Integrated analysis method and system for multi-scale damage of offshore wind power structure

Abstract

The application discloses a kind of offshore wind power structure multiscale damage integrated analysis method and system, method includes: based on the principle of continuum mechanics, by representative volume element (RVE) homogenization method, anisotropic elastoplasticity equivalent constitutive model of flange bolt connection is established, to replace the traditional bolt and flange split fine modeling mode.The equivalent flange model established is realized by parameterization calibration Macroscopic mechanical performance characterization of flange connection area, can accurately reflect its stiffness, strength and dynamic characteristics.Using the model can efficiently predict the nonlinear response of the connection part in the overall structure analysis, realize the collaborative analysis of global response and local damage of overall structure.Compared with traditional fine model, the application significantly improves the calculation efficiency under the premise of ensuring calculation accuracy, the analysis time can be reduced by about 62%, and stress distribution and dynamic response characteristics can be accurately reflected.

Description

Technical Field

[0001] This invention relates to the field of marine engineering structure analysis technology, and more specifically, to an integrated analysis method and system for multi-scale damage of offshore wind power structures. Background Technology

[0002] As a clean and renewable energy source, wind energy boasts advantages such as sustainability, low cost, and high power generation rate, playing an increasingly important role in the global energy structure. With technological advancements and cost reductions, wind power has become a crucial component of energy supply for many countries and regions. In recent years, offshore wind turbines have achieved rapid development due to their higher power generation efficiency and abundant wind energy resources. As a key device for converting wind energy into electricity, the performance and reliability of wind turbines directly affect the economic benefits and environmental impact of wind power projects. Compared to onshore wind turbines, offshore wind turbines must operate in harsh environments such as typhoons, hurricanes, earthquakes, and tsunamis, posing significant challenges to their structural stability and durability.

[0003] In existing studies on the dynamic response of offshore wind turbines under wind-wave coupling, numerical models typically simplify bolted connections by ignoring the modeling of the bolted components themselves and neglecting the contact interaction between the bolts and flanges. This simplification method improves modeling efficiency and reduces computational costs to some extent, especially in global dynamic response analysis. However, in extreme load studies, the local response of bolted connections often has a significant impact on overall performance.

[0004] The flange bolted connection area presents a computational challenge due to its complex nonlinear contact behavior. Detailed modeling typically requires high-density meshes and numerous nonlinear contact settings, leading to excessive computational resource consumption and low solution efficiency, hindering widespread implementation in practical research and engineering applications. Therefore, developing a low-cost analysis method is crucial, accurately capturing the dynamic interaction mechanism of the foundation-tower under wind-wave-current coupled loads while maintaining high computational efficiency. Some researchers have successfully analyzed the dynamic response of pile-supported wharves under seismic loads by simplifying non-critical areas using beam elements. Others have developed multi-scale models for reinforced concrete structures and verified their effectiveness in balancing computational efficiency and accuracy through comparison with experimental data. Currently, some researchers have proposed a multi-scale analysis method for monopile-supported offshore wind turbines, studying the overall structural response and local bolt damage evolution under extreme wind-wave coupled loads through refined finite element analysis of the wind turbine connecting bolts. Simplifying the settings of non-critical components is an effective means to improve computational efficiency; however, for wind turbines, the main consumption of computational resources is concentrated in the bolted connection area and the flange structure between the tower and the foundation. Bolted connections exhibit complex contact nonlinearity, preload, and stress concentration effects, typically requiring numerous fine meshes and contact pairs in the finite element model, along with explicit dynamic algorithms for solution. This leads to a dramatic increase in computational scale. Consequently, existing analysis methods for offshore wind turbines cannot balance the challenge of computational accuracy with computational efficiency. Summary of the Invention

[0005] To address the aforementioned shortcomings or improvement needs of existing technologies, this invention proposes a novel and efficient computational method. Based on the principles of continuum mechanics, an equivalent anisotropic elastoplastic constitutive model for flange-bolt connections is developed using the Representative Volume Element (RVE) homogenization method. This method replaces the traditional approach of modeling bolts and flanges separately with a parametrically calibrated homogenized flange model. This equivalent model can efficiently predict the nonlinear response of the connection in the overall structural analysis, effectively solving the flange-bolt modeling problem and significantly improving overall computational efficiency. A two-stage analysis method is employed. In the first stage, a homogenized equivalent model is established through parametric calibration, replacing the traditional separate modeling method of fine modeling of bolts and flanges. This equivalent model can accurately reflect the stiffness, strength, and dynamic response characteristics of the connection node at the macroscopic scale, thereby significantly reducing the computational scale of the overall system analysis. In the second stage, based on the stress analysis results of the overall structural model, a sub-model technique is used to perform local fine modeling of the critical flange-bolt region. Specifically, the displacement response of the overall model at the boundary of this region is used as the driving boundary condition of the sub-model, thereby reconstructing its accurate stress state at a fine scale. In the refined modeling sub-model, the GTN damage model is introduced to simulate the microscopic damage evolution process of bolt materials under load. Through a bidirectional data transmission mechanism between the macro and local models, the computational accuracy and efficiency of the complex model under multi-field coupled loads are collaboratively optimized. Finally, based on the established parametric equivalent model, the overall response characteristics of the wind turbine generator under seismic and wave loads are analyzed.

[0006] To achieve the above objectives, this invention provides an integrated analysis method for multi-scale damage to offshore wind power structures, comprising: S100: Establish an overall structural analysis model for offshore wind turbines; S200: A multi-scale numerical model of flange-bolt microstructure is established based on the GTN damage model. Based on the principle of continuous medium, a homogenized flange model is established according to the overall stiffness and stress characteristics of the flange-bolt connection structure. S300: Substitute the homogenized equivalent model established in step S200 into the overall structure and analyze the dynamic response of the overall structure under various loads such as wind, waves, and earthquakes. S400: The sub-model technique is used to evaluate the flange-bolt connection structure, establish a refined flange-bolt connection structure, and substitute the flange-bolt stress state in the overall structure calculation results obtained in step S300 into the sub-model to achieve collaborative analysis of global response and local damage.

[0007] Further, in step S100, the establishment of the overall structural analysis model includes: S101: Establish a corresponding numerical model based on the actual size of the offshore wind turbine, and adopt a three-pile rock-embedded jacket foundation for the wind farm foundation.

[0008] Further, in step S100, the establishment of the overall structural analysis model includes: S102: The offshore wind turbine consists of a rotor, nacelle, tower and support foundation.

