Intelligent prediction method and system for fatigue life of rusted steel plate

Abstract

The present application relates to the technical field of fatigue life prediction, in particular to a kind of rusted steel plate fatigue life intelligent prediction method and system, method includes: the three-dimensional topography of the rusted steel plate to be measured is measured to obtain its surface data, then data is input into the crack fatigue initiation life prediction model trained, the model contains crack initiation life prediction and stress concentration coefficient prediction module, whereby crack fatigue initiation life is obtained.Subsequently, the crack initiation life prediction model is analyzed to obtain the crack initiation life prediction model.Obtain the corresponding life by inputting the defect into the fatigue crack propagation life prediction model.Finally, according to the crack fatigue initiation life and the fatigue crack propagation life, the total life of the rusted steel plate is calculated, and the crack prediction accuracy is further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fatigue life prediction, and particularly relates to a fatigue life intelligent prediction method and system for a rusted steel plate. BACKGROUND

[0002] Steel structures are widely used in large-span structures, light steel structures, high-rise buildings, and bridges due to their high strength and excellent seismic performance. However, the components in engineering structures inevitably face fatigue problems during long-term use, that is, the material gradually damages under repeated or periodic loads, eventually leading to failure. Although the strength of single load is lower than the yield strength of the material, the material gradually initiates cracks under repeated loads, and these cracks continuously expand, eventually leading to material fracture. When there is water and air in the environment, steel will have a serious corrosion phenomenon. Rust has a very significant impact on fatigue and is the main driving force to accelerate the fatigue process. Rust forms pits and rough areas of varying depths on the surface of steel structures, and these irregular surface features greatly increase stress concentration, making cracks more likely to initiate and accelerate in the rusted area. Rust can significantly shorten the fatigue life of components, causing the structure to fail in a relatively short period of time.

[0003] Therefore, how to accurately predict the fatigue life of rusted components and take targeted preventive and reinforcement measures has become an important part of the service performance evaluation and improvement of existing structures in corrosive environments.

[0004] In fatigue life research, fatigue life is usually divided into crack initiation life and crack propagation life. By studying these two stages respectively, different analysis methods and measures are taken for different mechanical behaviors, which can more accurately predict the overall fatigue life of materials or structures.

[0005] Crack initiation refers to the time period from the beginning to the appearance of initial micro-cracks on the surface or inside of the material, and this stage is mainly affected by the surface state of the material. The rusted surface usually has a complex and irregular morphology, and the surface morphology characteristics of the rusted steel plate can be accurately obtained through three-dimensional morphology scanning. Fatigue cracks often originate from defects or the maximum stress of the sample. Stress concentration leads to an increase in local stress, making the local area more likely to exceed the fatigue limit of the material and enter the fatigue crack initiation stage in advance. The higher the stress concentration coefficient, the shorter the crack initiation life. The complexity of the rusted surface and stress concentration are important reasons for the variability of the fatigue crack initiation life of the rusted steel plate.

[0006] Crack propagation life is the time period from the initial crack to the crack propagation leading to the final failure, which is mainly affected by crack tip stress intensity factor and other factors. Through observation of the rusted surface, it is found that there are different local defects at the bottom of the rust pit, which will lead to the change of stress intensity factor at the bottom of the rust pit. Therefore, it can be considered that the local defects at the bottom of the rust pit are the main reason for the great variability of the crack propagation life of the rusted steel plate.

[0007] The crack initiation and propagation of the rusted component are affected by various factors, resulting in a complex relationship that is difficult to systematically analyze. Traditional fatigue life models mainly rely on material properties and test conditions, have low universality, and have numerous parameters and large calculation amount, which limits their engineering application value. SUMMARY

[0008] The technical problem to be solved by the present application is to provide a fatigue life intelligent prediction method and system for a rusted steel plate to solve the technical problem of low accuracy of existing rusted steel plate fatigue life prediction.

[0009] The purpose of the present application is achieved by the following technical solutions:

[0010] In a first aspect, the present application provides a fatigue life intelligent prediction method for a rusted steel plate, comprising:

[0011] S1, measuring the three-dimensional morphology of the rusted steel plate to be tested to obtain the surface data of the rusted steel plate to be tested;

[0012] S2, inputting the surface data of the rusted steel plate to be tested into a trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the rusted steel plate to be tested; the crack fatigue initiation life prediction model comprises a crack initiation life prediction module and a stress concentration coefficient prediction module; the stress concentration coefficient prediction module is used to obtain the stress concentration coefficient according to the input surface data of the rusted steel plate to be tested, and the crack initiation life prediction module is used to obtain the crack initiation life according to the stress concentration coefficient;

[0013] S3, analyzing the rusted detail information of the crack initiation to obtain the rusted defect characteristics;

[0014] S4, inputting the rusted defect of the rusted steel plate to be tested into a trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the rusted steel plate to be tested;

[0015] S5, obtaining the total life of the rusted steel plate according to the crack fatigue initiation life and the fatigue crack propagation life of the rusted steel plate to be tested.

