Optimization Method for Steel Structure Connection Welding Process Based on Microstructure Regulation

Abstract

This invention discloses a method for optimizing the welding process of steel structure connections based on microstructure control, belonging to the field of welding technology. The method first obtains the material parameters of the metal to be welded, the welding materials, and the welding process parameters; then, it combines experiments and numerical simulations to obtain the welding temperature time history distribution; next, it establishes a continuous cooling transformation curve, calibrates the phase transformation kinetic model and the grain size evolution model, and predicts the phase volume fraction and grain size in each region of the welded connection; it constructs a fatigue life prediction model considering the influence of phase volume fraction and grain size; finally, using fatigue life as the optimization objective, it performs reverse design of the welding process parameters to obtain a combination of process parameters that meets the target fatigue performance requirements. This invention establishes a quantitative correlation between phase volume fraction, grain size, and fatigue performance, enabling fatigue life-oriented welding process optimization for steel structure welded connections, significantly reducing experimental costs and improving optimization efficiency.

Description

Technical Field

[0001] This invention relates to the field of welding technology, and in particular to a method for optimizing the welding process of steel structure connections based on microstructure control. Background Technology

[0002] Welded connections in steel structures are widely used in bridges, marine engineering, building structures, pressure vessels, and engineering equipment. During actual service, these connections are frequently subjected to cyclic loads such as vehicle loads, wind loads, wave loads, and mechanical vibration loads, making them prone to fatigue damage and affecting the structural safety. Therefore, predicting the fatigue life of welded connections in steel structures and optimizing the welding process based on this prediction has significant engineering application value.

[0003] The welding process is essentially a localized, unsteady-state thermal cycle. Influenced by the welding heat input and cooling conditions, the weld metal and welding materials undergo complex microstructural evolution during welding, with significant differences in phase volume fraction, grain size, and mechanical properties in different regions. Especially in the weld seam and heat-affected zone, the material undergoes heating and cooling processes, easily forming a microstructure containing multiple metallographic types, which significantly impacts the fatigue performance of the welded joint.

[0004] Current research and engineering evaluation of the fatigue performance of welded joints typically employ methods such as the nominal stress method and the hot spot stress method. While these methods are relatively mature in engineering applications and can effectively assess the fatigue performance of welded joints, they primarily rely on the empirical relationship between macroscopic stress parameters and fatigue life, failing to adequately reflect the intrinsic connections between welding thermal cycles, microstructure evolution, and fatigue life. For welded joints, fatigue performance is not only related to external loads but also closely linked to the local microstructure and its spatial distribution formed during the welding process. Therefore, establishing a quantitative relationship between welding process parameters, thermal cycling, microstructure evolution characteristics, and fatigue life remains one of the pressing technical problems to be solved in this field.

[0005] On the other hand, the determination of existing welding process parameters largely relies on empirical selection or repeated experiments, often requiring numerous welding and fatigue tests. This results in drawbacks such as long development cycles, high testing costs, and low optimization efficiency. Although some numerical simulation methods can be used for welding temperature field analysis or microstructure evolution prediction, existing research mainly focuses on the temperature field, phase transformation process, or microstructure characteristics themselves, lacking a systematic method that further couples microstructure characteristics with fatigue life prediction and directly serves welding process optimization design. Furthermore, the temperature histories experienced by different regions of the welded joint are not the same, and the weld metal and welding material may also differ in composition, thermophysical parameters, and microstructure transformation behavior. Analysis based solely on a unified empirical model or single-region performance indicators often fails to comprehensively characterize the fatigue performance of welded butt joints. Currently, there is still a lack of a steel structure connection welding process optimization method that can comprehensively consider the differences in characteristics between the weld metal and welding material, welding temperature history, microstructure transformation process, grain size evolution, and fatigue performance. Summary of the Invention

[0006] To address the aforementioned issues, this invention aims to propose a method for optimizing the welding process of steel structure connections based on microstructure control. By establishing a quantitative correlation between microstructure characteristics and fatigue life, and optimizing welding process parameters accordingly, this method achieves synergistic analysis of the welding temperature process, microstructure evolution process, and fatigue life. This reduces experimental workload, lowers R&D costs, and improves welding process design efficiency and welded connection fatigue performance.

[0007] To achieve the above objectives, the technical solution of the present invention is implemented as follows: A method for optimizing steel structure connection welding process based on microstructure control includes the following steps: S1. Obtain the material parameters of the metal to be welded and the welding materials, as well as the welding process parameters; S2. Based on the material parameters and welding process parameters, a numerical model of the welding temperature field is established by combining experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. S3. Establish continuous cooling transformation curves for the welded metal and welding materials, and calibrate the phase transformation kinetics model and grain size evolution model for the welded metal and welding materials. S4. Combining the welding temperature time history distribution, phase transformation kinetics model, and grain size evolution model of the welded joint, obtain the phase volume fraction and grain size of each region of the welded joint; S5. Conduct thermal simulation tests and fatigue tests on the welded metal and welding materials in sequence, and construct a fatigue life prediction model that considers the effects of phase volume fraction and grain size. S6. Taking fatigue life as the optimization target, and combining the prediction results of phase volume fraction and grain size, the welding process parameters are reverse-engineered through the fatigue life prediction model to obtain a combination of process parameters that meet the target fatigue performance requirements.

[0008] Furthermore, in step S1, the material parameters of the metal to be welded and the welding material include the thermophysical parameters, chemical element content, volume fraction of each phase and grain size of each phase of the base metal, as well as the thermophysical parameters and chemical element content of the welding material; the welding material is welding wire or welding rod; the welding process parameters are arc welding process parameters, including welding voltage, welding current and welding speed.

[0009] Furthermore, in step S2, a numerical model of the welding temperature field is established using a combination of experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. Specifically, thermocouples are arranged in the heat-affected zone and its adjacent areas to measure the temperature time history during the actual welding process. Based on the measured geometric dimensions, constraints, heat dissipation conditions, thermophysical parameters of the welded metal and welding materials, and welding process parameters of the welded joint, a finite element model of the welding temperature field is established in conjunction with a moving heat source model. The finite element model of the welding temperature field is then verified using the temperature time history data obtained through thermocouple measurements to obtain the welding temperature time history distribution of the welded joint. The moving heat source model includes one or more of the following: a Gaussian surface heat source model, an ellipsoidal heat source model, and a double ellipsoidal heat source model.

