Correlation modeling method for in-situ measurement of deformation of rivet under load and failure point location of rivet

Abstract

The invention discloses a rivet under-load deformation amount in-situ measurement and failure point location correlation modeling method, which comprises the following steps: after riveting deformation of a structural member, carrying out in-situ measurement on the deformation amount under the condition that a rivet section area is under load to obtain an axial deformation amount curve of a feature point set along an outer contour line of the rivet in the rivet section area; decomposing the axial deformation amount of the feature point into a transverse expansion amount and a longitudinal deformation amount, and constructing a correlation degree measurement model for quantifying the correlation between the deformation amount and the failure point location in combination with the coordinates of the failure point location; in the modeling process, based on Hooke's law and Poisson's ratio definition, establishing a relationship between the transverse expansion amount and the transverse stress of the feature points; based on the Hooke's law and the yield strength plastic deformation rule, the relation between the longitudinal deformation amount and the longitudinal force of the feature points is established; and combining the influence of the axial deformation of the feature point and the weighting and section shape of the distance measurement of the failure point location on the failure point location so as to obtain the correlation degree measurement model.

Description

Technical Field

[0001] This invention belongs to the field of health monitoring of aircraft assembly riveting structures, specifically involving an in-situ measurement method for the deformation of rivets under load and a correlation modeling method for the failure points. Background Technology

[0002] Riveting is a common joining process widely used in various engineering and manufacturing fields, including aerospace, automotive manufacturing, and construction engineering. During riveting, a rivet is inserted into a pre-drilled through-hole in the parts to be joined. Then, under the action of a punch, the rivet undergoes axial compression and lateral expansion. The rivet body fills the through-hole under lateral expansion, while the portion exceeding the through-hole is uptaken and shaped under the punching force, ultimately clamping the joined plates tightly. Riveting offers advantages such as high speed, high impact force, low cost, high strength, corrosion resistance, and good reliability, thus it is widely used in engineering.

[0003] However, the performance and reliability of riveted connections are directly affected by deformation and residual stress. Rivets typically bear multiaxial loads from different directions, such as axial tension, axial compression, and transverse shear forces. These multiaxial forces lead to residual stress within the riveted components. If the value exceeds the design allowable range, it can cause riveting failure, resulting in component separation, structural damage, or even accidents. Rivet failure usually occurs near its load-bearing point. Currently, it is still impossible to accurately predict or eliminate residual stress. Therefore, accurate measurement and monitoring of residual stress in riveted components is crucial. Early detection of residual stress in components can help prevent potential problems, reduce maintenance and repair costs, and improve the reliability and safety of the connection structure.

[0004] With advancements in science and technology, the emergence of micro-sensors and data acquisition systems has made in-situ measurement of riveting deformation possible. These micro-sensors can be mounted on riveted components to collect deformation data in real time to monitor the riveting status. For residual stress, speckle interferometry and Digital Image Correlation (DIC) methods are used to measure and evaluate the residual stress outside the rivet head and uphead regions. For measuring residual stress inside the rivet, X-ray diffraction at different intercrystalline regions is employed. Existing speckle interferometry and DIC methods are limited to measuring the external regions of the rivet head and uphead, and cannot assess the internal state of the rivet. While X-ray methods can measure the internal structure, they are only effective at depths of 10-30 μm, unable to detect deeper areas or assess the overall internal condition of the rivet under load. Therefore, they have limitations in revealing subsequent fatigue failure mechanisms. Summary of the Invention

[0005] The purpose of this invention is to provide a method for in-situ measurement of rivet deformation under load and correlation modeling of its failure point, so as to detect potential problems in advance, reduce maintenance costs, and thus improve the reliability of engineering structures.

[0006] To achieve the above objectives, the present invention employs the following technical solution: In-situ measurement of rivet deformation under load and correlation modeling method for failure point location, including: After deformation occurs during the riveting of the structural component, the deformation amount under load in the rivet section area is measured in situ to obtain the deformation curve of the characteristic point set along the outer contour line of the rivet in the rivet section area. The axial deformation of the feature point is decomposed into lateral expansion and longitudinal deformation. Combined with the coordinates of the failure point, a correlation measurement model is constructed to quantify the correlation between the deformation and the failure point. In the modeling process, the relationship between the lateral expansion and lateral stress of the feature point is established based on Hooke's law and the definition of Poisson's ratio; the relationship between the longitudinal deformation and longitudinal force of the feature point is established based on Hooke's law and the law of plastic deformation of yield strength; and the correlation measurement model is obtained by combining the weighted sum of the axial deformation of the feature point and the distance to the failure point, and the influence of the cross-sectional shape on the failure point.