[0009] Further, in step S100, the establishment of the overall structural analysis model includes: S103: The simulation of the grouting material uses the concrete plastic damage model built into the ABAQUS software. This model captures the stiffness decay characteristics of concrete through the plastic damage factor. The compressive constitutive model of the grouting material adopts the Minder model. , , , in, It refers to the compressive strength of concrete. is the compressive strength of the concrete cylinder; x is the strain ratio; r is the curvature parameter. It is concrete strain. It is the peak strain of the concrete; It is the elastic modulus of concrete; E sec Yes, it refers to the stiffness of the concrete cleavage. The constitutive model under tension is shown below: , , , , in, It is the tensile damage factor; For stress; It is the elastic modulus of concrete; x is the concrete strain; x is the strain ratio. It refers to the tensile strength of concrete; It is the peak strain of the concrete; The strength-stiffness ratio parameter; This represents the softening rate.

[0010] Further, in step S100, the establishment of the overall structural analysis model includes: The damage factor of the grouting material is calculated as follows: , , in, It is a pressure-induced damage factor; It is the tensile damage factor; , These are the tensile and compressive stresses in concrete; , This represents the peak stress of concrete under tensile and compressive conditions. , Equivalent stiffness adjustment coefficients for compression and tension, respectively.

[0011] Further, in step S100, the establishment of the overall structural analysis model includes: S104: In this invention, since the pipe pile has a circular cross-section, in order to prevent the main surface and the subordinate surface from penetrating each other, the rock body is set as the main surface. At the same time, the normal behavior of the contact surface is set as hard contact, that is, the two surfaces are allowed to separate but not to penetrate each other. The tangential direction adopts the Coulomb friction model. The bolt and the flange are set as hard contact and tangential friction contact. The friction coefficient is usually between 0.2 and 0.6.

[0012] Further, in step S100, the establishment of the overall structural analysis model includes: S105: Initial load refers to the inherent load that the structure bears before it is put into operation and subjected to the action of the external environment. It mainly includes the self-weight of the entire structure (static load) and the bolt preload applied after installation. The self-weight of the structure consists of the sum of the static forces of all components such as the tower, monopile foundation, blades, nacelle, hub and connecting bolts. Before applying gravity loads to the structure, a static equilibrium analysis of the rock mass must be performed. The specific process includes: first, establishing a finite element model of the rock mass and setting material parameters; then, applying gravity loads and calculating the initial stress field under gravity by combining boundary conditions that reflect the actual engineering scenario; extracting the obtained stress field as the initial condition for the main analysis and importing it into the subsequent numerical model according to the principle of initial stress equilibrium; finally, reapplying gravity loads to ensure that the internal stress and external loads reach equilibrium, thereby obtaining a relatively accurate initial stress state without human interference.

[0013] Further, in step S100, the establishment of the overall structural analysis model includes: S106: Wind load is one of the most important external loads in the structural design of wind turbines, mainly consisting of tower load and rotor load. The wind load acting on the rotor can be calculated using this formula: , in, It is the wind load acting on the rotor; It is the air density, and its value is... ; It is the rotor radius; It refers to the wind speed at the wheel hub; It is the thrust coefficient. Depending on the operating conditions of the wind turbine, its value is usually between 0.2 and 0.8.

[0014] Further, in step S100, the establishment of the overall structural analysis model includes: When the wind speed exceeds the shutdown wind speed of the offshore wind turbine, the wind turbine stops rotating. At this time, the thrust coefficient approaches zero, and the horizontal thrust of the rotor in the shutdown state can be calculated using this equation: , in, It is the thrust acting on the rotor under extreme loads; This refers to the thrust coefficient under extreme operating conditions. Number of leaves; air density; This represents the projected area of ​​a single blade. This refers to wind speeds under extreme operating conditions.

[0015] Further, in step S100, the establishment of the overall structural analysis model includes: The wind load acting on the tower is calculated using the following formula. The tower is divided into 10 sections, and the wind load on each section is applied as a concentrated force. Through coupling constraints, the concentrated force is applied to the center of the windward side of the corresponding tower section. The expression is as follows: , in, Height of wind load application; The projected area of ​​each segment; air density; This is the tower shape factor (value 0.5). Let be a function of wind speed as a function of altitude, and its expression is as follows: , in, For wheel hub height, For the height of the tower, The power-law exponent, The wind speed at the wheel hub.

[0016] Further, in step S100, the establishment of the overall structural analysis model includes: S107: The wave load calculation for slender structures uses the Morrison equation, which assumes that the total wave force acting on the pile foundation consists of the sum of drag and inertial forces. Its expression is as follows: , in, This is the expression for inertial force; Here is the expression for the drag force; For water depth; This is the quality coefficient; This is the drag coefficient; The density of seawater; The outer diameter of a single pile foundation; For horizontal wave induced velocity; For horizontal wave-induced acceleration; The surface wave distribution function is expressed as follows: , , , in, Wave height; Wave number; The frequency of the wave; For wave period; This invention uses the power-law distribution method to calculate the flow load, and its expression is as follows: , in, The horizontal flow resistance per unit length, This represents the local flow velocity.

[0017] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S201: Create detailed models of the flange and bolts respectively; S202: Assign appropriate material properties to flanges and bolts respectively; S203: Assemble the completed models together and set their interactions; S204: Mesh the bolts and flanges separately, with the mesh type being an eight-node hexahedral linear reduced integral element.

[0018] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S205: As a vulnerable and critical component, bolts require a suitable damage model to accurately characterize their damage evolution in complex marine environments. This invention uses the GTN damage criterion to characterize the damage evolution process of bolts. The GTN damage criterion parameters for high-strength bolts are as follows: , , , , , , , , .

[0019] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S206: The stress-strain curve of a bolt can be divided into three stages: the elastic stage, the necking stage, and the post-necking stage. Numerical simulation requires the input of the material's actual stress-strain curve, which can be obtained by converting the engineering stress-strain curve. The calculation formula for the actual stress-strain curve is as follows: in, Bolt stress; It is the elastic modulus of the bolt; It is the strain of the bolt.

[0020] The true stress-strain curve during the necking stage is represented by the following equation: , , in, , For true stress and true strain; , To test the obtained engineering stress and engineering strain; The starting point of the necking stage is usually the ultimate stress point in the material's tensile test. , , , in, This is the weighting factor required for calibration, with a value between 0 and 1; other parameters must meet the continuity requirements of stress and strain. For linear strengthening modulus, ; The strain hardening index is... ; The stress intercept of the linear model. ; The strength coefficient, , ; , It represents the actual stress and strain at the material's ultimate stress point.