[0016] As a further improvement of the present application, the measurement and analysis of the three-dimensional morphology of the rusted steel plate to be tested to obtain the surface data of the rusted steel plate to be tested specifically comprises:

[0017] The surface of the rusted steel plate to be measured is measured in three dimensions to obtain surface data of the rusted steel plate to be measured, the surface data of the rusted steel plate to be measured including distribution, depth and size information of surface rust pits.

[0018] As a further improvement of the present application, the surface data of the rusted steel plate to be measured is input into the trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the rusted steel plate to be measured, specifically including:

[0019] The surface data of the rusted steel plate to be measured is input into the trained stress concentration coefficient prediction module to obtain the predicted maximum surface stress concentration coefficient;

[0020] The predicted maximum surface stress concentration coefficient is multiplied by the applied stress and then input into the trained crack initiation life prediction module to obtain the crack fatigue initiation life of the rusted steel plate to be measured;

[0021] The training process of the stress concentration coefficient prediction module and the crack initiation life prediction module respectively includes:

[0022] The collected surface data of the rusted steel plate to be measured is imported into a finite element software, a solid model of the rusted steel plate is constructed according to the finite element software, finite element analysis is performed on the rusted surfaces of different topographies according to the solid model of the rusted steel plate, and the stress concentration coefficient of the steel plate surface is calculated based on the solid model of the rusted steel plate; a stress concentration coefficient database of the rusted steel plate surface is constructed according to the geometric features of the rusted surface and the corresponding stress concentration coefficient;

[0023] The stress concentration coefficient prediction module is trained according to the stress concentration coefficient database of the rusted steel plate surface;

[0024] The fatigue life and the proportion of crack initiation life in total fatigue life of the non-rusted steel plate under different stress amplitudes are determined to construct a fatigue crack initiation life database of the non-rusted steel plate;

[0025] The crack initiation life prediction module is trained according to the fatigue crack initiation life database of the non-rusted steel plate, and in the training process, the crack initiation life prediction module uses the loss function corresponding to the neural network in the crack initiation life prediction module of the non-rusted steel plate to optimize the crack initiation life prediction module.

[0026] As a further improvement of the present application, the crack initiation life prediction module adopts a physical information neural network model; the corresponding physical information of the crack initiation life prediction module is obtained according to an S-N curve, and the physical information includes a monotonically decreasing trend between the crack initiation life and the corresponding stress amplitude applied to the rusted steel plate, and the slope of the S-N curve gradually increases.

[0027] As a further improvement of the present application, the loss function corresponding to the neural network in the crack initiation life prediction module of the non-rusting steel plate specifically comprises:

[0028]

[0029]

[0030]

[0031]

[0032] In the formula, y true is the actual value of the crack initiation life, y pred is the predicted value of the crack initiation life, Penaltys s (1) for punishing the case that the predicted function and the first-order derivative of stress are positive, Penaltys s (2) for punishing the case that the predicted function and the second-order derivative of stress are negative.

[0033] As a further improvement of the present application, the fatigue crack characteristics of the rusted steel plate are analyzed in detail, specifically comprising:

[0034] According to the maximum stress concentration coefficient, the crack initiation corresponding to the rust pit of the to-be-tested rusted steel plate is determined, a two-dimensional section view of the rust pit is obtained, and the macroscopic characteristics of the rust pit on the surface of the rusted steel plate and the microscopic defects at the bottom of the rust pit are characterized;

[0035] An association model of the macroscopic geometric characteristics of the rust pit and the microscopic defect characteristics at the bottom of the rust pit is established through regression analysis.

[0036] As a further improvement of the present application, the fatigue crack propagation life prediction model is obtained by training a fatigue crack propagation life database, and the training process of the fatigue crack propagation life prediction model comprises:

[0037] According to the association model, the microscopic defects at the bottom of the rust pit on the surface of the rusted steel plate are simplified, the irregular rust pit is simplified into an elliptical regular pit, and the microscopic defects at the bottom of the rust pit are simplified into a plurality of initial cracks in the crack propagation stage, and a simplified model of the fatigue crack propagation of the rusted steel plate is obtained according to the defect characteristics;

[0038] The stress state of the simplified model of the rusted steel plate is analyzed by using a finite element model, and the crack propagation behavior under cyclic loading in different conditions is simulated by using FRANC3D;

[0039] The crack propagation continues until the stress intensity factor of the rusted steel plate exceeds the fracture toughness and fracture occurs, and the number of cycles of the cyclic load at this time is recorded as the crack propagation life;

[0040] The crack propagation simulation results of different rusted steel plates are collected, the corresponding rust pit geometric characteristics, crack characteristics, cyclic load characteristics and crack propagation life of each model are recorded, a fatigue crack propagation life database is constructed, and a fatigue crack propagation life prediction model is trained based on the fatigue crack propagation life database.