[0007] Furthermore, in-situ measurements of deformation in the rivet cross-sectional area under load are performed, including: The rivet and its surrounding area are cut around the structural component to prepare a test block. The test block is then ground to create a cross-section parallel to the rivet axis. The clamping shank is then bonded with metal adhesive at the rivet head end and the uphead end along the rivet axis. Determine the maximum elastic load of the current rivet and set the experimental loading load accordingly; conduct a cyclic loading test on the specimen block on the fatigue testing machine using the clamping handle, and use the DIC method to measure the in-situ deformation of the rivet cross-section area under load during this process. Feature points are set along the outer contour of the rivet in the rivet profile area. The feature points at both ends of the contour are symmetrical along the axis of the rivet. The axial deformation of the feature points on the outer contour of the rivet is extracted. The feature point number is used as the abscissa and the deformation is used as the ordinate. The rivet axis is aligned with the abscissa to obtain the axial deformation curve of the feature points.

[0008] Furthermore, based on Hooke's law and the definition of Poisson's ratio, the relationship between the lateral expansion and lateral stress at the characteristic point is established, expressed as:

[0009] Among them, D xi Let σ be the lateral expansion amount of the i-th feature point. x Let E be the transverse stress of the rivet, E be the elastic modulus of the rivet material, and A be the elastic modulus.i Let A0 be the area of ​​the cross section containing the i-th feature point, and L be the total cross-sectional area of ​​the rivet. i The axial position of the i-th feature point after loading, from the origin, is obtained through the deformation curve.

[0010] Furthermore, based on Hooke's law and the law of plastic deformation of yield strength, the relationship between the longitudinal deformation and longitudinal force at the characteristic point is established, expressed as:

[0011] Among them, D yj Let F be the longitudinal deformation of the j-th feature point. y Let E be the longitudinal force acting on the rivet, and σ be the elastic modulus of the rivet material. s σ is the yield strength of the rivet material. y For longitudinal stress, L j The axial position of the j-th feature point after it is loaded, relative to the origin.

[0012] Furthermore, the failure point refers to the specific location where the rivet suffers irreversible damage or failure under stress; the failure point corresponds to the area where the rivet material undergoes plastic yielding or failure.

[0013] Furthermore, by combining the weighted average of the axial deformation of the feature point and the distance to the failure point, and the influence of the cross-sectional shape on the failure point location, the correlation measurement model is obtained, including:

[0014] in, For the correlation measurement model, P k These are the coordinates of the k-th failure point, where α1 and α2 are the weighting coefficients for the influence of lateral expansion and longitudinal deformation on the failure point, respectively. |D xi -P k | n and |D yj -P k | m The distance between the axial deformation of the feature point and the failure point is measured, n and m are the nonlinear relationship exponents between the deformation and the failure point, and β is the influence coefficient of the rivet material properties on the failure point. (A) i / A0) p The influence of cross-sectional shape on the failure point is introduced, where p is the sensitivity index of the area ratio.

[0015] Furthermore, an association matrix R is constructed based on the association measurement model, where each element R in the association matrix R is... ijkThe correlation between the lateral expansion of the k-th failure point and the longitudinal deformation of the i-th feature point is expressed as: .

[0016] Furthermore, the parameters of the correlation measurement model are optimized using an algorithm based on actual test data. The optimization objective is to minimize the error in the correlation matrix R, and the objective function is:

[0017] The algorithm includes regression analysis, least squares method, and neural network; N is the number of feature points, and K is the number of failure points.