[0021] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S207: In establishing the equivalent model of the flange bolt connection, due to the significant differences in stress response in the three principal stress directions, the overall connection structure exhibits obvious anisotropic mechanical behavior. Although the Mises yield criterion is simple in form, its isotropic assumption makes it difficult to accurately describe the yield characteristics of such anisotropic materials. Therefore, this invention adopts the Hill anisotropic yield criterion for equivalent simulation. By introducing multi-directional yield strength parameters, this criterion can more effectively characterize the plastic deformation behavior of materials in different directions, thereby more accurately reflecting the actual mechanical response of the flange bolt connection under complex stress states. The specific expression of the Hill anisotropic yield criterion is as follows: , , , , in, The yield stress; These are the yield stress values ​​in various directions, measured along the three principal stress directions. , , It is normal stress. , , For shear stress components; F, G, and H are anisotropic yield parameters; It is the user-defined yield stress; It is the anisotropic yield stress ratio; ; The flow rule for the Hill anisotropic yield criterion is as follows: , .

[0022] in, This is the plastic strain increment vector; These are plastic multiplier vectors; is the normal vector of the yield surface in stress space; b is the plastic flow direction vector; F, G, H are the normal stress anisotropy weighting parameters, and N is the shear stress anisotropy parameter.

[0023] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S208: Based on the mechanical response characteristics of bolt and flange connection structures under external loads, this invention selects a representative volume element (RVE) as the basic unit for equivalent analysis. The RVE consists of a single bolt and its surrounding flange region. Its design fully reflects the key geometric features of bolted connections and can effectively capture local stress concentration phenomena and the anisotropic response characteristics of materials. At the same time, a representative volume element is established without considering the interaction between the flange and the bolt, that is, the bolt modeling is ignored and only the flange is established (i.e., a homogenized model).

[0024] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S209: Bolt preload is applied to the bolt by a cooling method, and its calculation formula is as follows: Bolt deformation under preload : , , Where F is the bolt preload; For bolt rod stiffness; It is the elastic modulus of the bolt; It is the length of the bolt rod; It is the cross-sectional area of ​​the bolt rod; It is the yield point of the bolt material (the yield point of a 10.9 grade bolt is 900 MPa). This is the bolt preload, which is typically [value missing]. ; Deformation of the connector under preload : , This expression can be further written as: , in, It is the coefficient of linear expansion of the bolt material. For the stiffness of the connection part, This is the initial temperature of the bolt, with a value of 0. The applied temperature is determined based on the magnitude of the bolt preload.

[0025] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S210: To clarify the influence of bolted connections on yielding behavior, this invention first compares the differences in yield points in the plane between the refined model and the model that ignores bolts, and obtains a comparison of the mechanical properties of the refined model and the model that ignores bolts. The calculation process is as follows: yield data of representative volume elements (RVE) under different stress states were obtained through numerical simulation; first, uniaxial tensile and compressive loads were applied to the three principal stress directions of the bolt-flange model and the traditional simplified model respectively to obtain their uniaxial yield strength; then, the influence of biaxial stress state on yield behavior was characterized by applying composite biaxial loads along different principal stress directions; finally, based on the above series of numerical test results, the yield surface of RVE in the principal stress plane was obtained.

[0026] Further, in step S200, the establishment of the numerical model of the flange-bolt microscopic multi-scale includes: S211: Substitute the material parameters calculated by the refined model in the three principal stress directions into the homogenized model based on the Hill criterion.

[0027] Furthermore, the homogenized equivalent model established in step S200 is substituted into the overall structure, and the steps for analyzing the dynamic response of the overall structure under various loads such as wind, waves, and earthquakes include: S301: Substitute the established homogeneous flange-bolt connection model into the overall structure; S302: First, perform ground stress balance, then apply gravity load to the entire model; S303: Apply wave load to the jacket foundation, wind load to the tower and rotor, and then apply seismic load to the overall model; S304: Analyze the dynamic response of the overall model under wind, wave, and seismic loads.

[0028] Furthermore, based on the results of the overall structural calculations in S300, the flange-bolt connection structure is evaluated using sub-modeling techniques, including: S401: Establish a detailed model of the flange and bolts; S402: Assign material parameters to the flange and bolt respectively; S403: Assemble the flange and bolt models together, mesh them, and set the corresponding contact relationships; S404: Extract the stress state at the flange-bolt connection in the overall model under wind and wave loads, substitute the extracted stress state at the flange-bolt connection into the sub-model, and analyze the established refined flange-bolt model.

[0029] An integrated analysis system for multi-scale damage of offshore wind turbine structures, applying the integrated analysis method for multi-scale damage of offshore wind turbine structures, includes: The data acquisition and preprocessing module collects multi-field load data and structural geometry and material parameters of offshore wind power structures, and completes filtering, noise reduction and standardization preprocessing. The multi-scale model building module establishes an overall structural model of the offshore wind turbine, constructs a flange-bolt micro-multi-scale numerical model based on the GTN damage model, establishes a homogenized flange model through RVE homogenization and Hill anisotropic yield criterion, and builds a refined flange-bolt sub-model. The multi-field coupling analysis module embeds a homogenized equivalent model into the overall structure to analyze the overall dynamic response under multiple loads such as wind, waves, and earthquakes. It imports the flange-bolt stress state of the overall structure through sub-model technology to achieve multi-scale coupling calculation. The damage co-assessment module combines the overall structural dynamic response with the analysis results of local sub-models to characterize the stress distribution, plastic development, and damage evolution of the flange-bolted connection, and completes the co-assessment of global response and local damage. The results output and optimization module outputs structural damage assessment reports and remaining life prediction results. It optimizes the accuracy of the model through parameterized calibration and simultaneously provides graded early warning and operation and maintenance decision suggestions.

[0030] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: 1. The method of this invention, by establishing an anisotropic equivalent constitutive model, transforms the hundreds of thousands of solid elements and contact elements in the traditional flange-bolt region into a small number of continuous medium elements, significantly reducing the degrees of freedom of the overall finite element model. Under the same computational accuracy conditions, the computation time for overall structural analysis can be reduced by approximately 60% to 90%, and the dependence on high-performance computing platforms can be reduced, transforming the overall dynamic analysis of large wind turbine units from "difficult to carry out" to "efficient to carry out".

[0031] 2. The method of this invention proposes an equivalent flange model that can accurately reproduce the nonlinear mechanical behavior of actual bolted connection structures, ensuring high accuracy and reliability in engineering calculations. The equivalent model is established based on the RVE homogenization theory and the Hill anisotropic yield criterion, realistically simulating the yielding mechanism, plastic development law, macroscopic stiffness performance, and complex dynamic response of flange-bolted connections under different principal stress directions. Comparison with the refined bolt model shows that the errors in key response quantities are generally less than 5%, including stress distribution, natural frequencies, mode shapes, displacement, and acceleration, significantly outperforming existing empirical simplified models.