[0041] As a further improvement of the application, the fatigue crack propagation life prediction model adopts a physical information neural network model optimized based on the physical information in the Paris formula, and the physical information includes the influence of the changes of the maximum stress and the initial crack length on the crack propagation life.

[0042] As a further improvement of the application, the fatigue crack propagation life prediction model loss function is:

[0043]

[0044] Wherein,

[0045]

[0046] In the formula, P1 and P2 are used to punish the first derivative of the prediction function on the maximum stress and the initial crack length.

[0047] In a second aspect, the application provides a rusted steel plate fatigue life intelligent prediction system for realizing the rusted steel plate fatigue life intelligent prediction method described above, comprising:

[0048] A data acquisition unit measures the three-dimensional morphology of the to-be-tested rusted steel plate and acquires the surface data of the to-be-tested rusted steel plate.

[0049] A crack fatigue initiation life prediction unit inputs the surface data of the to-be-tested rusted steel plate into the trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the to-be-tested rusted steel plate; the crack fatigue initiation life prediction model comprises a crack initiation life prediction module and a stress concentration coefficient prediction module; the stress concentration coefficient prediction module obtains a stress concentration coefficient, and the crack initiation life prediction module is used to obtain a crack initiation life according to the stress concentration coefficient;

[0050] A data analysis unit analyzes the rusting detail information of crack initiation to obtain rust defect characteristics.

[0051] A fatigue crack propagation life prediction unit inputs the rust defect of the to-be-tested rusted steel plate into the trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the to-be-tested rusted steel plate.

[0052] A rusted steel plate total life prediction unit obtains the total life of the rusted steel plate according to the crack fatigue initiation life and the fatigue crack propagation life of the to-be-tested rusted steel plate.

[0053] The fatigue life intelligent prediction method of the rusted steel plate has the advantages that the rusted steel plate to be measured is measured in three dimensions to obtain surface data, which can comprehensively and accurately reflect the actual rusting condition of the steel plate. Compared with the traditional single-dimensional measurement method, this three-dimensional measurement can capture more detailed information, such as the depth, shape and distribution of rust pits, to provide more reliable basic data for subsequent life prediction. The use of the crack fatigue initiation life prediction model and the fatigue crack propagation life prediction model makes the evaluation of the life of the rusted steel plate more precise. The crack initiation life prediction module combined with the stress concentration coefficient prediction module considers the key influence of stress concentration on crack initiation, and can accurately predict when and where the crack starts to appear. The fatigue crack propagation life prediction model focuses on the propagation stage after the crack appears, and the separate prediction of the two stages can more accurately grasp the entire process from the beginning of the problem to the final failure of the rusted steel plate. The analysis of the rusting detail information of the crack initiation area obtains the rusting defect characteristics, which helps to understand the specific influence of rusting on the steel plate structure. The present application can avoid unnecessary early replacement or excessive maintenance, thereby saving costs. By accurately understanding the remaining life of the steel plate, maintenance or replacement can be performed at the most appropriate time, avoiding waste of resources.

[0054] Further, compared with the traditional contact measurement method, the non-contact measurement does not damage the surface of the rusted steel plate, ensuring the authenticity and reliability of the measurement results. At the same time, it can avoid the deformation of the rust pits or measurement errors caused by contact pressure, further improving the measurement accuracy.

[0055] Further, by inputting the surface data of the rusted steel plate to be measured into the specially trained stress concentration coefficient prediction module and crack initiation life prediction module, the fatigue crack initiation life can be accurately predicted. The stress concentration coefficient is an important parameter for evaluating the possibility of crack initiation in a structure, and accurate prediction of the stress concentration coefficient provides a reliable basis for subsequent calculation of the fatigue crack initiation life. At the same time, determining the crack initiation helps to take targeted maintenance measures in advance to prevent further damage to the structure. Combining the prediction results of the two modules, the surface characteristics and mechanical properties of the rusted steel plate are comprehensively considered to improve the accuracy of the prediction. Compared with a single prediction method, this step-by-step prediction method can more carefully analyze the influence of rusting on the fatigue performance of the steel plate. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to make the objects, technical scheme and advantages of the embodiments of the present application or the prior art clearer, the accompanying drawings that need to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those of ordinary skill in the art without any creative work.

[0057] Figure 1 It is an intelligent prediction flowchart of fatigue life of a rusted steel plate.

[0058] Figure 2 It is a structure diagram of a stress concentration coefficient artificial neural network.

[0059] Figure 3 It is a scanning result diagram of a three-dimensional surface morphology of a rusted steel plate.

[0060] Figure 4 It is a surface stress nephogram of a solid model of a rusted steel plate.

[0061] Figure 5 It is a physical information neural network diagram for predicting crack initiation life of an un-rusted steel plate.

