's modulus is 71.5GPa, the Poisson's ratio is 0.33, and the experimentally measured root mean square roughness RMS and average roughness RA of 2219 aluminum alloy are 1.480um and 1.184um respectively. Figure 1 As shown. 42CrMo steel, density ρ=8240kg / m 3 , the Young's modulus is 202GPa at 25℃, the Poisson's ratio is 0.288, the root mean square roughness RMS and average roughness RA of 42CrMo steel measured experimentally are 2.178um and 1.788um respectively. Figure 2 As shown. Since 42CrMo steel is used as a spinning core mold in the actual example, its deformation can be ignored, so it is equivalent to a rigid body. Since ABAQUS has no units, the models in this implementation example all use a rectangular roughness model with a length, width and height of 800um, 800um and 200um for resistance prediction. Therefore, the total contact area of the roughness model is calculated to be A = 800×800um2. Generate the initial parameters of the predicted three-dimensional model, X L , Y L The length in the direction is 800, the surface size of the generated three-dimensional model is 640000, the number of grid points N = 640000 and the power spectrum density C = 1;
[0135] Use the meshgrid function in MATLAB to generate a two-dimensional coordinate grid with a grid size of 800×800;
[0136] Use the wgn function to generate a Gaussian distributed white noise two-dimensional random sequence η(800,800);
[0137] Calculate the Fourier transform A(ω) of the random sequence η(800,800) 800 ,ω 800 );
[0138] According to R = σ 2 exp(-2.3((τ x / β y ) 2 +(τ y / β y ) 2 )) the Fourier transform of the filter output signal power spectrum density G z (ω 800 ,ω 800 ), while determining the input sequence power spectrum density C = 1;
[0139] The formula H(ω x ,ω 800 ) = (G z (ω 800 ,ω 800 ) / C) 1 / 2 Calculate the transfer function H(ω 800 ,ω 800 ) of the filter;
[0140] The Fourier transform Z(ω x ,ω y ) of the output sequence of the input sequence after two-dimensional filtering is obtained by the formula Z(ω x ,ω y ) = H(ω x ,ω y )A(ω 800 ,ω 800 );
[0141] Z(ω 800 ,ω 800 ) is inverse Fourier transformed to obtain the height distribution function Z1(x,y) of the surface;
[0142] Z2(x,y) is given to be a trigonometric function measured by experiment in accordance with the true law. The law function of aluminum alloy AL2219 is Z2(x,y) = 1.787*sin(32.708*x), and the law function of 42CrMo steel is Z2(x,y) = 3.119*sin(61.084*x). The measurement results are shown in Figure 3 、 Figure 4 The same size grid and grid point number, i.e. 640000 scattered points, are generated by using the previous steps;
[0143] The two height distribution functions Z1(x,y) and Z2(x,y) are fused to obtain the final height distribution function Z(x,y), and the fusion effect is shown in Figure 5 The roughness of the generated model is compared with that of the real model as shown in Figure 1 and Figure 2As shown, the comparison of the generated model and the real model obeys Gaussian distribution and amplitude distribution Figure 6 As shown (refer to Lu Hui's Microstructure Surface Topography Modeling and Optical Performance Analysis);
[0144] The final height distribution function Z(x, y) is exported in the form of three-dimensional point cloud data, that is, X, Y, and Z three-axis data are exported, and then the X, Y, and Z three-axis data are imported into the COMSOL software for parameterized modeling;
[0145] In the COMSOL software, the imported X, Y, and Z three-axis data are selected to parameterize the surface fitting of the scattered points into a surface, that is, a three-dimensional roughness plane is obtained. Then, according to the determined X L , Y L length and the self-determined ZL height and other system default values, a cuboid with a length of 800 um, a width of 800 um, and a height of 400 um is established. According to the parameterized surface fitting of the rough surface, the cuboid is divided into two parts by the division surface, and then one of the two parts after division is deleted, that is, the roughness model is obtained;
[0146] In the project tree of the COMSOL software, the established roughness model is exported in the.dat format, and then the roughness model is imported into ABAQUS;
[0147] The imported roughness model is assembled, the interaction is set, the material properties are defined, the analysis step is set, and the mesh is divided in ABAQUS (model diagram, assembly diagram, and mesh division diagram as shown in Figure 7 ) ;
[0148] The contact area A c of the two roughness planes (contact strain diagrams under 0-degree and 90-degree loading as shown in Figure 8 、 Figure 9 ) is obtained by creating a job and performing calculation, and the total area of the roughness planes is approximately calculated as A = XL × YL = 800 × 800 um 2 ;
[0149] The contact resistance (the inverse of conductivity) between the contact surfaces is calculated and predicted using (the calculation result diagrams under 0-degree and 90-degree loading as shown in Figure 10 、 Figure 11 ).