[0032] 3. The method of this invention, after obtaining the dynamic response of the overall structure under wind, wave, current, earthquake, or extreme conditions, uses a sub-model to transfer boundary displacements to a local fine bolt zone model, achieving precise prediction of stress concentration, plastic zone development, damage accumulation, and potential failure modes. This cross-scale analysis framework can simultaneously evaluate global structural performance and local connection safety, providing high-precision criteria for the reliability of offshore wind turbines under abnormal waves, extreme wind fields, seismic coupling effects, and service life fatigue conditions.

[0033] 4. The method of this invention can simultaneously acquire the overall dynamic response of the wind turbine and the damage evolution of the bolted connection area, realizing a unified analytical framework for global response and local failure. By using a homogenized flange equivalent model for overall structural analysis and combining it with local sub-model technology, this invention achieves the collaborative prediction of the dynamic characteristics of the overall structure and the stress concentration, plastic development, damage accumulation, and potential failure modes of the bolted area within the same computational system. This cross-scale coupling capability overcomes the technical bottlenecks of traditional methods where the overall model cannot reflect local damage and the local model cannot obtain real boundary conditions. It enables a unified description of the entire process of the structure from overall stress behavior to local failure mechanisms, thus providing more complete and reliable technical support for the safety assessment of offshore wind turbines under extreme wind, wave, current, and seismic coupled environments.

[0034] 5. The system of this invention features a data acquisition and preprocessing module that ensures the accuracy and standardization of multi-load, structural geometry, and material parameters, laying a reliable data foundation for subsequent analysis. The multi-scale model construction module, utilizing the GTN damage model, RVE homogenization, and Hill anisotropic yield criterion, balances the efficiency of overall structural modeling with the refinement of the flange-bolt local model, significantly reducing the computational cost of traditional fine modeling. The multi-field coupling analysis module can accurately simulate the overall dynamic response under multiple loads such as wind, waves, and earthquakes, achieving multi-scale coupling calculations of global and local dimensions through sub-model technology, overcoming the limitations of single-scale analysis. The damage collaborative assessment module integrates overall dynamic response and local damage evolution data, accurately characterizing the stress distribution, plastic development, and damage state of the flange-bolt connection, achieving collaborative judgment of global response and local damage. The result output and optimization module not only outputs damage assessment reports and remaining life prediction results but also continuously optimizes model accuracy through parametric calibration, simultaneously providing graded early warnings and operation and maintenance decision-making suggestions, providing comprehensive and practical technical support for the safe operation and risk management of offshore wind power structures. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating the integrated analysis method for multi-scale damage of offshore wind power structures according to an embodiment of the present invention. Figure 2This is a schematic diagram illustrating the integrated analysis method for multi-scale damage of offshore wind power structures according to an embodiment of the present invention. Figure 3 This is a numerical model diagram of an offshore wind turbine in an embodiment of the present invention; Figure 4 This is a detailed view of the flange and bolts in an embodiment of the present invention; Figure 5 This is a constitutive curve diagram of a bolt in an embodiment of the present invention; Figure 6 This is a diagram showing the balance between bolt preload and ground stress in an embodiment of the present invention. Figure 7 Cell diagrams of the refined model, traditional model, and homogenized model in embodiments of the present invention; Figure 8 The stress distribution diagrams for the refined model and the homogenized model in this embodiment of the invention are shown. Figure 9 These are stress-strain curves of the refined model and the homogenized model in the embodiments of the present invention; Figure 10 This is a schematic diagram of the tower deflection when the overall structure is subjected to wind and wave loads in an embodiment of the present invention; Figure 11 A schematic diagram of the foundation deflection when the overall structure is subjected to wind and wave loads in an embodiment of the present invention; Figure 12 The stress distribution and damage diagrams of the flange-bolt in the refined model, traditional model, and homogenized model of the overall structure under wind and wave loads in the embodiments of the present invention are shown. Figure 13 This is a comparison diagram of the refined model, the traditional model, and the homogeneous model of the flange-bolt in the embodiments of the present invention; Figure 14 This is a numerical model diagram of a multi-scale offshore wind turbine in an embodiment of the present invention; Figure 15 This is a load loading direction diagram of a multi-scale numerical model of an offshore wind turbine in an embodiment of the present invention; Figure 16 This is a schematic diagram of the time-varying displacement curve of the tower top of a multi-scale numerical model of an offshore wind turbine in an embodiment of the present invention; Figure 17 This is a schematic diagram of the time-varying displacement curve of the jacket foundation of a multi-scale numerical model of an offshore wind turbine in an embodiment of the present invention; Figure 18 This is a schematic diagram of the time-varying acceleration curve at the top of the tower of a multi-scale numerical model of an offshore wind turbine in an embodiment of the present invention. Detailed Implementation

[0036] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0037] Example 1 like Figure 2 As shown, this invention proposes a novel and efficient computational method employing a two-stage analysis approach. The first stage establishes a homogenized equivalent model through parametric calibration, replacing the traditional separate modeling method that requires detailed modeling of bolts and flanges. This equivalent model accurately reflects the stiffness, strength, and dynamic response characteristics of the connection nodes at the macroscopic scale, thereby significantly reducing the computational scale of the overall system analysis. The second stage, based on the stress analysis results of the overall structural model, employs sub-modeling technology to perform detailed local modeling of the critical flange-bolt region. Specifically, the displacement response of the overall model at the boundary of this region is used as the driving boundary condition for the sub-model, thereby reconstructing its accurate stress state at a fine scale. In the detailed modeled sub-model, the GTN damage model is introduced to simulate the microscopic damage evolution process of the bolt material under load. Through a bidirectional data transmission mechanism between the macro-model and the local model, the computational accuracy and efficiency of the complex model under multi-field coupled loads are collaboratively optimized. Finally, based on the established parametric equivalent model, the overall response characteristics of the wind turbine generator under seismic and wave loads are analyzed.