[0062] Figure 6 It is a simplified model diagram of irregular rust pits and micro-defects at the bottom of the rust pits of a rusted steel plate.

[0063] Figure 7 It is a simplified result diagram obtained by the simplified model of irregular rust pits and micro-defects at the bottom of the rust pits of a rusted steel plate.

[0064] Figure 8 It shows the expansion and fusion of multiple cracks at the bottom of the rust pits during the crack expansion process. DETAILED DESCRIPTION

[0065] In order to make the objects, technical scheme and advantages of the embodiments of the present application or the prior art clearer, the accompanying drawings that need to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those of ordinary skill in the art without any creative work.

[0066] The technical scheme of the present application will be described clearly and completely in combination with the accompanying drawings and specific embodiments. The described embodiments are only some embodiments of the present application, not all the embodiments.

[0067] Embodiment 1

[0068] As Figures 1-8As shown, the present embodiment provides an intelligent fatigue life prediction method for rusted steel plates. The fatigue life of a rusted steel plate is divided into crack initiation life and crack propagation life. The life prediction of the crack initiation stage of the rusted steel plate includes stress concentration analysis based on the three-dimensional topographic features of the rusted surface, rusted surface stress concentration coefficient prediction based on a deep artificial neural network, and crack initiation life prediction based on a physical information neural network. The life prediction of the crack propagation stage includes crack propagation life simulation based on FRANC3D, and crack propagation life prediction based on a physical information neural network. Based on the prediction of the two stages, the accuracy of the fatigue life prediction of the rusted steel plate and the interpretability of the artificial neural network are greatly improved. The following is a specific implementation.

[0069] S1, measure the three-dimensional topography of the rusted steel plate to be measured, and obtain the surface data of the rusted steel plate to be measured; in the present embodiment, a three-dimensional non-contact optical profiler is used to record the topography of the corrosion surface of the rusted steel plate to be measured, the three-dimensional topography of the surface of the rusted steel plate to be measured is measured, and the surface data of the rusted steel plate to be measured is obtained, the surface data of the rusted steel plate to be measured includes rust pit details, and the rust pit details include the distribution, depth, size, etc. of the surface rust pits.

[0070] Specifically, a three-dimensional non-contact optical profiler (3D Profiling) produced by the American NANOVER company is used to perform high-precision three-dimensional scanning on a large number of actually rusted steel plates, capture the geometric features of the surface rust, including the depth, width, distribution shape, etc. of the rust pits, and the results are as shown in Figure 3 The results of the three-dimensional scanning are imported into Abaqus to establish a solid model of the rusted steel plate, which accurately reflects the rusting features of the surface. Through numerical simulation calculation, the stress concentration effect of different rusted surfaces is studied, and the results are as shown in Figure 4

[0071] S2, input the surface data of the rusted steel plate to be measured into the trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the rusted steel plate to be measured. The crack fatigue initiation life prediction model includes a crack initiation life prediction module and a stress concentration coefficient prediction module; the stress concentration coefficient prediction module obtains the stress concentration coefficient, and the crack initiation life prediction module is used to obtain the crack initiation life according to the stress concentration coefficient. Specifically, it includes:

[0072] The surface data of the rusted steel plate to be measured is input into the trained stress concentration coefficient prediction module to obtain the predicted stress concentration coefficient; after the predicted stress concentration coefficient is multiplied by the applied stress, it is input into the trained crack initiation life prediction module to obtain the crack fatigue initiation life of the rusted steel plate to be measured.

[0073] The training process of the stress concentration coefficient prediction module and the crack initiation life prediction module includes: ​

[0074] The collected surface data of the rusted steel plate to be tested is imported into a finite element software, a rusted steel plate entity model is constructed according to the finite element software, finite element analysis is performed on rusted surfaces with different appearances according to the rusted steel plate entity model, and the stress concentration coefficient of the surface of the steel plate is calculated based on the rusted steel plate entity model; specifically, a plurality of finite element analyses are performed on a large number of rusted surface models with different appearances, and these finite element models cover various parameters such as rust depth, pit shape, distribution, and density.

[0075] A rusted steel plate surface stress concentration coefficient database is obtained according to the geometric characteristics of the rusted surface and the corresponding stress concentration coefficient; the database contains the geometric characteristics of the rusted surface and the corresponding stress concentration coefficient, and provides an adequate data basis for the subsequent training of the data-driven rusted surface stress concentration coefficient artificial neural network.