[0010] Furthermore, in step S3, establishing the continuous cooling transformation curves of the welded metal and the welding material, and calibrating the phase transformation kinetic model and grain size evolution model of the welded metal and the welding material specifically refers to: conducting thermal expansion tests on the welded metal and the welding material, applying thermal cycles with different cooling rates; acquiring thermal expansion curves under each thermal cycle condition using a dilatometer, and determining the phase transformation initiation temperature and phase transformation termination temperature using the tangent method; obtaining the volume fraction of each phase under each thermal cycle condition using electron backscattering diffraction experiments; and establishing the correspondence between cooling conditions, phase transformation temperature, and phase composition by combining the phase transformation temperature range determined by the thermal expansion curves and the electron backscattering diffraction experiment results. The continuous cooling transformation curves of the weld metal and the welding material were obtained respectively. Based on the continuous cooling transformation curves of the weld metal and the welding material, the linear interpolation model used to predict the austenite volume fraction in the heating stage was calibrated, the modified Johnson–Mehl–Avrami–Kolmogorov model used to predict the ferrite, pearlite and bainite volume fraction in the cooling stage was calibrated, the Koistinen–Marburger model used to predict the martensite volume fraction in the cooling stage was calibrated, the Sellars model used to predict austenite grain growth in the heating stage was calibrated, and the Suehiro model used to predict the ferrite, pearlite, bainite and martensite grain size in the cooling stage was calibrated.

[0011] Furthermore, the linear interpolation model is calculated according to the following formula:

[0012] Among them, f A A represents the volume fraction of austenite; T represents the current temperature; A represents the volume fraction of austenite. c1 and A c3 These are the start and end temperatures of the austenite transformation, respectively. The modified Johnson–Mehl–Avrami–Kolmogorov model is calculated according to the following formula:

[0013]

[0014] Among them, f i Let f be the phase volume fraction of the metallographic type calculated at time i, where f is the phase volume fraction for ferrite, pearlite, and bainite respectively. F f P f B ;k i and n i The kinetic parameter is related to temperature, alloy element content, and metallographic type; Δt i The time increment from time i-1 to time i; The Koistinen–Marburger model is calculated according to the following formula:

[0015] Among them, f M b is the martensite volume fraction; b is the martensite transformation rate constant; M s This is the temperature at which the martensitic transformation begins.

[0016] The Sellars model is calculated using the following formula:

[0017] Where, d A d0 is the average austenite grain size during the heating stage; t is the initial parent material grain size; Q is the equivalent holding time; R is the activation energy for grain growth; A is the gas constant, taken as 8.314 J / (mol·K); and A and n are material constants. The Suehiro model is calculated using the following formula:

[0018] Where d is the average grain diameter of each metallographic structure when cooled to room temperature, and d is the average grain diameter of ferrite, pearlite, bainite, and martensite, respectively. F d P d B d M ;d max The average grain diameter at the end of the heating phase; T 5% The temperature at which the volume fraction of each metallographic phase reaches 5% during the cooling process; C1 and K are material constants; f is the volume fraction of the metallographic structure calculated when cooled to room temperature.

[0019] Furthermore, the parameter A in the linear interpolation model c1 and A c3 The parameter k in the modified Johnson–Mehl–Avrami–Kolmogorov model i and n i The parameter M in the Koistinen–Marburger model s b, obtained through continuous cooling transformation curves; the parameters Q, A, and n in the Sellars model, and the parameters C1 and K in the Suehiro model, obtained through electron backscattering diffraction experiments and phase transition kinetic models.

[0020] Furthermore, in step S5, the specific procedure for the thermal simulation test is as follows: referring to the actual welding thermal cycle process and the established continuous cooling transformation curve, a combination of key parameters for thermal simulation is determined; the determined welding thermal cycle process is applied to the welded metal and welding material using a thermal simulation testing machine to prepare heat-affected specimens with different phase volume fractions and grain sizes; the key parameters for thermal simulation include initial temperature, heating rate, peak temperature, peak temperature duration, and cooling rate; the phase volume fraction and grain size of the heat-affected specimens are predicted by the phase transformation kinetics model and the grain size evolution model.

[0021] Furthermore, in step S5, the specific process of the fatigue test is as follows: based on the proposed combination of key parameters for thermal simulation, a thermal simulation testing machine is used to apply thermal cycling to the fatigue specimens corresponding to the welded metal and welding material to prepare thermally affected fatigue specimens with different phase volume fractions and grain sizes; fatigue tests are carried out on the thermally affected fatigue specimens under various stress amplitudes and stress ratios to obtain the correspondence between phase volume fraction, grain size, stress amplitude, stress ratio and fatigue life, and a fatigue life prediction model considering the influence of phase volume fraction and grain size is established based on the test results; The fatigue life prediction model is calculated using the following formula:

[0022]

[0023]

[0024] Where β, η, M, P, and Ф are material constants; R is the stress ratio; D and N represent the damage variable and the number of cycles, respectively; A0 is the equivalent stress amplitude; S ij,max and S ij,min This represents the maximum and minimum values ​​of the deviatoric stress tensor within a loading cycle.

[0025] Furthermore, the material constants β, η, M, P, and Φ in the fatigue life prediction model are calibrated using experimental results from heat-affected fatigue specimens. Specifically, this includes: preparing heat-affected fatigue specimens with different phase volume fractions and grain sizes based on different key thermal simulation parameters; conducting fatigue tests under different stress amplitudes and stress ratios to obtain the phase volume fraction, grain size, stress amplitude, stress ratio, and fatigue life data for each heat-affected fatigue specimen; based on this, the material constants β, η, M, P, and Φ in the fatigue life prediction model are inverted and calibrated; further, the relationship between the material constants β, η, M, P, and Φ and the phase volume fraction f is established. A f F f P f B f M and grain size dA d F d P d B d M The functional relationship between them; the functional relationship can be established using regression analysis or machine learning methods; the regression analysis method includes one or more of linear regression and nonlinear regression methods; the machine learning method includes one or more of artificial neural networks, support vector machines, random forests, and gradient boosting trees.

[0026] Furthermore, in step S6, when optimizing the welding process for a specific welded metal and welding material, the material parameters and welding process parameters of the welded metal and welding material are substituted into the established numerical model of the welding temperature field. The phase volume fraction and grain size of each region of the welded connection are predicted by the phase transformation kinetic model and the grain size evolution model, and the fatigue life of the welded connection is predicted by the fatigue life prediction model. A steel structure connection welding process parameter library is established based on a large number of parameter combinations, thereby selecting the process parameter combination that meets the target fatigue performance requirements. The fatigue life of the welded connection is determined according to the fatigue life predicted by the welded metal and the welding material respectively, and the smaller value of the two is taken as the actual fatigue life of the welded connection.

[0027] Beneficial effects: This invention systematically establishes the correlation between welding process parameters, welding temperature time history distribution, phase volume fraction, grain size, and fatigue life of the welded metal and welding materials, realizing multi-scale prediction of fatigue life for butt joints in steel structures. Compared with methods that only rely on nominal stress or hot spot stress for fatigue assessment, this invention can further reflect the influence of welding thermal cycles and microstructure evolution on fatigue performance, improving the pertinence and accuracy of fatigue life prediction.