[0038] Example 2 like Figure 1 As shown, the integrated analysis method for multi-scale damage of offshore wind power structures includes the following steps: Step S100: Establish an overall structural analysis model for the offshore wind turbine. This model is based on an 8.5 MW offshore wind turbine in the South China Sea off Fangchenggang, Guangxi, and includes the tower, flange connection, three-legged frame, rock-embedded foundation, and rock mass. Figure 3As shown. The wind turbine has a diameter of 227 meters. The hub center height is 133 meters, the tower height is 111 meters, the tower base diameter is 7.5 meters, and the tower top diameter is 5.3 meters. The upper nacelle foundation adopts a space truss structure, consisting of three steel pipe piles with a diameter of 2.2 meters and a wall thickness of 40 millimeters, evenly distributed in an equilateral triangle. The top dimensions of the nacelle are 17 meters × 17 meters × 17 meters, and the bottom dimensions are 23 meters × 23 meters × 23 meters. The lower steel pipe piles are embedded in the rock mass, with a length of 27 meters and a wall thickness of 40 millimeters. The upper nacelle legs are connected to the steel pipe piles via grouting connections, with a grouting connection section thickness of 250 millimeters. The main support leg steel pipes are interconnected via cross-bracing steel pipes with a diameter of 0.5 meters. For simplified calculations, components such as the rotor and nacelle are represented by mass points of equal mass. The total mass of the blades and hub is 215 tons, and the nacelle mass is 250 tons. Detailed drawings of the flanges and bolts are shown below. Figure 4 As shown.

[0039] The seafloor structure in the rock mass model can be divided into eight layers: the uppermost layer is silty soil, followed by clastic rock, slate, sedimentary tuff, sandstone, siltstone, limestone, and mudstone. The silty soil layer is omitted from the modeling because its physical properties differ significantly from the rock mass.

[0040] The wind turbine tower, nacelle foundation, pipe piles, flanges, and other components are all made of Q345 grade steel. The composite hardening model in ABAQUS is used for simulation. This model can simultaneously consider the nonlinear isotropic hardening and kinematic hardening effects of materials, accurately simulating the Bauschinger effect of metals while effectively presenting their mean stress relaxation characteristics.

[0041] The simulation of the grouting material was performed using the concrete plastic damage (CDP) model built into the ABAQUS software. This model captures the stiffness decay characteristics of concrete through the plastic damage factor. The compressive constitutive model of the grouting material was calculated using the Minder model. (1) (2) (3) in, The concrete compressive strength is 130 MPa. This is the peak strain of the concrete, which is 0.003. It is the elastic modulus of concrete, which is 40 GPa. The concrete secant stiffness is [value missing].

[0042] The constitutive model under tension is shown below: (4) (5) (6) (7) in, It is the tensile strength of concrete, which is 5 MPa; It is the peak strain of the concrete; The value is 5.

[0043] The damage factor of the grouting material is calculated as follows: (8) (9) in, , This represents the peak stress of concrete under tensile and compressive conditions. , The values ​​are 2 and 1 respectively.

[0044] In this invention, since the pipe pile has a circular cross-section, the rock mass is designated as the main surface to prevent the primary and secondary surfaces from penetrating each other. Simultaneously, the normal behavior of the contact surfaces is set to hard contact, meaning the two surfaces are allowed to separate but not penetrate each other, and the tangential direction uses the Coulomb friction model. The contact between the bolts and flanges is set to hard contact and tangential friction contact, with a friction coefficient of 0.3.

[0045] Before applying structural gravity loads, a static equilibrium analysis of the rock mass is performed. The specific process includes: first, establishing a finite element model of the rock mass and setting material parameters; then, applying gravity loads and calculating the initial stress field under gravity, combined with boundary conditions reflecting the actual engineering scenario; extracting the obtained stress field as the initial condition for the main analysis based on the principle of initial stress equilibrium and importing it into the subsequent numerical model; finally, reapplying gravity loads to ensure that the internal stresses and external loads reach equilibrium, thereby obtaining a relatively accurate initial stress state without human interference.

[0046] The wind load on the rotor can be calculated using equation (10): (10), in, It is the wind load acting on the rotor; It is the air density, and its value is... ; It is the rotor radius; It refers to the wind speed at the wheel hub; It is the thrust coefficient. Depending on the operating conditions of the wind turbine, its value is usually between 0.2 and 0.8.

[0047] When the wind speed exceeds the shutdown wind speed of the offshore wind turbine, the wind turbine stops rotating. At this time, the thrust coefficient approaches zero, and equation (10) no longer applies. The horizontal thrust of the rotor in the shutdown state can be calculated by equation (11): (11) in, The thrust coefficient under extreme conditions is 1.6; Number of leaves; This represents the projected area of ​​a single blade. This refers to wind speeds under extreme operating conditions.

[0048] The wind load acting on the tower is calculated according to formula (12). The tower is divided into 10 sections, and the wind load on each section is applied as a concentrated force. The concentrated force is applied to the center of the windward side of the corresponding tower section through coupling constraints, and its expression is as follows: (12) in, Height of wind load application; The projected area of ​​each segment; This is the tower shape factor (value 0.5). Let be a function of wind speed as a function of altitude, and its expression is as follows: (13) in, For wheel hub height, For the height of the tower, It is 0.2. The wind speed at the wheel hub.

[0049] The wave load calculation for slender structures uses the Morrison equation. This equation assumes that the total wave force acting on the pile foundation consists of the sum of drag and inertial forces, and its expression is as follows: (14) in, This is the expression for inertial force; Here is the expression for the drag force; For water depth; The quality coefficient (the value is 2 in this invention); The drag coefficient (the value in this invention is 1.2); The density of seawater; The outer diameter of a single pile foundation; For horizontal wave induced velocity; For horizontal wave-induced acceleration; Let be the surface wave distribution function, and its expression is as follows: (15) (16) (17) in, Wave height; Wave number; The frequency of the wave; It is the wave period.

[0050] This invention uses the power-law distribution method to calculate the flow load, and its expression is as follows: (18) in, The horizontal flow resistance per unit length, This represents the local flow velocity.

[0051] Step S200: Establish detailed models for the flange and bolts respectively, select representative cells, and assign them material properties. The flange uses Q345 steel, and the bolts are 10.9 grade high-strength bolts, assigned a GTN damage model. The stress-strain curve of the bolts can be divided into three stages: elastic stage, necking stage, and post-necking stage. Numerical simulation requires the input of the material's true stress-strain curve, which can be obtained by converting the engineering stress-strain curve. The calculation formula for the true stress-strain curve is as follows: (19) The true stress-strain curve of the necking stage is represented by formulas (2) and (3): where, , For true stress and true strain; , The engineering stress and strain obtained from the test.