[0018] A method for predicting rivet failure includes: First, axial cyclic loading or static load tests are conducted on the riveted structure under controlled loading conditions. The deformation curves of each characteristic point on the outer contour of the rivet profile are obtained by using the DIC method or strain measurement technology. At the same time, local geometric parameters and material performance indicators are collected. Subsequently, based on the geometric parameters, material performance indicators and deformation curves, the lateral expansion and longitudinal deformation of each feature point are calculated, and the corresponding correlation matrix is ​​constructed using the correlation degree measurement model to obtain the correlation degree between the feature point and the failure point. Establish correlation degree-load or time curves and dynamically update the correlation degree in the correlation degree matrix under different working conditions; use extreme value criteria or threshold method to determine that the area where the failure point is located has entered a potential failure state when the correlation degree of a certain failure point exceeds the set threshold or shows a sudden increase trend.

[0019] Compared with the prior art, the present invention has the following technical features: 1. This invention proposes an in-situ measurement scheme for riveting deformation, which measures the axial deformation of the rivet under load based on the DIC method, and uses the deformation at a specific point as an observation to characterize the magnitude of the internal residual stress, thus solving the difficult problem of measuring the internal residual stress after riveting.

[0020] 2. This invention proposes a modeling method that correlates deformation with failure location. By analyzing the correlation between deformation and failure location, the failure mechanism of riveting is revealed. Furthermore, model correlation analysis and model optimization are performed to make the connection between rivet deformation and axial failure more intuitive, thereby optimizing the method for analyzing the influence of deformation on axial failure. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the specimen block preparation and clamping process; Figure 2 This is a flowchart illustrating the method of the present invention. Detailed Implementation

[0022] See Figure 1 and Figure 2 This invention discloses an in-situ measurement method for the deformation of a rivet under load and a method for associating the failure point with it, comprising the following steps: Step 1: After deformation occurs at the riveting joint of the structural component, in-situ measurements of the deformation in the rivet cross-section area under load are performed. The magnitude of the residual stress formed is characterized by the deformation as an observable. The specific details are as follows: First, the rivet and its surrounding area are cut around the structural component to prepare a test piece. The test piece is then ground to create a cross-section parallel to the rivet axis. The clamping shank is then bonded to the rivet head and the uphead end with metal adhesive along the rivet axis.

[0023] Secondly, the maximum elastic load of the rivet is calculated based on its mechanical properties. 1 / 10 to 1 / 2 of the maximum elastic load is selected as the final experimental load. Cyclic loading experiments are performed on the specimen block using the clamping handle on a fatigue testing machine. During this process, the in-situ deformation of the rivet profile area under load is measured using the DIC method. Feature points are set along the outer contour of the rivet in the rivet profile area. Starting from one end of the rivet head, the feature points are marked as #1, #2, #3, etc. The feature points at both ends of the rivet are symmetrical along the axis, so the feature point numbers at the other end are #1', #2', #3', etc.

[0024] Finally, based on the DIC measurement results, the axial deformation of the feature points on the outer contour line of the rivet is extracted. A coordinate graph is plotted with the feature point number as the abscissa and the deformation as the ordinate, thus obtaining the deformation information of the overall outer contour of the rivet cross-section area.

[0025] Step 2 involves performing correlation modeling between deformation and rivet failure points; specifically, correlation analysis between deformation and failure points reveals the riveting failure mechanism, thereby optimizing the riveting structure.

[0026] By aligning the rivet axis with the horizontal axis of the coordinate graph, and considering that the specimen block is subjected to varying tensile loads after riveting, each feature point on the outer contour will produce a certain amount of deformation, thus obtaining the axial deformation curve of the feature point.

[0027] Then, for the rivets that have been identified as having axial failure, the failure points on the rivets are matched with the feature points with large deformations. The analysis shows that there is a certain correlation between the deformation of the rivets under axial load and the failure points. When the rivets are compressed, they expand in the transverse direction, and when they are stretched, they deform in the longitudinal direction. Under each load condition, the feature points on the outer contour of the rivets have corresponding deformations. The deformation of the feature points along the axial direction is decomposed into transverse expansion and longitudinal deformation. Furthermore, the residual stress of the rivets is large at the feature points with large deformations, making them prone to failure.

[0028] Finally, by analyzing the correlation between the deformation of the rivet under load and the failure location, a multi-dimensional mathematical model was established, which can more accurately predict the failure location of the riveted structure. This model not only considers the lateral and longitudinal deformation of the rivet, but also introduces multiple variables that affect rivet failure, such as stress and material plasticity, and models the failure through the coupling relationship between multiple parameters.