[0028] This invention establishes a numerical model of the welding temperature field by combining experimental measurement and numerical simulation. By combining continuous cooling transformation curves, phase transformation kinetics models and grain size evolution models, it can predict the phase volume fraction and grain size in each region of the welded joint. This can comprehensively characterize the microstructure features during the welding process and provide a basis for fatigue life prediction.

[0029] This invention conducts thermal simulation tests and fatigue tests on the welded metal and welding materials respectively, and establishes a fatigue life prediction model that considers the effects of phase volume fraction, grain size, stress amplitude and stress ratio. This allows for a comprehensive consideration of the performance differences of different materials and regions in steel structure welded connections, thereby improving the applicability and predictive reliability of the method.

[0030] This invention reverse-engineers welding process parameters with fatigue life as the target. It can quickly screen combinations of process parameters such as welding voltage, welding current and welding speed according to the target fatigue performance requirements, reducing the workload of repeated trial welding and fatigue testing based on experience. This helps to shorten the R&D cycle, reduce testing costs and improve process optimization efficiency. Attached Figure Description

[0031] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a flowchart illustrating the main process of the steel structure connection welding process optimization method based on microstructure control as described in this embodiment of the invention. Figure 2 This is a schematic diagram of the double ellipsoidal heat source model described in an embodiment of the present invention; Figure 3 This is a schematic diagram of the continuous cooling transition curve described in an embodiment of the present invention; Figure 4 This is a schematic diagram of the welded butt joint test piece and its processing procedure as described in an embodiment of the present invention; Figure 5 This is a schematic diagram of a fatigue test specimen with weld reinforcement ground down according to an embodiment of the present invention; Figure 6 This is a microstructure morphology diagram of the heat-affected zone of the docking connection described in an embodiment of the present invention; Figure 7 This is a schematic diagram of the K-type thermocouple measuring point arrangement according to an embodiment of the present invention; Figure 8 This is a diagram showing the numerical simulation results of the welding temperature field according to an embodiment of the present invention; Figure 9 This is a comparison diagram of the measured and simulated thermal cycle time histories described in the embodiments of the present invention; Figure 10 This is a simulation result of the phase volume fraction evolution in the heat-affected zone of the welded butt joint of Q690E high-strength steel according to an embodiment of the present invention. Detailed Implementation

[0032] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0033] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0034] Definitions: To facilitate understanding of the technical solution of this invention, some terms in this document are explained as follows: Microstructure characteristics: These refer to the organizational state of steel at the microscale, mainly including parameters such as metallographic type, phase volume fraction, and grain size. The ferrite, pearlite, bainite, martensite, and austenite structures discussed in this paper are typical microstructure types formed in steel under different thermal cycling conditions. Their phase volume fraction and grain size significantly affect the fatigue performance of the material.

[0035] Continuous cooling transformation curve: refers to the curve showing the microstructure transformation of a material in different temperature ranges under continuous cooling conditions. It is used to characterize the relationship between cooling conditions, phase transformation temperature and final microstructure composition of steel.

[0036] Phase transformation kinetics model: refers to a mathematical model used to describe the transformation of various microstructures of materials with temperature or time during the welding thermal cycle.

[0037] Thermal expansion test: refers to the test method that determines the phase transition initiation temperature, phase transition termination temperature and phase transition process characteristics by measuring the dimensional changes of the sample during heating and cooling.

[0038] Thermal simulation test: refers to the test method that uses thermal simulation equipment to apply a predetermined thermal cycle process to the sample in order to reproduce the local thermal history and microstructure evolution process during welding.

[0039] Stress ratio: refers to the ratio of minimum stress to maximum stress under cyclic loading conditions.

[0040] Stress amplitude: refers to the difference between the maximum and minimum stress in a stress cycle, used to characterize the fluctuation range of cyclic load.

[0041] Thermophysical parameters: These refer to the physical parameters involved in the analysis of a material's thermal processes, including parameters such as thermal conductivity, specific heat capacity, and density.

[0042] Electron backscatter diffraction test: refers to the test method that uses electron backscatter diffraction technology in scanning electron microscopy to characterize the microstructure information of materials, such as crystal orientation, microstructure type, phase volume fraction and grain size.

[0043] Example 1 See Figure 1-10 A method for optimizing the welding process of steel structure connections based on microstructure control includes the following steps: S1. Obtain the material parameters of the metal to be welded and the welding materials, as well as the welding process parameters; In step S1, the material parameters of the metal to be welded and the welding material include the thermophysical parameters, chemical element content, volume fraction of each phase and grain size of each phase of the base metal, as well as the thermophysical parameters and chemical element content of the welding material; the welding material is welding wire or welding rod; the welding process parameters are arc welding process parameters, including welding voltage, welding current and welding speed.

[0044] It should be noted that welding voltage, welding current, and welding speed are the three parameters that collectively determine the welding heat input, which is the most critical factor affecting the welding temperature field and the final microstructure. The distribution of the welding temperature field, the evolution of the microstructure, and the level of fatigue performance are essentially the result of the combined effects of the material's inherent properties and external process conditions; only by obtaining accurate input parameters can the accuracy of all subsequent simulations and predictions be guaranteed.

[0045] S2. Based on the material parameters and welding process parameters, a numerical model of the welding temperature field is established by combining experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. In step S2, a numerical model of the welding temperature field is established using a combination of experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. Specifically, thermocouples are arranged in the heat-affected zone and its adjacent areas to measure the temperature time history during the actual welding process. Based on the measured geometric dimensions, constraints, heat dissipation conditions, thermophysical parameters of the welded metal and welding materials, and welding process parameters of the welded joint, a finite element model of the welding temperature field is established using a moving heat source model. The finite element model of the welding temperature field is then verified using the temperature time history data obtained through thermocouple measurements to obtain the welding temperature time history distribution of the welded joint. The moving heat source model includes one or more of the following: a Gaussian surface heat source model, an ellipsoidal heat source model, and a double ellipsoidal heat source model.

[0046] In a specific implementation, step S2 includes the following specific steps: S21. Determine the location of temperature measuring points based on the structural form, geometric dimensions and weld arrangement of the welded connection, and arrange K-type thermocouples in the weld area, heat-affected zone and its adjacent areas to measure the temperature changes during the actual welding process and obtain the temperature time history data of each measuring point. S22. Based on the measured geometric dimensions, constraints, heat dissipation conditions, thermophysical parameters of the welded metal and welding materials, and welding process parameters of the welded connection, a finite element model of the welding temperature field is established using a double ellipsoidal heat source model, such as... Figure 2 As shown; The mathematical equations for the double ellipsoidal heat source model are expressed as follows:

[0047]

[0048]

[0049] Where, q f (x, y, z) and q r (x, y, z) represent the heat flow rates at (x, y, z) of the front and rear halves of the ellipsoid at the heat source, respectively; η, U, and I represent the welding thermal efficiency, welding current, and welding voltage, respectively; f f and f r , respectively, represent the energy distribution coefficients of the front and rear halves of the heat source; a and b are the width and depth of the double ellipsoidal heat source, respectively; c f and c r These are the lengths of the front and rear axes of the double ellipsoidal heat source, respectively.