[0052] (20) (twenty one) The starting point of the necking stage is usually the ultimate stress point in the material's tensile test. (twenty two) (twenty three) (twenty four) in, This is the weighting factor required for calibration, with a value between 0 and 1; other parameters must meet the continuity requirements of stress and strain. , , . . , These are the actual stress and strain at the material's ultimate stress point. The constitutive curve of the bolt is shown below. Figure 5 As shown. The bolts of the representative cell are assembled with the flange to establish interaction, and a bolt preload is applied. The effect of the applied bolt preload is as follows. Figure 6 As shown.

[0053] Based on the principle of continuous media, a homogenized flange model is established according to the overall stiffness and stress characteristics of the selected flange-bolt connection structure. Due to the anisotropy of the refined flange-bolt model under the three principal stress directions, the Hill yield criterion is selected as the yield criterion for the homogenized flange model. The Hill yield criterion can effectively reflect the anisotropic properties of metallic materials in different directions.

[0054] The specific expression for Hill's criterion is as follows: (25) (26) (27) (28) in, It is the yield stress value, which is measured in the three principal stress directions; It is the user-defined yield stress; It is the anisotropic yield stress ratio; .

[0055] The flow rule for the Hill anisotropic yield criterion is as follows: (29) (30) Bolt deformation under preload : (31) (32) Where F is the bolt preload; For bolt rod stiffness; It is the elastic modulus of the bolt; It is the length of the bolt rod; It is the cross-sectional area of ​​the bolt rod; It is the yield point of the bolt material (the yield point of a 10.9 grade bolt is 900 MPa). This is the bolt preload, which is typically [value missing]. ; Deformation of the connector under preload : (33) This expression can be further written as: (34) in, It is the coefficient of linear expansion of the bolt material. For the stiffness of the connection part, This is the initial temperature of the bolt, with a value of 0. The applied temperature is determined based on the magnitude of the bolt preload.

[0056] Figure 7 This invention presents cell diagrams of the refined model (FB model), homogenized model (PH model), and traditional model (TS model) in embodiments of the present invention. The invention first compares the differences in in-plane yield points between the refined model and the model ignoring bolts. A comparison of the mechanical properties of the refined model and the model ignoring bolts is obtained. Yield data of representative volume elements (RVEs) under different stress states are obtained through numerical simulation. First, uniaxial tensile and compressive loads are applied to the three principal stress directions of the bolt-flange model and the traditional simplified model, respectively, to obtain their uniaxial yield strength. Then, by applying combined biaxial loads along different principal stress directions, the influence of biaxial stress states on yield behavior is characterized. Finally, based on the above series of numerical test results, the yield surface of the RVE in the principal stress plane is obtained. The material parameters calculated by the refined model in the three principal stress directions are then substituted into the homogenized model based on the Hill criterion. This yields the homogenized flange-bolt equivalent model established based on the Hill anisotropic yield criterion. Figure 8 This is a stress distribution diagram of the refined model and the homogenized model in an embodiment of the present invention. For example... Figure 8 As shown, the mechanical responses of the PH model and the FB model were compared and analyzed along the three principal stress directions. Figure 8 (a) shows the simulation results of FBmodel in the direction. The necking phenomenon is obvious in the middle of the bolt rod, indicating that the region has entered the plastic deformation stage. Figure 8 (b) The simulation results of the PH model in the ... Figure 8 In (c)-(e), the deformation modes and stress distributions of the PH model and the FB model are highly consistent.

[0057] Figure 9 The stress-strain curves of the refined model and the homogenized model in the embodiments of the present invention are shown. Figure 9The stress-strain curves of the PH model and the FB model in the three principal stress directions are compared. It can be seen that the curve trends of the two models are basically consistent in both the elastic and plastic stages, with the curves of the PH model and the FB model almost completely overlapping in the elastic stage. In the plastic stage, the hardening trend of the PH model is also basically consistent with that of the FB model. The results show that the equivalent model based on the Hill criterion can accurately predict the elastoplastic response of the original structure in different directions, verifying the effectiveness and reliability of the model in simulating the nonlinear mechanical behavior of structures.

[0058] Step S300: After obtaining the homogeneous flange-bolt connection equivalent constitutive model based on the HILL anisotropic yield criterion in step S200, the equivalent model is embedded into the overall finite element structural model of the offshore wind turbine to replace the traditional fine bolt-flange modeling method, thereby constructing a unified overall structural analysis framework.

[0059] Before conducting dynamic response analysis, initial stress equilibrium calculations were first performed to ensure that the seabed rock mass and the rock-embedded foundation reached a stable equilibrium state under gravity, thus eliminating the interference of the initial stress field on the subsequent dynamic response. After obtaining the stable initial field, a gravity load was applied to the overall structure to obtain the static response of the wind turbine under its own weight. Subsequently, wave loads were applied to the jacket foundation, using the wave pressure field calculated by the Morison equation. Simultaneously, random wind loads or equivalent wind pressures were applied to the tower, nacelle, and rotor to simulate the combined effects of wind, waves, and currents in the actual offshore operating environment.

[0060] Figure 10 This invention relates to the deflection of the tower when the overall structure is subjected to wind and wave loads. The invention will compare the horizontal displacement response of the FB model, PH model, and TS model under rated wind speed. Figure 10 The horizontal displacement of the tower under three different load directions is shown: Figure 10 (a) Under the load along the positive y-axis, one pile leg is under compression and the other two are under tension, and the maximum displacement of the tower top reaches 1016 mm; Figure 10 (b) Under the condition of load along the negative y-axis, with two pile legs under compression and one under tension, the maximum displacement of the tower top is 996 mm; Figure 10(c) Under x-axis loading, one pile leg is under compression, another under tension, and the third maintains a near-neutral axial state. The maximum displacement at the top of the tower is 1006 mm. It is noteworthy that the x-axis loading condition results in the largest displacement at the top of the tower, indicating that this is the most unfavorable stress state for the structure and requires attention and appropriate protective measures in structural design. The results show that the top of the tower is the most sensitive part of the structural system to horizontal displacement, and its displacement is also the most significant. Under a standard wind speed of 11.4 m / s, the horizontal displacement results of the PH, FB, and TS models in the three different loading directions are highly consistent and far below the safety limits set by the code. This not only verifies the accuracy of the established parametric homogenization (PH) model in overall stiffness simulation but also demonstrates its applicability and reliability in actual engineering bearing capacity assessment.