[0076] A stress concentration coefficient prediction module based on a deep neural network framework is trained according to the rusted steel plate surface stress concentration coefficient database, and a trained stress concentration coefficient prediction module is obtained; in this embodiment, the stress concentration coefficient prediction module adopts a physical information neural network model. The trained physical information neural network model quickly obtains the corresponding stress concentration coefficient by inputting any rusted surface characteristics, without the need to perform finite element simulation every time, greatly improving the calculation efficiency. A database of “rusted surface geometric characteristics and stress concentration coefficient” is generated, recording the corresponding relationship between various rust characteristics (such as pit shape and size) and the corresponding stress concentration coefficient; the database is trained by an artificial neural network model as shown in FIG. 8, and a module for quickly predicting the stress concentration coefficient according to the scanned rust characteristics is established, which is used for subsequent rusted steel plate crack initiation life prediction. Figure 2

[0077] According to the loss function corresponding to the neural network in the crack initiation life prediction module of the un-rusted steel plate, the fatigue life and the proportion of crack initiation life in the total fatigue life of the un-rusted steel plate under different stress amplitudes are determined, and an un-rusted steel plate fatigue crack initiation life database is established;

[0078] A crack initiation life prediction module is trained according to the un-rusted steel plate fatigue crack initiation life database, and a trained crack initiation life prediction module is obtained.

[0079] In the un-rusted steel plate fatigue crack initiation life neural network based on physical information, the physical quantity introduced in the loss function is the stress amplitude applied to the component. According to the physical information of the S-N curve, the relationship between the crack initiation life and the corresponding stress amplitude applied to the rusted steel plate shows a monotonically decreasing trend, and the slope of the S-N curve gradually increases, so the formula of the physical information loss function used in this network is as follows:

[0080] ​

[0081]

[0082]

[0083]

[0084] where y true is the actual value of crack initiation life, y pred is the predicted value of crack initiation life, Penaltys s (1) for penalizing the case that the predicted function is positive with the first derivative of stress. s (2) for penalizing the case that the predicted function is negative with the second derivative of stress.

[0085] Through the fatigue crack initiation life database of uncorroded steel plates, a physical information neural network model is trained to predict the crack initiation life of uncorroded steel plates according to the input material attribute characteristics and load characteristics.

[0086] Specifically, according to the existing experimental data, the crack initiation life of uncorroded steel plates under different stress amplitudes is determined, and then a fatigue crack initiation life database of uncorroded steel plates is established, which includes loading system (stress amplitude, stress ratio, etc.) and crack initiation life; a physical information neural network model (PINN) is constructed, as shown in Figure 5 , which combines physical laws to obtain the relationship between stress amplitude and crack initiation life from the S-N curve. Based on the crack initiation life database, the neural network model is trained to obtain a tool that can quickly predict the crack initiation life under different load conditions. The process of applying the above module to predict the crack initiation life of a specific corroded steel plate is shown in Figure 6 .

[0087] S3, analyze the corrosion defect characteristics of the region where the crack initiates.

[0088] According to the maximum stress concentration coefficient, the crack initiation corresponding to the rust pit of the to-be-tested rusted steel plate is determined, the three-dimensional scanning technology is used to finely analyze the bottom of the rust pit corresponding to the crack initiation of the to-be-tested rusted steel plate, and the two-dimensional section diagram of the rust pit is extracted. And all the defects at the bottom of the rust pit on the surface of the rusted steel plate are characterized, the number, size and distribution of the local defects at the bottom of the rust pit are analyzed. According to different corrosion characteristics, the number, size and other characteristics of the local defects at the bottom of the rust pit are regressed with the macro-features of the rust pit (such as the depth and width of the rust pit). The shape, size and distribution of the micro-defects at the bottom of the rust pit and other factors will make the fatigue crack propagation life show great uncertainty. Among them, the micro-defects refer to the discontinuous or abnormal structures that cannot be directly distinguished by the naked eye at the bottom of the rust pit, but can be seen at the micro-scale. These defects will affect the mechanical properties of the steel plate, such as reducing the strength and toughness, increasing the stress concentration degree, and thus becoming the potential starting point of crack initiation and propagation under the load of the steel plate, which has an important influence on the fatigue life of the steel plate.

[0089] Through regression analysis, a correlation model of the macro-geometric features of the rust pit and the micro-defect features (such as number, size, etc.) at the bottom of the rust pit is established.

[0090] According to the correlation model, the micro-defects at the bottom of the rust pit on the surface of the rusted steel plate are simplified, the irregular rust pit is simplified into an elliptical regular pit, and the micro-defects at the bottom of the rust pit are simplified into multiple initial cracks in the crack propagation stage. According to the defect characteristics, a simplified model of the fatigue crack propagation of the rusted steel plate is obtained, and the defect characteristics are obtained according to the simplified model of the rusted steel plate. The simplified defect characteristics are shown in FIG. 8. Figure 7 The stress state of the simplified model is analyzed by using the finite element model, and FRANC3D is used to simulate the crack propagation behavior under cyclic load in different conditions; the crack propagates until the stress intensity factor of the rusted steel plate exceeds the fracture toughness and breaks, and the number of cycles at this time is recorded as the crack propagation life; the crack propagation simulation results of different rusted steel plates are collected, and the rust pit geometric characteristics, crack characteristics, cyclic load characteristics and crack propagation life corresponding to each model are recorded, a fatigue crack propagation life database is constructed, and a fatigue crack propagation life prediction model is trained based on the fatigue crack propagation life database.