[0050] S23. Combining the temperature time history data obtained from each measuring point by the K-type thermocouple, the parameter f in the double ellipsoidal heat source model is... f f r a, b, c f c r Verification was performed to ensure that the numerical simulation results matched the experimental measurement results; S24. Based on the verified finite element model of the welding temperature field, the welding temperature time history distribution of each region of the welded joint is obtained.

[0051] Actual measurement of K-type thermocouples: Multiple sets of K-type thermocouples were arranged in the weld heat-affected zone and its adjacent area. Figure 7 The system collects temperature change curves in real time during the actual welding process. This is the most accurate temperature data used to calibrate the numerical model.

[0052] Modeling the double ellipsoidal heat source: The Goldak double ellipsoidal moving heat source model (a heat source model suitable for gas shielded welding) can be used, implemented by writing the DFLUX subroutine in ABAQUS. This model divides the arc into two ellipsoids, simulating the rapid heating of the front half and the slow cooling of the rear half of the arc, respectively, and can accurately reflect the three-dimensional heat distribution of the arc.

[0053] Model parameter calibration: Using thermocouple-measured temperature time history data, the six key parameters (energy distribution coefficient f) of the double ellipsoidal heat source model were inverted and calibrated. f f r Heat source width a, depth b, front and rear axis length c f c r This ensures that the error between the simulated temperature curve and the measured curve is less than 5%. Figure 9 ).

[0054] Full-field temperature output: The calibrated finite element model can output the temperature value at any position and time of the weld joint, thus obtaining the complete three-dimensional welding temperature time history distribution. Figure 8 ).

[0055] S3. Establish continuous cooling transformation curves for the welded metal and welding materials, and calibrate the phase transformation kinetics model and grain size evolution model for the welded metal and welding materials. In step S3, establishing the continuous cooling transformation curves of the welded metal and welding material, and calibrating the phase transformation kinetic model and grain size evolution model of the welded metal and welding material specifically refers to: conducting thermal expansion tests on the welded metal and welding material, applying thermal cycles with different cooling rates; acquiring thermal expansion curves under each thermal cycle condition using a dilatometer, and determining the phase transformation initiation temperature and phase transformation termination temperature using the tangent method; obtaining the volume fraction of each phase under each thermal cycle condition using electron backscattering diffraction experiments; and establishing the correspondence between cooling conditions, phase transformation temperature, and phase composition based on the phase transformation temperature range determined by the thermal expansion curves and the electron backscattering diffraction experiment results, thereby obtaining the phase transformation kinetic model and grain size evolution model of the welded metal and welding material. Obtain the continuous cooling transformation curves of the weld metal and the welding material; based on the continuous cooling transformation curves of the weld metal and the welding material, calibrate the linear interpolation model used to predict the austenite volume fraction in the heating stage, the modified Johnson–Mehl–Avrami–Kolmogorov model used to predict the ferrite, pearlite and bainite volume fractions in the cooling stage, the Koistinen–Marburger model used to predict the martensite volume fraction in the cooling stage, the Sellars model used to predict austenite grain growth in the heating stage, and the Suehiro model used to predict the ferrite, pearlite, bainite and martensite grain sizes in the cooling stage.

[0056] The linear interpolation model is calculated according to the following formula:

[0057] Among them, f A A represents the volume fraction of austenite; T represents the current temperature; A represents the volume fraction of austenite. c1 and A c3 These are the start and end temperatures of the austenite transformation, respectively.

[0058] The modified Johnson–Mehl–Avrami–Kolmogorov model is calculated according to the following formula:

[0059]

[0060] Among them, f iLet f be the phase volume fraction of the metallographic type calculated at time i, where f is the phase volume fraction for ferrite, pearlite, and bainite respectively. F f P f B ;k i and n i The kinetic parameter is related to temperature, alloy element content, and metallographic type; Δt i This is the time increment from time i-1 to time i.

[0061] The Koistinen–Marburger model is calculated according to the following formula:

[0062] Among them, f M b is the martensite volume fraction; b is the martensite transformation rate constant; M s This is the temperature at which the martensitic transformation begins.

[0063] The Sellars model is calculated using the following formula:

[0064] Where, d A d0 is the average austenite grain size during the heating stage; t is the initial parent material grain size; Q is the equivalent holding time; R is the activation energy for grain growth; A is the gas constant, taken as 8.314 J / (mol·K); and A and n are material constants.

[0065] The Suehiro model is calculated using the following formula:

[0066] Where d is the average grain diameter of each metallographic structure when cooled to room temperature, and d is the average grain diameter of ferrite, pearlite, bainite, and martensite, respectively. F d P d B d M ;d max T represents the average grain diameter at the end of the heating phase. 5% The temperature at which the volume fraction of each metallographic phase reaches 5% during the cooling process; C1 and K are material constants; f is the volume fraction of the metallographic structure calculated when cooled to room temperature.

[0067] Furthermore, the parameter A in the linear interpolation model c1 and A c3 The parameter k in the modified Johnson–Mehl–Avrami–Kolmogorov model i and n i The parameter M in the Koistinen–Marburger models b, obtained through continuous cooling transformation curves; the parameters Q, A, and n in the Sellars model, and the parameters C1 and K in the Suehiro model, obtained through electron backscattering diffraction experiments and phase transition kinetic models.

[0068] In a specific implementation, step S3 includes the following specific steps: S31. Prepare thermal expansion specimens for the metal to be welded and the welding material respectively, and apply a preset thermal cycle to the thermal expansion specimens using a thermal simulation testing machine; the preset thermal cycle includes a heating stage and a cooling stage, and sets multiple sets of different peak temperatures and cooling rates to simulate the thermal cycle conditions that different regions may experience during the welding process. S32. The length change curves of thermal expansion samples under each group of thermal cycling conditions are collected by a dilatometer, and the phase change initiation temperature and phase change termination temperature under each group of thermal cycling conditions are determined by the tangent method based on the length change curves. S33. Microstructure characterization of thermal expansion samples was carried out using an electron backscatter diffraction system in conjunction with a field emission scanning electron microscope to obtain the phase volume fraction and grain size of austenite, ferrite, pearlite, bainite and martensite in each group of thermal expansion samples. S34. Combining the phase transformation initiation temperature, phase transformation termination temperature, and phase volume fraction results obtained from electron backscatter diffraction experiments, establish the correspondence between cooling rate, phase transformation temperature, and microstructure composition. Obtain the continuous cooling transformation curves of the welded metal and welding material according to YB / T 5128-2018, "Determination Method of Continuous Cooling Transformation Curve of Steel (Expansion Method)". Figure 3 As shown; S35. Based on the continuous cooling transformation curve, the variation law of austenite volume fraction during the heating stage is fitted, and the parameters A of the linear interpolation model used to predict the austenite volume fraction are calibrated. c1 and A c3 ; The linear interpolation model is calculated according to the following formula:

[0069] Among them, f A A represents the volume fraction of austenite; T represents the current temperature; A represents the volume fraction of austenite. c1 and A c3 These are the start and end temperatures of the austenite transformation, respectively. S36. Based on the continuous cooling transformation curve, the evolution law of the volume fraction of ferrite, pearlite and bainite during the cooling stage is fitted, and the kinetic parameter k in the modified Johnson–Mehl–Avrami–Kolmogorov model is calibrated. i and n iThe evolution law of martensite volume fraction was fitted, and the parameter M in the Koistinen–Marburger model was calibrated. s and b; The modified Johnson–Mehl–Avrami–Kolmogorov model is calculated according to the following formula:

[0070]

[0071] Among them, f i Let f be the phase volume fraction of the metallographic type calculated at time i, where f is the phase volume fraction for ferrite, pearlite, and bainite respectively. F f P f B ;k i and n i The kinetic parameter is related to temperature, alloy element content, and metallographic type; Δt i The time increment from time i-1 to time i; The Koistinen–Marburger model is calculated according to the following formula:

[0072] Among them, f M b is the martensite volume fraction; b is the martensite transformation rate constant; M s This is the temperature at which the martensitic transformation begins. S37. Based on the grain size data and phase transformation kinetic model obtained from electron backscatter diffraction experiments, the austenite grain growth law during the heating stage is fitted, and the parameters Q, A and n in the Sellars model are calibrated; the grain size evolution law of ferrite, pearlite, bainite and martensite before and after cooling is fitted, and the parameters C1 and K in the Suehiro model are calibrated. The Sellars model is calculated using the following formula:

[0073] Where, d A d0 is the average austenite grain size during the heating stage; t is the initial parent material grain size; Q is the equivalent holding time; R is the activation energy for grain growth; A is the gas constant, taken as 8.314 J / (mol·K); and A and n are material constants. The Suehiro model is calculated using the following formula:

[0074] Where d is the average grain diameter of each metallographic structure when cooled to room temperature, and d is the average grain diameter of ferrite, pearlite, bainite, and martensite, respectively.F d P d B d M ;d max T represents the average grain diameter at the end of the heating phase. 5% The temperature at which the volume fraction of each metallographic phase reaches 5% during the cooling process; C1 and K are material constants; f is the volume fraction of the metallographic structure calculated when cooled to room temperature.

[0075] It is understandable that the microstructure transformation of steel under welding thermal cycling is a kinetic process closely related to temperature and time. This step establishes a benchmark for the phase transformation law of steel (CCT curve) through experiments and calibrates the parameters of a dedicated phase transformation and grain evolution model, providing mathematical tools for subsequent microstructure prediction.

[0076] S4. Combining the welding temperature time history distribution, phase transformation kinetics model, and grain size evolution model of the welded joint, obtain the phase volume fraction and grain size of each region of the welded joint; In a specific implementation, step S4 includes the following specific steps: S41. Substitute the welding temperature time history distribution of each region of the welded joint obtained in step S24 into the phase transformation kinetic model established in steps S35 and S36, calculate the evolution process of austenite volume fraction during the heating stage, and the evolution process of ferrite, pearlite, bainite and martensite volume fraction during the cooling stage, and obtain the phase volume fraction of each region after cooling to room temperature. S42. Substitute the welding temperature time history distribution of each region of the welded joint obtained in step S24 into the phase transformation kinetic model established in steps S35 and S36 and the grain size evolution model established in step S37 to calculate the austenite grain growth process during the heating stage and the grain size of ferrite, pearlite, bainite and martensite after cooling.

[0077] This step is a quantitative mapping process from temperature field to microstructure. The principle is to substitute the temperature time history curve of each finite element obtained in step S2 into the microstructure evolution model calibrated in step S3 according to the time step, and calculate the final microstructure characteristics of each region after cooling to room temperature.

[0078] S5. Conduct thermal simulation tests and fatigue tests on the welded metal and welding materials in sequence, and construct a fatigue life prediction model that considers the effects of phase volume fraction and grain size. In step S5, the specific procedure for the thermal simulation test is as follows: referring to the actual welding thermal cycle process and the established continuous cooling transformation curve, a combination of key parameters for thermal simulation is determined; the determined welding thermal cycle process is applied to the welded metal and welding material using a thermal simulation testing machine to prepare heat-affected specimens with different phase volume fractions and grain sizes; the key parameters for thermal simulation include initial temperature, heating rate, peak temperature, peak temperature duration, and cooling rate; the phase volume fraction and grain size of the heat-affected specimens are predicted by the phase transformation kinetics model and the grain size evolution model.

[0079] In step S5, the specific process of the fatigue test is as follows: based on the proposed combination of key parameters for thermal simulation, a thermal simulation testing machine is used to apply thermal cycling to the fatigue specimens corresponding to the welded metal and the welding material to prepare thermally affected fatigue specimens with different phase volume fractions and grain sizes; fatigue tests are carried out on the thermally affected fatigue specimens under various stress amplitudes and stress ratios to obtain the correspondence between phase volume fraction, grain size, stress amplitude, stress ratio and fatigue life, and a fatigue life prediction model considering the influence of phase volume fraction and grain size is established based on the test results; The fatigue life prediction model is calculated using the following formula:

[0080]

[0081]

[0082] Where β, η, M, P, and Ф are material constants; R is the stress ratio; D and N represent the damage variable and the number of cycles, respectively; A0 is the equivalent stress amplitude; S ij,max and S ij,min This represents the maximum and minimum values ​​of the deviatoric stress tensor within a loading cycle.

[0083] Furthermore, the material constants β, η, M, P, and Φ in the fatigue life prediction model are calibrated using experimental results from heat-affected fatigue specimens. Specifically, this includes: preparing heat-affected fatigue specimens with different phase volume fractions and grain sizes based on different key thermal simulation parameters; conducting fatigue tests under different stress amplitudes and stress ratios to obtain the phase volume fraction, grain size, stress amplitude, stress ratio, and fatigue life data for each heat-affected fatigue specimen; based on this, the material constants β, η, M, P, and Φ in the fatigue life prediction model are inverted and calibrated; further, the relationship between the material constants β, η, M, P, and Φ and the phase volume fraction f is established. A f F f P f B f M and grain size d A dF d P d B d M The functional relationship between them; the functional relationship can be established using regression analysis or machine learning methods; the regression analysis method includes one or more of linear regression and nonlinear regression methods; the machine learning method includes one or more of artificial neural networks, support vector machines, random forests, and gradient boosting trees.