[0061] Figure 11 In this embodiment of the invention, the foundation deflection when the overall structure is subjected to wind and wave loads is referenced. Figure 10 The pile foundation is embedded in the seabed soil layer, with the pile tops being free ends. Therefore, when subjected to horizontal loads, the foundation will undergo overall lateral displacement deformation. Since the displacement responses of the three numerical models (PH model, TS model, and FB model) under horizontal loads are highly consistent, due to space limitations, this invention only selects the PH model as a representative result. Figure 11 The horizontal displacement of the pile body under three different load directions is shown using the FB model. Since the displacement response of the three pile legs is basically the same, and the horizontal displacement of the pile body is small, with the maximum displacement at the pile top being only 1.7 mm, the average displacement value of the three pile legs is used in this paper. Figure 11 (a) Shows the load along the positive y-axis. Figure 11 (b) Show the load along the negative y-axis. Figure 11 (c) Show the load along the x-axis.

[0062] After the solution is completed, the post-processing module extracts key calculation data such as the displacement of the tower top, the stress of key sections, the peak acceleration, the dynamic bending moment, and the stress evolution process of the flange connection area, providing accurate input conditions for subsequent local sub-model damage analysis and structural safety evaluation.

[0063] Step S400: After obtaining the dynamic response of the overall structure in step S300, in order to accurately describe the local stress concentration and damage evolution behavior in the flange-bolt connection area, a refined local sub-model is further established and connected with the solution results of the overall model, thereby realizing cross-scale collaborative analysis between the overall and local models.

[0064] Based on the actual engineering structure dimensions, a high-precision solid model including the geometric features of bolts, bolt holes, and flange transition zones is constructed to ensure the accuracy of the local stress field. According to the actual material selection in the project, separate material parameters are applied to the bolts and flange components. A Q345 metal constitutive model is applied to the flange, and 10.9 grade high-strength bolts are used. The GTN damage model is employed to simulate damage initiation, propagation, and failure processes. The three-dimensional solid models of the bolts and flange are assembled according to the actual assembly relationship, and the contact relationship between the bolts and flange is defined. The contact form can adopt surface-to-surface contact, friction coefficient model, or friction slip model to simulate the stress state of the actual connection interface. Based on the magnitude of the local stress gradient, local fine meshing is performed on key parts such as the bolt shank and flange hole wall to improve the accuracy of stress and damage prediction.

[0065] From the overall structural dynamic analysis in step S300, nodal displacement, interface stress, or strain data of the flange-bolted connection region are extracted and applied as boundary conditions to the corresponding locations of the local refined sub-model. By applying full-field displacement or mixed loads, the actual stress state of the local model under the overall environment is reconstructed. Subsequently, nonlinear static or dynamic analysis is performed on the local model to calculate stress concentration, plastic development, damage parameter evolution, and potential failure modes in the bolt and hole wall regions, thereby obtaining the local damage response of the flange-bolted connection under complex environmental conditions.

[0066] Figure 12 This invention presents stress distribution and damage diagrams of flange-bolt components in a refined model, a traditional model, and a homogenized model of the overall structure subjected to typhoon and wave loads in an embodiment of the invention. To study the evolution of bolt damage under typhoon loads, this invention systematically compares the differences in bolt damage evolution processes using the FB model, PH model, and TS model. The typhoon wind speed is set to 50 m / s. For the FB model, the bolt damage evolution process can be obtained by directly applying the typhoon load to the overall structure. This model can directly capture the stress state and plastic evolution behavior of the bolts. For the PH model and TS model, a two-stage analysis method is used: first, load analysis is performed at the overall structural level, and then sub-model technology is used to transfer the stress state of key parts to a locally refined model containing only flange-bolt connections. Figure 12This study demonstrates the stress distribution of bolts under typhoon loads. The final bolt stress consists of two parts: the initial preload and the additional stress caused by the external load. The maximum bolt stress values ​​calculated by each model are: FB model 995 MPa, PH model 997 MPa, and TS model 943 MPa. The comparison shows that the maximum stress deviation between the PH model and the FB model is only 0.2%, proving that the parametric homogenization method and the refined model have equivalent accuracy. This result verifies the reliability of the PH model in predicting bolt stress levels. However, the TS model does not consider the bolt's contribution to the connection stiffness, and its predicted stress value is 5.4% lower than the baseline model. This may lead to an inability to accurately assess the actual stress state of the bolts during the design process, thereby masking the potential failure risk of the connection system and posing a threat to structural safety. Figure 12 The damage evolution process of bolts under typhoon load is illustrated. VVF is a parameter representing the volume ratio of micropores or defects in a material; a higher VVF value indicates more porosity or damage within the material. The results from the FB model are largely consistent with those from the PH model, while the TS model underestimates the extent of bolt damage evolution under typhoon load.

[0067] Figure 13 This is a comparison diagram of the refined model, traditional model, and homogeneous model of the flange-bolt in this embodiment of the invention. In terms of modeling complexity, the FB model requires establishing the geometric details of the bolts and flanges. To ensure the accuracy of the calculation results, the bolt and flange areas are usually refined through mesh processing (e.g., ...). Figure 13 (a) shows the complex interactions. In the dynamic analysis of the overall structure, the finite element software needs to iteratively calculate and determine the convergence of these complex contact states in each incremental step. At the same time, the large number of meshes significantly increases computational resource consumption. For example... Figure 13 As shown in (b), the TS model simplifies the modeling of bolts and flanges, ignoring the interaction mechanism between them and their respective material properties. While this simplified analysis method shortens computation time, it cannot accurately reflect the stress distribution and damage evolution process at the flange contact surface. This efficiency improvement at the expense of key mechanical properties may lead to biases in structural safety assessments. The PH model achieves the best balance between computational accuracy and efficiency. Figure 13 As shown in (c), this model accurately simulates the nonlinear mechanical behavior of the bolt-flange connection system by introducing precise calibration parameters and selecting an appropriate yield criterion. Meanwhile, its computational efficiency is comparable to that of the TS model.

[0068] Step S500: When analyzing the impact of earthquakes on the overall structure, seismic motion time history input can be applied to the nodes at the bottom of the rock mass, and time history coupling can be achieved with wind and wave loads, thereby constructing a dynamic calculation case under the combined action of multiple fields of wind, waves, current, and earthquake. The overall structural model adopts implicit or explicit dynamic solution methods to calculate the response characteristics such as dynamic displacement, acceleration, stress field, natural frequency, and dynamic amplification effect of the tower-flange-frame-foundation system. This invention uses the PH model to analyze the dynamic response of the overall structure under seismic loads. Figure 14 This is a numerical model diagram of a multi-scale offshore wind turbine in an embodiment of the present invention. Figure 15 The loading direction is shown in the multi-scale numerical model of offshore wind turbines in this embodiment of the invention.