[0029] (1) Lateral expansion amount D x Definition.

[0030] Lateral expansion refers to the deformation of the rivet body in the lateral (perpendicular to the axial) direction after it is subjected to force. Assuming the rivet is subjected to axial load, the rivet expands in the lateral direction, and the lateral expansion at n feature points can be expressed by the following formula: D x ={D x1 D x2 ,...,D N} Based on the material properties, cross-sectional shape, and force applied to the rivet, the lateral expansion D at the i-th (i=1,2,…,N) characteristic point can be established using Hooke's Law and the definition of Poisson's ratio. xi Relationship with transverse stress; transverse expansion D xi This can be expressed by the following formula:

[0031] Where, σ x Let E be the transverse stress of the rivet, E be the elastic modulus of the rivet material, and A be the elastic modulus. i Let A0 be the area of ​​the cross section containing the i-th feature point, and L be the total cross-sectional area of ​​the rivet. i The axial position of the i-th feature point after loading, from the origin, is obtained through the deformation curve. The concept of area ratio is introduced here to consider the response of lateral deformation to different cross-sectional areas. The change in lateral expansion is highly related to the force distribution, shape, and stress state of the rivet.

[0032] (2) Longitudinal deformation D y Definition.

[0033] Longitudinal deformation describes the deformation of a rivet under axial load; longitudinal deformation mainly refers to the change in length of the rivet under axial tension or compression, and is usually more significant under axial load. Longitudinal deformation can be defined by the following formula: D y ={D y1 D y2 ,...,D yN} Longitudinal deformation is typically influenced by factors such as axial force on the rivet, material plasticity, and stress concentration. When considering deformation caused by axial force, the yield strength of the rivet and the strain hardening behavior of the material also need to be taken into account. The longitudinal deformation D at the j-th (j=1,2…,N) characteristic point is established using Hooke's law and the yield strength-plastic deformation law. yj The expression is:

[0034] Among them, F y σ is the longitudinal force acting on the rivet. s σ is the yield strength of the rivet material. y For longitudinal stress, L j The axial position of the j-th feature point after loading, from the origin, is obtained through the deformation curve. This formula incorporates a yield strength term, reflecting the yield and strain hardening characteristics of the rivet material, ensuring accurate calculation of longitudinal deformation under different stress states.

[0035] (3) Definition of the coordinates of the failure point.

[0036] The failure point refers to the specific location where a rivet suffers irreversible damage or failure under stress. Rivet failure is usually caused by excessive stress concentration or plastic deformation. Determining the failure point involves a combination of stress analysis and deformation analysis.

[0037] The coordinates P of the failure point are defined as follows: P={P1,P2,...,P K} Among them, P k These are the coordinates of the kth (k=1,2,…,K) failure point, where K is the number of failure points. These failure points typically correspond to the areas where the rivet material undergoes plastic yielding or failure. The occurrence of failure points is closely related to the amount of deformation, stress distribution, and the geometry of the rivet.

[0038] (4) Association degree measurement model f(D) xi D yj ,P k ).

[0039] To quantify the correlation between deformation and failure location, this scheme introduces a correlation degree measurement model f(D) xi D yj ,P k This function reflects the relationship between deformation and failure location; considering the influence between deformation and failure location, it can be expressed by the following weighted model:

[0040] Where α1 and α2 are the weighting coefficients for the influence of lateral expansion and longitudinal deformation on the failure location, respectively, |D xi -P k | n and |D yj -P k | m The distance between the deformation and the failure point is measured, n and m are the nonlinear relationship exponents between the deformation and the failure point, and β is the influence coefficient of the rivet material properties on the failure point. (A) i / A0) p The influence of cross-sectional shape on the failure location is introduced. p is the sensitivity index of area ratio, which is used to adjust the weight sensitivity of cross-sectional shape or local geometric features on the failure correlation degree. When p=1, it means that the influence of area ratio on correlation degree is linear. When p>1, it means that the influence of area difference is amplified (the influence of cross-sectional contraction zone or stress concentration zone is more prominent). When p<1, it means that the influence of area ratio is weakened (the influence of shape change is small, which is suitable for approximately uniform rods).

[0041] This model introduces a more complex nonlinear relationship between lateral and longitudinal deformation, which can more accurately reflect the distribution of failure points.