[0084] In a specific implementation, step S5 includes the following specific steps: S51. Referring to the actual welding thermal cycle process, multiple sets of key thermal simulation parameters are proposed for the welded metal and the welding material respectively; the key thermal simulation parameters include initial temperature, heating rate, peak temperature, peak temperature duration and cooling rate. S52. Using a thermal simulation testing machine, thermal cycling processes corresponding to the combination of key thermal simulation parameters are applied to the welded metal and welding material samples to prepare thermally affected fatigue specimens with different phase volume fractions and grain sizes. S53. Based on the phase transformation kinetics model established in steps S35 and S36 and the grain size evolution model established in step S37, predict the phase volume fraction and grain size of each heat-affected fatigue specimen; or use electron backscatter diffraction test to determine the phase volume fraction and grain size of each heat-affected fatigue specimen. S54. Under multiple stress amplitudes and stress ratios, fatigue tests are carried out on the heat-affected fatigue specimens to obtain fatigue life data of each heat-affected fatigue specimen under different phase volume fractions, grain sizes, stress amplitudes and stress ratios. S55. Based on the fatigue life data, the material constants in the fatigue life prediction model are inverted and calibrated to obtain a fatigue life prediction model applicable to the welded metal and the welding material. The fatigue life prediction model is calculated using the following formula:

[0085]

[0086]

[0087] Where β, η, M, P, and Ф are material constants; R is the stress ratio; D and N represent the damage variable and the number of cycles, respectively; A0 is the equivalent stress amplitude; S ij,max and S ij,min This represents the maximum and minimum values ​​of the deviatoric stress tensor within a loading cycle.

[0088] S56. Further, based on the phase volume fraction, grain size, and corresponding calibration results of the heat-affected fatigue specimen, establish the material constants β, η, M, P, Φ, and phase volume fraction f. A f F f P f B f M and grain size d A d F d P d B d M The functional relationship between them; S57. Based on the aforementioned functional relationship, a fatigue life prediction model considering the effects of phase volume fraction and grain size is constructed by using phase volume fraction, grain size, stress amplitude, and stress ratio as input parameters.

[0089] This step breaks through the limitation of the traditional fatigue model's "fixed material constants" by establishing a quantitative correlation between the material constants of the fatigue model and microstructure parameters (phase volume fraction, grain size), so that the same model can be applied to all regions with different microstructures in the welded joint.

[0090] S6. Taking fatigue life as the optimization target, and combining the prediction results of phase volume fraction and grain size, the welding process parameters are reverse-engineered through the fatigue life prediction model to obtain a combination of process parameters that meet the target fatigue performance requirements.

[0091] In step S6, when optimizing the welding process for a specific welded metal and welding material, the material parameters and welding process parameters of the welded metal and welding material are substituted into the established numerical model of the welding temperature field. The phase volume fraction and grain size of each region of the welded joint are predicted by the phase transformation kinetic model and the grain size evolution model, and the fatigue life of the welded joint is predicted by the fatigue life prediction model. A welding process parameter library for steel structure welded joints is established based on a large number of parameter combinations, thereby selecting the process parameter combination that meets the target fatigue performance requirements. The fatigue life of the welded joint is determined according to the fatigue life predicted by the welded metal and welding material respectively, and the smaller value of the two is taken as the actual fatigue life of the welded joint.

[0092] In a specific implementation, step S6 includes the following specific steps: S61. For a specific welded metal and welding material, determine the range of welding process parameters to be optimized, and establish a combination of welding process parameters; the welding process parameters are arc welding process parameters, including welding voltage, welding current and welding speed. S62. Substitute each group of welding process parameter combinations and the corresponding welded metal and welding material parameters into the welding temperature field finite element model established in step S22 and verified in step S23 to obtain the welding temperature time history distribution of each region of the welded connection under each group of welding process parameter combinations. S63. Substitute the welding temperature time history distribution obtained under each combination of welding process parameters into the phase transformation kinetic model established in steps S35 and S36 and the grain size evolution model established in step S37, and calculate the phase volume fraction and grain size of each region of the welded connection. S64. Substitute the phase volume fraction, grain size, and stress amplitude and stress ratio under service load conditions into the fatigue life prediction model constructed in steps S55 and S57 to obtain the fatigue life of the welded metal and the fatigue life of the welding material corresponding to each combination of welding process parameters. S65. Determine the fatigue life evaluation value of the welded connection based on the fatigue life of the welded metal and the fatigue life of the welding material, and preferably take the smaller value of the two as the fatigue life of the welded connection corresponding to the combination of welding process parameters. S66. Compare the fatigue life of the welded connection corresponding to each group of welding process parameter combinations, select the welding process parameter combinations that meet the target fatigue performance requirements, and establish a welding process parameter library for steel structure welded connections. S67. Based on the welding process parameter library for steel structure welding connections, select the welding process parameter combination with the maximum fatigue life or that meets the preset fatigue life requirement and has the best process feasibility as the optimization result.

[0093] This step utilizes the previously established full-chain prediction model to traverse all possible combinations of process parameters, calculate the fatigue life corresponding to each set of parameters, and finally select the optimal process that meets the requirements, replacing the traditional trial-and-error method.

[0094] In summary, this embodiment effectively solves the problem that existing technologies do not fully consider the intrinsic relationship between welding thermal cycling, microstructure evolution, and fatigue life, and realizes a systematic correlation between welding process parameters, welding temperature time history distribution, phase volume fraction, grain size, stress amplitude, stress ratio, and fatigue life. By combining experimental measurement with numerical simulation, the microstructure characteristics and fatigue life of various regions in welded steel structure connections under different welding process parameters can be predicted, and based on this, reverse optimization design of welding process parameters can be completed. Compared with traditional methods that rely on experience to select process parameters and involve numerous repeated experiments, this embodiment can reduce experimental workload, lower R&D costs, improve welding process design efficiency and welded connection fatigue performance, and has good engineering application value.

[0095] Data verification: In practice, two sets of gas-shielded welding processes were used to weld Q690E high-strength steel to prepare welded butt joint test pieces, such as... Figure 4 As shown. The test piece prepared using the first set of welding process parameters (220A welding current, 24.8V welding voltage, 4.5mm / s welding speed) was named Q690E-A; the test piece prepared using the second set of welding process parameters (220A welding current, 24.6V welding voltage, 3.2mm / s welding speed) was named Q690E-B. Subsequently, the butt joint test pieces were processed into fatigue test pieces using wire cutting equipment to reduce the influence of welding residual stress on the fatigue test results; and the weld reinforcement of the fatigue test pieces was ground smooth to reduce the stress concentration effect caused by the weld reinforcement, such as... Figure 5 This is a schematic diagram of a fatigue test specimen with the weld reinforcement ground down. Figure 5 Figure a shows a fatigue specimen of a butt joint (the weld excess has been ground down for comparison). Figure 5 Figure b shows the fatigue specimen of the base material (which itself has no excess height). Subsequently, fatigue tests were conducted on fatigue specimens prepared using different welding process parameters and on fatigue specimens of the base material. The test results are shown in Table 1.