[0091] S4, input the rusted steel plate corrosion defect to be tested into the trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the to-be-tested rusted steel plate.

[0092] The fatigue crack propagation life prediction model adopts a physical information neural network model constructed based on the Paris formula, and the physical information includes the influence of the changes of the maximum stress and the initial crack length on the crack propagation life. The fatigue crack propagation life prediction model training process includes:

[0093] In the area of the corrosion steel plate corresponding to the maximum stress concentration factor, a two-dimensional cross-sectional view of the rust pit is obtained, and the corrosion characteristics are determined according to the two-dimensional cross-sectional view;

[0094] The corrosion characteristics are subjected to regression analysis, the micro-defects at the bottom of the rust pit on the surface of the corrosion steel plate are subjected to simplified processing, the irregular rust pit is simplified into an elliptical regular pit, and the micro-defects at the bottom of the rust pit are simplified into multiple initial cracks in the crack propagation stage, to obtain a simplified model of the corrosion steel plate and different defect characteristics; wherein, the simplified model is imported into Franc3D software, and necessary material parameters including elastic modulus, Poisson's ratio, fracture toughness, etc. are set according to the actual mechanical properties of the material. In addition, the loading conditions required for fatigue crack propagation analysis are set according to the working conditions of the fatigue load.

[0095] The finite element model is used to perform parameterized analysis on the crack propagation characteristics of the corrosion steel plate, and the Paris formula is used to simulate the crack propagation behavior under cyclic load in different conditions; in addition, the embodiment also focuses on the change of the stress intensity factor (SIF) and the propagation path of the crack at the bottom of the rust pit, and the propagation and fusion of the multiple cracks at the bottom of the rust pit are as shown in Figure 8 .

[0096] The crack propagation continues until the stress intensity factor of the corrosion steel plate exceeds the fracture toughness and fracture occurs, and the number of cycles at this time is recorded as the crack propagation life; the above simulation process is repeated to establish a database containing “loading system, initial crack length, rust pit macroscopic characteristics, etc. and crack propagation life”.

[0097] The simulation results of the crack propagation of different corrosion steel plates are collected, and the rust pit geometric characteristics, crack characteristics, cyclic load characteristics and crack propagation life corresponding to each model are recorded to construct a fatigue crack propagation life database, and a fatigue crack propagation life prediction model is trained based on the fatigue crack propagation life database.

[0098] According to the Paris formula, the crack propagation rate is related to the stress intensity factor (ΔK), and the stress intensity factor depends on the maximum stress and the crack length. Therefore, the model should learn that: for a given initial crack length, the increase of the maximum stress will lead to the decrease of the crack propagation life; for a given maximum stress, the increase of the initial crack length will also lead to the decrease of the crack propagation life; in the fatigue crack propagation life prediction model for predicting the crack propagation life, the physical quantities introduced in the loss function include the cyclic load stress applied to the component and the length of the initial crack at the bottom of the rust pit; the formula of the physical information loss function used in the fatigue crack propagation life prediction model is as follows:

[0099]

[0100] wherein,

[0101]

[0102] where P1 and P2 are used to penalize the positive first derivative of the prediction function at the maximum stress and initial crack length, respectively.

[0103] S5, obtaining the total life of the rusted steel plate according to the crack fatigue initiation life and the fatigue crack propagation life of the rusted steel plate to be measured. Specifically, the fatigue crack initiation life and the propagation life of the rusted steel plate predicted based on the neural network are added to obtain the total life of the rusted steel plate.

[0104] For a certain existing rusted steel plate, after obtaining its rusting characteristics through three-dimensional scanning, the crack initiation life, the crack propagation life of the steel plate under different loading conditions are predicted respectively, and the total fatigue life of the rusted steel plate is given.

[0105] In summary, by measuring the three-dimensional morphology of the rusted steel plate to be measured by the three-dimensional non-contact optical profiler, detailed information including the distribution, depth and size of the rust pits can be obtained, providing an accurate data basis for subsequent analysis. Combined with the finite element software to construct the entity model and analyze, the understanding of the surface characteristics and stress concentration effect of the rusted steel plate is further improved. By measuring the three-dimensional morphology of the rusted steel plate to be measured by the three-dimensional non-contact optical profiler, detailed information including the distribution, depth and size of the rust pits can be obtained, providing an accurate data basis for subsequent analysis. Combined with the finite element software to construct the entity model and analyze, the understanding of the surface characteristics and stress concentration effect of the rusted steel plate is further improved. The rusted steel plate is analyzed in detail in the crack initiation area, the two-dimensional cross-sectional view of the rust pit is extracted by three-dimensional scanning technology, the bottom defect of the rust pit is characterized and regression analyzed, the correlation model is established and the defect characteristics are simplified. This detailed analysis helps to more accurately understand the influence of the rust defect on the fatigue crack propagation life, and provides a reliable basis for the subsequent prediction of the fatigue crack propagation life.