[0150] Each embodiment in the specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between each embodiment can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.
[0151] The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A finite element-based method of predicting contact resistance between contact surfaces, characterized by, The method comprises the following steps: Confirm the initial parameters of the contact resistance prediction three-dimensional model, generate a Gaussian distribution white noise two-dimensional random sequence in MATLAB, calculate the Fourier transform, and obtain the first function of the surface height distribution through filtering processing; A second function conforming to the real surface law is given, and the first function is fused to obtain a final height distribution function; The final height distribution function is exported as three-dimensional point cloud data, imported into COMSOL software for parameterized modeling, and a roughness model is generated; In ABAQUS, finite element modeling and calculation are performed, the roughness model is imported into ABAQUS, assembly, interaction setting, material property definition, analysis step, and mesh division are performed; The contact area of the contact surface to be predicted is calculated, and the contact resistance between the contact surfaces is calculated according to the contact area.
2. The method of claim 1, wherein, The root mean square roughness RMS and the average roughness RA in the roughness model are represented by the following formulas, respectively: L denotes the sampling length, x denotes the sampling position, y (x) is the perpendicular distance of the profile from the center line.
3. The method of claim 1, wherein, The general form of the autocovariance function in the roughness model is represented by the following formula: C(h) = E[(z(x)-μ)(z(x+h)-μ)]; Where h is the displacement vector between space points; z(x) represents the height of the surface at position x; μ is the mean of the surface height; E[·] represents the expectation; On a three-dimensional surface, h contains a vector with three components, and the calculation of the autocovariance function involves summing the displacements in three directions, that is: N is the number of surface sample points, x i and x j are two surface sample point locations; The calculation formula of the autocovariance function of the three-dimensional surface is given as follows: where L x and L y are the size of the surface in the x and y axis directions, respectively.
4. The method of claim 1, wherein, The power spectral density function in the roughness model is used to describe the frequency domain characteristics of the rough signal, and the calculation formula is as follows: S(ω x ,ω y ) is the two-dimensional power spectral density function, ω x and ω y represent the components of spatial frequency in x and y directions, respectively, ρ(σ x ,σ y ) is the autocorrelation function of the rough signal, and cos(ω x σ x +ω y σ y ) represents the phase relationship of the signal at different frequency components.
5. The method of claim 1, wherein, In ABAQUS, the current flowing through the two contact surfaces is determined by the following formula: where J is the current density from a point A on one face to a point B on the other face, and is the potential at the corresponding point on the contact surface, σ g is the gap conductivity.
6. The finite element based method of predicting contact resistance between contact surfaces according to claim 1, wherein, The gap conductivity is determined by the following formula: wherein, is the average temperature at point A and point B, d is the gap between point A and point B, p is the contact pressure transferred across the interface between point A and point B, is the average value of a predefined field between point A and point B, 7. The method of claim 1, wherein, The contact resistance between the contact surfaces is the inverse of the conductivity, where the conductivity is calculated using the following formula: where A c is the contact area, A is the total area; σ1 and σ2 represent the electrical conductivity of the two different materials, respectively.