[0096] Furthermore, material samples were taken from the coarse-grained and fine-grained regions of the heat-affected zone of the butt joint and metallographic samples were prepared. The microstructure morphology of the butt joint was observed using a scanning electron microscope, such as... Figure 6 As shown, Figure 6 Figure a shows the base material. Figure 6 Figure b shows the coarse-grained region of Q690E-A. Figure 6 The middle image (c) shows the fine-grained region of Q690E-A. Figure 6 The middle d-axis diagram represents the coarse-grained region of Q690E-B. Figure 6 Figure e shows the Q690E-B fine-grained region; electron backscatter diffraction (ESD) was used to quantitatively obtain the phase volume fractions and overall weighted average grain size of martensite, bainite, ferrite, pearlite, and austenite structures. The overall weighted average grain size is calculated using the formula... Calculations were performed. The test results are shown in Table 2.

[0097] Table 1

[0098] Table 2

[0099] Table 1 shows that the average number of stress cycles for fatigue specimens prepared using the first set of welding process parameters, the second set of welding process parameters, and the base material were 125,472, 97,805, and 355,714, respectively. Since the fracture locations of the butt-joint fatigue specimens were all located in the fine-grained region far from the weld, the main differences compared to the base material lay in the phase volume fraction and grain size. Correspondingly, the fatigue lives of the Q690E-A and Q690E-B fatigue specimens were approximately 35% and 27% of those of the base material fatigue specimens, respectively, and the difference in fatigue life between the two sets of welding process parameters was significant. These results demonstrate that phase volume fraction and grain size have a significant impact on fatigue life, thus verifying the necessity and rationality of constructing a fatigue life prediction model considering the influence of phase volume fraction and grain size in this invention.

[0100] In a specific example, two sets of gas-shielded welding processes were used to weld Q690E high-strength steel, and K-type thermocouples were used to measure the thermal cycle time history of the weld heat-affected zone and its adjacent areas. Figure 4 and Figure 7 As shown. Subsequently, ABAQUS software was used to conduct corresponding welding temperature field simulations. A finite element model was established based on the measured geometric dimensions of the specimen, constraint conditions, heat dissipation conditions, and thermophysical parameters. A Goldak double-ellipsoidal moving heat source model was constructed by writing the DFLUX subroutine. The mathematical equations of the double-ellipsoidal heat source model are expressed as follows:

[0101]

[0102]

[0103] Where, q f (x, y, z) and q r (x, y, z) represent the heat flow rates at (x, y, z) of the front and rear halves of the ellipsoid at the heat source, respectively; η, U, and I represent the welding thermal efficiency, welding current, and welding voltage, respectively; f f and f r , respectively, represent the energy distribution coefficients of the front and rear halves of the heat source; a and b are the width and depth of the double ellipsoidal heat source, respectively; c f and c r These represent the lengths of the front and rear axes of the double ellipsoidal heat source, respectively. The measured and simulated welding temperature fields and thermal cycling time histories are shown below. Figure 8 and Figure 9 Therefore, the numerical model of welding temperature field established in this invention can accurately reflect the temperature field distribution and thermal cycling time history variation law during the welding process, thus verifying the feasibility and accuracy of predicting the welding temperature time history distribution by combining experimental measurement and numerical simulation.

[0104] In a specific example, based on the chemical element content and grain size of the base material, the continuous cooling transformation curve of Q690E high-strength steel was obtained using JMatPro software. A linear interpolation model was established in the ABAQUS finite element model using the USDFLD subroutine to predict the austenite volume fraction during the heating stage, a modified Johnson–Mehl–Avrami–Kolmogorov model for predicting the ferrite, pearlite, and bainite volume fractions during the cooling stage, and a Koistinen–Marburger model for predicting the martensite volume fraction during the cooling stage. The phase volume fraction evolution of the heat-affected zone of Q690E high-strength steel during welding was simulated. The simulation results of the phase volume fraction evolution in the coarse-grained region of the butt joint of Q690E high-strength steel, using a welding current of 220A, a welding voltage of 24.6V, and a welding speed of 3.2mm / s, are shown below. Figure 10 As shown in the figure; the corresponding metallographic structure content measured by electron backscatter diffraction is shown in the Q690E-B coarse-grained region data in Table 2. Therefore, the phase transformation kinetic model established in this invention can accurately characterize the evolution of phase volume fraction under welding thermal cycling, thus verifying the feasibility and effectiveness of this invention in predicting the evolution of butt joint microstructure.

[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for optimizing the welding process of steel structure connections based on microstructure control, characterized in that, Includes the following steps: S1. Obtain the material parameters of the metal to be welded and the welding materials, as well as the welding process parameters; S2. Based on the material parameters and welding process parameters, a numerical model of the welding temperature field is established by combining experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. S3. Establish continuous cooling transformation curves for the welded metal and welding materials, and calibrate the phase transformation kinetics model and grain size evolution model for the welded metal and welding materials. S4. Combining the welding temperature time history distribution, phase transformation kinetics model, and grain size evolution model of the welded joint, obtain the phase volume fraction and grain size of each region of the welded joint; S5. Conduct thermal simulation tests and fatigue tests on the welded metal and welding materials in sequence, and construct a fatigue life prediction model that considers the effects of phase volume fraction and grain size. S6. Taking fatigue life as the optimization target, and combining the prediction results of phase volume fraction and grain size, the welding process parameters are reverse-engineered through the fatigue life prediction model to obtain a combination of process parameters that meet the target fatigue performance requirements.

2. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S1, the material parameters of the metal to be welded and the welding material include the thermophysical parameters, chemical element content, volume fraction of each phase and grain size of each phase of the base metal, as well as the thermophysical parameters and chemical element content of the welding material; the welding material is welding wire or welding rod; the welding process parameters are arc welding process parameters, including welding voltage, welding current and welding speed.

3. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S2, a numerical model of the welding temperature field is established using a combination of experimental measurement and numerical simulation to obtain the welding temperature time history distribution of the welded joint. Specifically, thermocouples are arranged in the heat-affected zone and its adjacent areas to measure the temperature time history during the actual welding process. Based on the measured geometric dimensions, constraints, heat dissipation conditions, thermophysical parameters of the welded metal and welding materials, and welding process parameters of the welded joint, a finite element model of the welding temperature field is established using a moving heat source model. The finite element model of the welding temperature field is then verified using the temperature time history data obtained through thermocouple measurements to obtain the welding temperature time history distribution of the welded joint. The moving heat source model includes one or more of the following: a Gaussian surface heat source model, an ellipsoidal heat source model, and a double ellipsoidal heat source model.

4. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S3, establishing the continuous cooling transformation curves of the welded metal and welding material, and calibrating the phase transformation kinetic model and grain size evolution model of the welded metal and welding material specifically refers to: conducting thermal expansion tests on the welded metal and welding material, applying thermal cycles with different cooling rates; acquiring thermal expansion curves under each thermal cycle condition using a dilatometer, and determining the phase transformation initiation temperature and phase transformation termination temperature using the tangent method; obtaining the volume fraction of each phase under each thermal cycle condition using electron backscattering diffraction experiments; and establishing the correspondence between cooling conditions, phase transformation temperature, and phase composition based on the phase transformation temperature range determined by the thermal expansion curves and the electron backscattering diffraction experiment results, thereby obtaining the phase transformation kinetic model and grain size evolution model of the welded metal and welding material. Obtain the continuous cooling transformation curves of the weld metal and the welding material; based on the continuous cooling transformation curves of the weld metal and the welding material, calibrate the linear interpolation model used to predict the austenite volume fraction in the heating stage, the modified Johnson–Mehl–Avrami–Kolmogorov model used to predict the ferrite, pearlite and bainite volume fractions in the cooling stage, the Koistinen–Marburger model used to predict the martensite volume fraction in the cooling stage, the Sellars model used to predict austenite grain growth in the heating stage, and the Suehiro model used to predict the ferrite, pearlite, bainite and martensite grain sizes in the cooling stage.

5. The method for optimizing steel structure connection welding process based on microstructure control according to claim 4, characterized in that, The linear interpolation model is calculated according to the following formula: Among them, f A A represents the volume fraction of austenite; T represents the current temperature; A represents the volume fraction of austenite. c1 and A c3 These are the start and end temperatures of the austenite transformation, respectively. The modified Johnson–Mehl–Avrami–Kolmogorov model is calculated according to the following formula: Among them, f i Let f be the phase volume fraction of the metallographic type calculated at time i, where f is the phase volume fraction for ferrite, pearlite, and bainite respectively. F f P f B ;k i and n i The kinetic parameter is related to temperature, alloy element content, and metallographic type; Δt i The time increment from time i-1 to time i; The Koistinen–Marburger model is calculated according to the following formula: Among them, f M b is the martensite volume fraction; b is the martensite transformation rate constant; M s This is the temperature at which the martensitic transformation begins. The Sellars model is calculated using the following formula: Where, d A d0 is the average austenite grain size during the heating stage; t is the initial parent material grain size; Q is the equivalent holding time; R is the activation energy for grain growth; A is the gas constant, taken as 8.314 J / (mol·K); and A and n are material constants. The Suehiro model is calculated using the following formula: Where d is the average grain diameter of each metallographic structure when cooled to room temperature, and d is the average grain diameter of ferrite, pearlite, bainite, and martensite, respectively. F d P d B d M ;d max T represents the average grain diameter at the end of the heating phase. 5% The temperature at which the volume fraction of each metallographic phase reaches 5% during the cooling process; C1 and K are material constants; f is the volume fraction of the metallographic structure calculated when cooled to room temperature.

6. The method for optimizing steel structure connection welding process based on microstructure control according to claim 5, characterized in that, The parameter A in the linear interpolation model c1 and A c3 The parameter k in the modified Johnson–Mehl–Avrami–Kolmogorov model i and n i The parameter M in the Koistinen–Marburger model s b, obtained through continuous cooling transformation curves; the parameters Q, A, and n in the Sellars model, and the parameters C1 and K in the Suehiro model, obtained through electron backscattering diffraction experiments and phase transition kinetic models.

7. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S5, the specific procedure for the thermal simulation test is as follows: referring to the actual welding thermal cycle process and the established continuous cooling transformation curve, a combination of key parameters for thermal simulation is determined; the determined welding thermal cycle process is applied to the welded metal and welding material using a thermal simulation testing machine to prepare heat-affected specimens with different phase volume fractions and grain sizes; the key parameters for thermal simulation include initial temperature, heating rate, peak temperature, peak temperature duration, and cooling rate; the phase volume fraction and grain size of the heat-affected specimens are predicted by the phase transformation kinetics model and the grain size evolution model.

8. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S5, the specific process of the fatigue test is as follows: based on the proposed combination of key parameters for thermal simulation, a thermal simulation testing machine is used to apply thermal cycling to the fatigue specimens corresponding to the welded metal and the welding material to prepare thermally affected fatigue specimens with different phase volume fractions and grain sizes. Fatigue tests were conducted on the thermally affected fatigue specimens under various stress amplitudes and stress ratios to obtain the corresponding relationships between phase volume fraction, grain size, stress amplitude, stress ratio and fatigue life. Based on the test results, a fatigue life prediction model considering the effects of phase volume fraction and grain size was established. The fatigue life prediction model is calculated using the following formula: Where β, η, M, P, and Ф are material constants; R is the stress ratio; D and N represent the damage variable and the number of cycles, respectively; A0 is the equivalent stress amplitude; S ij,max and S ij,min This represents the maximum and minimum values ​​of the deviatoric stress tensor within a loading cycle.

9. The method for optimizing steel structure connection welding process based on microstructure control according to claim 8, characterized in that, The material constants β, η, M, P, and Φ in the fatigue life prediction model are calibrated using experimental results from heat-affected fatigue (HAF) specimens. Specifically, this includes: preparing HAF specimens with different phase volume fractions and grain sizes based on different key thermal simulation parameters; conducting fatigue tests under different stress amplitudes and stress ratios to obtain phase volume fractions, grain sizes, stress amplitudes, stress ratios, and fatigue life data for each HAF specimen; based on this, the material constants β, η, M, P, and Φ in the fatigue life prediction model are inverted and calibrated; further, a relationship is established between the material constants β, η, M, P, and Φ and the phase volume fraction f. A f F f P f B f M and grain size d A d F d P d B d M The functional relationship between them; the functional relationship can be established using regression analysis or machine learning methods; the regression analysis method includes one or more of linear regression and nonlinear regression methods; the machine learning method includes one or more of artificial neural networks, support vector machines, random forests, and gradient boosting trees.

10. The method for optimizing steel structure connection welding process based on microstructure control according to claim 1, characterized in that, In step S6, when optimizing the welding process for a specific welded metal and welding material, the material parameters and welding process parameters of the welded metal and welding material are substituted into the established numerical model of the welding temperature field. The phase volume fraction and grain size of each region of the welded joint are predicted by the phase transformation kinetic model and the grain size evolution model, and the fatigue life of the welded joint is predicted by the fatigue life prediction model. A steel structure connection welding process parameter library is established based on a large number of parameter combinations, thereby selecting the process parameter combination that meets the target fatigue performance requirements. The fatigue life of the welded joint is determined according to the fatigue life predicted by the welded metal and welding material respectively, and the smaller value of the two is taken as the actual fatigue life of the welded joint.

Images