[0106] The present embodiment adopts advanced three-dimensional scanning technology and physical information neural network model, greatly improving the prediction efficiency. Without the need for a large number of traditional experiments and complex calculations, only by obtaining the surface data of the rusted steel plate through three-dimensional scanning and inputting into the trained model, the prediction results of the crack fatigue initiation life, the crack propagation life and the total life can be quickly obtained.

[0107] Embodiment 2

[0108] The present embodiment provides a fatigue life intelligent prediction system of a rusted steel plate, which is used to realize the fatigue life intelligent prediction method of the rusted steel plate in the above-mentioned embodiment 1. The system comprises a data acquisition unit, a crack fatigue initiation life prediction unit, a data analysis unit, a fatigue crack propagation life prediction unit and a rusted steel plate total life prediction unit. Each unit specifically comprises:

[0109] a data acquisition unit, which measures the three-dimensional morphology of the rusted steel plate to be measured and acquires surface data of the rusted steel plate to be measured;

[0110] a crack fatigue initiation life prediction unit, which inputs the surface data of the rusted steel plate to be measured into a trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the rusted steel plate to be measured; the crack fatigue initiation life prediction model comprises a crack initiation life prediction module and a stress concentration coefficient prediction module; the stress concentration coefficient prediction module obtains a stress concentration coefficient, and the crack initiation life prediction module is used to obtain the crack initiation life according to the stress concentration coefficient;

[0111] a data analysis unit, which analyzes the rusting detail information of the crack initiation to obtain rust defect features;

[0112] a fatigue crack propagation life prediction unit, which inputs the rust defect of the rusted steel plate to be measured into a trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the rusted steel plate to be measured;

[0113] a rusted steel plate total life prediction unit, which obtains the total life of the rusted steel plate according to the crack fatigue initiation life and the fatigue crack propagation life of the rusted steel plate to be measured.

Claims

1. A method for intelligently predicting the fatigue life of corroded steel plates, characterized in that, include: S1. Measure the three-dimensional morphology of the rusted steel plate to be tested and obtain the surface data of the rusted steel plate to be tested; S2. Input the surface data of the corroded steel plate to be tested into the trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the corroded steel plate to be tested; the crack fatigue initiation life prediction model includes a crack initiation life prediction module and a stress concentration factor prediction module; the stress concentration factor prediction module is used to obtain the stress concentration factor based on the input surface data of the corroded steel plate to be tested, and the crack initiation life prediction module is used to obtain the crack initiation life based on the stress concentration factor. S3. Analyze the detailed information of corrosion on crack initiation to obtain the characteristics of corrosion defects; S4. Input the corrosion defects of the steel plate to be tested into the trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the steel plate to be tested. S5. Based on the crack fatigue initiation life and fatigue crack propagation life of the rusted steel plate to be tested, the total life of the rusted steel plate is obtained. The step of inputting the surface data of the corroded steel plate to be tested into the trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the corroded steel plate to be tested specifically includes: The surface data of the rusted steel plate to be tested is input into the trained stress concentration factor prediction module to obtain the predicted maximum surface stress concentration factor. The predicted maximum surface stress concentration factor is multiplied by the applied stress and then input into the trained crack initiation life prediction module to obtain the crack fatigue initiation life of the corroded steel plate under test. The training processes for the stress concentration factor prediction module and the crack initiation life prediction module include: The collected surface data of the rusted steel plate to be tested is imported into the finite element software. A solid model of the rusted steel plate is constructed based on the finite element software. Finite element analysis is performed on rusted surfaces with different morphologies based on the solid model of the rusted steel plate. The maximum stress concentration factor on the surface of the steel plate is calculated based on the solid model of the rusted steel plate. A database of stress concentration factors on the surface of the rusted steel plate is constructed based on the geometric characteristics of the rusted surface and the corresponding maximum stress concentration factor. A stress concentration factor prediction module was trained based on a database of stress concentration factors on the surface of corroded steel plates. Determine the fatigue life and crack initiation life of non-rusted steel plates under different stress amplitudes, and construct a fatigue crack initiation life database for non-rusted steel plates. A crack initiation life prediction module is trained based on a fatigue crack initiation life database of non-rusted steel plates. During the training process, the crack initiation life prediction module is optimized by utilizing the loss function corresponding to the neural network in the crack initiation life prediction module.