[0042] (5) Correlation matrix R.

[0043] The correlation matrix R is used to describe the correlation between the lateral expansion and longitudinal deformation of all feature points and all failure points, and has a multi-dimensional structure.

[0044] Each element R in the correlation matrix R ijk The correlation between the lateral expansion of the k-th failure point and the longitudinal deformation of the i-th feature point can be obtained using the following formula:

[0045] Each element in the correlation matrix R expresses the relationship between the deformation and the failure point through a weighting function, further revealing the coupling relationship between the deformation area and the failure point.

[0046] (6) Association analysis and model optimization.

[0047] To obtain a more accurate prediction model, various algorithms, including regression analysis, least squares method, and neural networks, were optimized using actual test data to optimize the parameters (α1 and α2, n and m, p) of the correlation degree measurement model. The optimization objective is to minimize the error in the correlation matrix R, and the objective function is:

[0048] This formula describes minimizing the objective function. By training and optimizing the parameters of the correlation measurement model, a more accurate correlation matrix R can be obtained, thereby accurately predicting the failure point of the rivet.

[0049] Using this method, the proposed solution can accurately predict and optimize the failure points of riveted structures. The formula not only introduces more physical meaning and complexity, but also improves the accuracy and practical application value of the model through nonlinear correlation functions and optimization algorithms.

[0050] Step 3: After obtaining the optimized correlation matrix R, the prediction of rivet failure can be achieved by calculating and comparing the correlation degree between the actual measurement data and the correlation matrix.

[0051] Specifically, firstly, under controlled loading conditions, axial cyclic loading or static load tests are conducted on the riveted structure. The deformation curves of characteristic points on the outer contour of the rivet profile are obtained using the DIC method or strain measurement technology. Simultaneously, local geometric parameters (such as cross-sectional area) are acquired. Compared with the benchmark area The input dataset consists of material properties (yield strength, elastic modulus, Poisson's ratio) and other parameters.

[0052] Subsequently, the lateral expansion D at each feature point was calculated. x and longitudinal deformation D y， Using the parameter-optimized correlation measurement model, construct the corresponding correlation matrix R, and calculate the correlation between each feature point and the failure point. .

[0053] By normalizing and statistically analyzing the correlation of each failure point, the failure points that show a significant increase in the rate of change of correlation during loading and that stably exhibit extreme values ​​in multiple loading cycles are identified. These failure points represent the locations where the local strain energy accumulation and geometric deformation are most synergistic, and represent the areas where the structure is most susceptible to damage or yielding. Specifically, in engineering applications, correlation degree-load (or time) curves can be established to dynamically update the correlation degree matrix R under different working conditions, and extreme value criteria or threshold methods can be used to determine the failure risk area. When the correlation degree of a certain failure point exceeds the set threshold or shows a sudden increase trend (the correlation degree change rate is greater than the change rate threshold), it can be determined that the area where the failure point is located has entered a potential failure state. Through this method, potential problems can be detected in advance, thereby improving the reliability of engineering structures.

[0054] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for in-situ measurement of rivet deformation under load and correlation modeling of its failure point, characterized in that, include: After deformation occurs during the riveting of the structural component, the deformation amount under load in the rivet section area is measured in situ to obtain the deformation curve of the characteristic point set along the outer contour line of the rivet in the rivet section area. The axial deformation of the feature point is decomposed into lateral expansion and longitudinal deformation. Combined with the coordinates of the failure point, a correlation measurement model is constructed to quantify the correlation between the deformation and the failure point. In the modeling process, the relationship between the lateral expansion and lateral stress of the feature point is established based on Hooke's law and the definition of Poisson's ratio; the relationship between the longitudinal deformation and longitudinal force of the feature point is established based on Hooke's law and the law of plastic deformation of yield strength; and the correlation measurement model is obtained by combining the weighted sum of the axial deformation of the feature point and the distance to the failure point, and the influence of the cross-sectional shape on the failure point.

2. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point as described in claim 1, is characterized in that, In-situ measurements of deformation in the rivet cross-sectional area under load were performed, including: The rivet and its surrounding area are cut around the structural component to prepare a test block. The test block is then ground to create a cross-section parallel to the rivet axis. The clamping shank is then bonded with metal adhesive at the rivet head end and the uphead end along the rivet axis. Determine the maximum elastic load of the current rivet and set the experimental loading load accordingly; conduct a cyclic loading test on the specimen block on the fatigue testing machine using the clamping handle, and use the DIC method to measure the in-situ deformation of the rivet cross-section area under load during this process. Feature points are set along the outer contour of the rivet in the rivet profile area. The feature points at both ends of the contour are symmetrical along the axis of the rivet. The axial deformation of the feature points on the outer contour of the rivet is extracted. The feature point number is used as the abscissa and the deformation is used as the ordinate. The rivet axis is aligned with the abscissa to obtain the axial deformation curve of the feature points.

3. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point as described in claim 1, is characterized in that... Based on Hooke's law and the definition of Poisson's ratio, the relationship between the lateral expansion and lateral stress at a characteristic point is established and expressed as follows: Among them, D xi Let σ be the lateral expansion amount of the i-th feature point. x Let E be the transverse stress of the rivet, E be the elastic modulus of the rivet material, and A be the elastic modulus. i Let A0 be the area of ​​the cross section containing the i-th feature point, and L be the total cross-sectional area of ​​the rivet. i The axial position of the i-th feature point after loading, from the origin, is obtained through the deformation curve.

4. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point as described in claim 1, characterized in that, The relationship between the longitudinal deformation and longitudinal force at a characteristic point, based on Hooke's law and the law of plastic deformation due to yield strength, is established as follows: Among them, D yj Let F be the longitudinal deformation of the j-th feature point. y Let E be the longitudinal force acting on the rivet, and σ be the elastic modulus of the rivet material. s σ is the yield strength of the rivet material. y For longitudinal stress, L j The axial position of the j-th feature point after it is loaded, relative to the origin.

5. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point as described in claim 1, characterized in that, The failure point refers to the specific location where the rivet suffers irreversible damage or failure under stress; the failure point corresponds to the area where the rivet material undergoes plastic yielding or failure.

6. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point according to claim 1, characterized in that, By combining the weighted average of the axial deformation of the feature point and the distance to the failure point, and the influence of the cross-sectional shape on the failure point location, the correlation measurement model is obtained, including: in, For the correlation measurement model, P k These are the coordinates of the k-th failure point, where α1 and α2 are the weighting coefficients for the influence of lateral expansion and longitudinal deformation on the failure point, respectively. |D xi -P k | n and |D yj -P k | m The distance between the axial deformation of the feature point and the failure point is measured, n and m are the nonlinear relationship exponents between the deformation and the failure point, and β is the influence coefficient of the rivet material properties on the failure point. (A) i / A0) p The influence of cross-sectional shape on the failure point is introduced, where p is the sensitivity index of the area ratio.

7. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point as described in claim 1, characterized in that, Construct an association matrix R based on the association measurement model. Each element R in the association matrix R is R_i. ijk The correlation between the lateral expansion of the k-th failure point and the longitudinal deformation of the i-th feature point is expressed as: .

8. The method for in-situ measurement of rivet deformation under load and the associated modeling of its failure point according to claim 1, characterized in that, The parameters of the correlation measurement model are optimized using an algorithm based on actual test data. The optimization objective is to minimize the error in the correlation matrix R. The objective function for optimization is: The algorithm includes regression analysis, least squares method, and neural network; N is the number of feature points, and K is the number of failure points.

9. A method for predicting rivet failure, characterized in that, include: First, axial cyclic loading or static load tests are conducted on the riveted structure under controlled loading conditions. The deformation curves of each characteristic point on the outer contour of the rivet profile are obtained by using the DIC method or strain measurement technology. At the same time, local geometric parameters and material performance indicators are collected. Subsequently, based on the geometric parameters, material performance indicators and deformation curves, the lateral expansion and longitudinal deformation of each feature point are calculated, and the corresponding correlation matrix is ​​constructed using the correlation degree measurement model to obtain the correlation degree between the feature point and the failure point. Establish correlation degree-load or time curves and dynamically update the correlation degree in the correlation degree matrix under different working conditions; use extreme value criteria or threshold method to determine that the area where the failure point is located has entered a potential failure state when the correlation degree of a certain failure point exceeds the set threshold or shows a sudden increase trend.

Images