2. The intelligent fatigue life prediction method for corroded steel plates according to claim 1, characterized in that, The measurement and analysis of the three-dimensional morphology of the rusted steel plate to be tested, and the acquisition of surface data of the rusted steel plate to be tested, specifically include: A three-dimensional non-contact optical profilometer is used to record the morphology of the corroded surface of the steel plate to be tested. The three-dimensional morphology of the surface of the steel plate to be tested is measured to obtain the surface data of the steel plate to be tested. The surface data of the steel plate to be tested includes the distribution, depth and size information of the surface rust pits.

3. The intelligent fatigue life prediction method for corroded steel plates according to claim 1, characterized in that, The crack initiation life prediction module adopts a physical information neural network model; the physical information corresponding to the crack initiation life prediction module is obtained from the SN curve. The physical information includes the relationship between crack initiation life and the stress amplitude applied to the corresponding rusted steel plate, which shows a monotonically decreasing trend, and the slope of the SN curve gradually increases.

4. The intelligent fatigue life prediction method for corroded steel plates according to claim 1, characterized in that, The loss function corresponding to the neural network in the crack initiation life prediction module for non-rusted steel plates specifically includes: In the formula, y true y represents the actual value of crack initiation lifetime. pred Penaltys is a predicted value for crack initiation lifetime. s (1) Penaltys is used to penalize the case where the first derivative of the prediction function and the stress are positive. s (2) This is used to penalize cases where the prediction function and the second derivative of the stress are negative.

5. The intelligent fatigue life prediction method for corroded steel plates according to claim 1, characterized in that, A detailed analysis of the fatigue crack characteristics of corroded steel plates is conducted, including: The rust pits corresponding to the crack initiation of the steel plate under test are determined based on the maximum stress concentration factor. Two-dimensional cross-sectional views of the rust pits are obtained, and the macroscopic features of the rust pits on the surface of the steel plate and the microscopic defects at the bottom of the rust pits are characterized. A correlation model between the macroscopic geometric features of rust pits and the microscopic defect features at the bottom of rust pits was established using regression analysis.

6. The intelligent fatigue life prediction method for corroded steel plates according to claim 5, characterized in that, The fatigue crack propagation life prediction model is trained using a fatigue crack propagation life database. The training process of the fatigue crack propagation life prediction model includes: Based on the aforementioned correlation model, the microscopic defects at the bottom of the rust pits on the surface of the corroded steel plate are simplified, the irregular rust pits are simplified into elliptical regular pits, and the microscopic defects at the bottom of the rust pits are simplified into multiple initial cracks in the crack propagation stage. Based on the defect characteristics, a simplified model of fatigue crack propagation of the corroded steel plate is obtained. The simplified model was subjected to parametric analysis of stress state using the finite element model, and the crack propagation behavior under cyclic loading was simulated using FRANC3D under different conditions. The crack propagates until the stress intensity factor of the corroded steel plate exceeds the fracture toughness and fracture occurs. The number of cyclic loading at this point is recorded as the crack propagation life. The simulation results of crack propagation for different rusted steel plates were collected, and the geometric features of rust pits, crack features, cyclic loading features and crack propagation life corresponding to each model were recorded. A fatigue crack propagation life database was constructed, and a fatigue crack propagation life prediction model was trained based on the fatigue crack propagation life database.

7. The intelligent fatigue life prediction method for corroded steel plates according to claim 1, characterized in that, The fatigue crack propagation life prediction model adopts a physical information neural network model based on the physical information in the Paris formula. The physical information includes the influence of the changes in maximum stress and initial crack length on crack propagation life.

8. The intelligent fatigue life prediction method for corroded steel plates according to claim 7, characterized in that, The loss function of the fatigue crack propagation life prediction model is: in, In the formula, P1 and P2 are used to penalize the case where the first derivative of the prediction function is positive at the maximum stress and the initial crack length, respectively.

9. A fatigue life intelligent prediction system for corroded steel plates, characterized in that, The method for intelligently predicting the fatigue life of corroded steel plates according to any one of claims 1-8 includes: The data acquisition unit measures the three-dimensional morphology of the corroded steel plate to be tested and acquires the surface data of the corroded steel plate to be tested. The crack fatigue initiation life prediction unit inputs surface data of the corroded steel plate to be tested into a trained crack fatigue initiation life prediction model to obtain the crack fatigue initiation life of the corroded steel plate to be tested. The crack fatigue initiation life prediction model includes a crack initiation life prediction module and a stress concentration factor prediction module. The stress concentration factor prediction module obtains the stress concentration factor, and the crack initiation life prediction module is used to obtain the crack initiation life based on the stress concentration factor. The data analysis unit analyzes the detailed information of rust initiation in cracks to obtain the characteristics of rust defects; The fatigue crack propagation life prediction unit inputs the corrosion defects of the steel plate to be tested into the trained fatigue crack propagation life prediction model to obtain the fatigue crack propagation life of the steel plate to be tested. The total life prediction unit for rusted steel plates obtains the total life of the rusted steel plate based on the crack fatigue initiation life and fatigue crack propagation life of the rusted steel plate under test.

Images