[0069] Figure 16 The displacement time history curves of the tower top along the x-axis and y-axis are shown. Under the coupled effects of multiple hazards such as wind, waves, and earthquakes, the maximum displacement of the tower top in the x-axis direction reached 810 mm, and the maximum displacement in the y-axis direction was 783 mm. Figure 17 This demonstrates the horizontal displacement of an offshore platform structure at a height of 20 meters above the seabed. Under load in the x-direction, the maximum horizontal displacement at this location is 34 mm, while under load in the y-direction, the maximum horizontal displacement is 11 mm, three times that in the x-direction. Figure 18 As shown, the comparative analysis of acceleration response reveals significant directional differences: under load in the x-axis direction, the maximum acceleration of the structure reaches 6.4 m / s², while under load in the y-axis direction, it drops to 3.1 m / s². This indicates that the homogenized flange-bolt connection model (PH model) constructed based on the Hill yield criterion of this invention can accurately reflect the stress and deformation characteristics of the overall structure under seismic loads, demonstrating good applicability and reliability.

[0070] This invention, based on the principles of continuum mechanics, develops an equivalent anisotropic elastoplastic constitutive model for flange-bolt connections using a representative volume element (RVE) homogenization method. This method replaces the traditional approach of modeling bolts and flanges separately with a parametrically calibrated homogenized flange model. This equivalent model can efficiently predict the nonlinear response of the connection in the overall structural analysis, effectively solving the flange-bolt modeling problem and significantly improving overall computational efficiency. A two-stage analysis method is employed. The first stage establishes a homogenized equivalent model through parametric calibration, replacing the traditional separate modeling method of finely modeling bolts and flanges. This equivalent model accurately reflects the stiffness, strength, and dynamic response characteristics of the connection node at the macroscopic scale, thereby significantly reducing the computational scale of the overall system analysis. The second stage, based on the stress analysis results of the overall structural model, uses sub-modeling technology to perform local fine modeling of the critical flange-bolt region. Specifically, the displacement response of the overall model at the boundary of this region is used as the driving boundary condition of the sub-model, thereby reconstructing its accurate stress state at a fine scale. In the refined modeling sub-model, the GTN damage model is introduced to simulate the microscopic damage evolution process of bolt materials under load. Through a bidirectional data transmission mechanism between the macro and local models, the computational accuracy and efficiency of the complex model under multi-field coupled loads are collaboratively optimized. Finally, based on the established parametric equivalent model, the overall response characteristics of the wind turbine generator under seismic and wave loads are analyzed.

[0071] An integrated analysis system for multi-scale damage to offshore wind turbine structures includes: The data acquisition and preprocessing module collects multi-field load data and structural geometry and material parameters of offshore wind power structures, and completes filtering, noise reduction and standardization preprocessing. The multi-scale model building module establishes an overall structural model of the offshore wind turbine, constructs a flange-bolt micro-multi-scale numerical model based on the GTN damage model, establishes a homogenized flange model through RVE homogenization and Hill anisotropic yield criterion, and builds a refined flange-bolt sub-model. The multi-field coupling analysis module embeds a homogenized equivalent model into the overall structure to analyze the overall dynamic response under multiple loads such as wind, waves, and earthquakes. It imports the flange-bolt stress state of the overall structure through sub-model technology to achieve multi-scale coupling calculation. The damage co-assessment module combines the overall structural dynamic response with the analysis results of local sub-models to characterize the stress distribution, plastic development, and damage evolution of the flange-bolted connection, and completes the co-assessment of global response and local damage. The results output and optimization module outputs structural damage assessment reports and remaining life prediction results. It optimizes the accuracy of the model through parameterized calibration and simultaneously provides graded early warning and operation and maintenance decision suggestions.

[0072] The system of this invention features a data acquisition and preprocessing module that ensures the accuracy and standardization of multi-load, structural geometry, and material parameters, laying a reliable data foundation for subsequent analysis. The multi-scale model construction module, utilizing the GTN damage model, RVE homogenization, and Hill anisotropic yield criterion, balances the efficiency of overall structural modeling with the refinement of the flange-bolt local model, significantly reducing the computational cost of traditional fine modeling. The multi-field coupling analysis module can accurately simulate the overall dynamic response under multiple loads such as wind, waves, and earthquakes, achieving multi-scale coupling calculations of global and local dimensions through sub-model technology, overcoming the limitations of single-scale analysis. The damage collaborative assessment module integrates overall dynamic response and local damage evolution data, accurately characterizing the stress distribution, plastic development, and damage state of the flange-bolt connection, achieving collaborative judgment of global response and local damage. The result output and optimization module not only outputs damage assessment reports and remaining life prediction results but also continuously optimizes model accuracy through parametric calibration, simultaneously providing graded early warnings and operation and maintenance decision-making suggestions, offering comprehensive and practical technical support for the safe operation and risk management of offshore wind power structures.

[0073] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An integrated analysis method for multi-scale damage of offshore wind power structures, characterized in that, include: S100: Establish an overall structural analysis model for offshore wind turbines; S200: A multi-scale numerical model of flange-bolt microstructure is established based on the GTN damage model. Based on the principle of continuous medium, a homogenized flange model is established according to the overall stiffness and stress characteristics of the flange-bolt connection structure. S300: Substitute the homogenized equivalent model established in step S200 into the overall structure and analyze the dynamic response of the overall structure under various loads such as wind, waves, and earthquakes. S400: The sub-model technique is used to evaluate the flange-bolt connection structure, establish a refined flange-bolt connection structure, and substitute the flange-bolt stress state in the overall structure calculation results obtained in step S300 into the sub-model to achieve collaborative analysis of global response and local damage.

2. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 1, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S101: Establish a corresponding numerical model based on the actual size of the offshore wind turbine, and adopt a three-pile rock-embedded jacket foundation for the wind farm foundation; S102: The offshore wind turbine consists of a rotor, nacelle, tower and support foundation.

3. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 2, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S103: The simulation of the grouting material uses the concrete plastic damage model built into the ABAQUS software. This model captures the stiffness decay characteristics of concrete through the plastic damage factor. The compressive constitutive model of the grouting material adopts the Minder model. ， ， ， in, It refers to the compressive strength of concrete. is the compressive strength of the concrete cylinder; x is the strain ratio; r is the curvature parameter. It is concrete strain. It is the peak strain of the concrete; It is the elastic modulus of concrete; E sec Yes, it refers to the stiffness of the concrete cleavage. The constitutive model under tension is shown below: ， ， ， ， in, It is the tensile damage factor; For stress; It is the elastic modulus of concrete; x is the concrete strain; x is the strain ratio. It refers to the tensile strength of concrete; It is the peak strain of the concrete; The strength-stiffness ratio parameter; This represents the softening rate.

4. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 3, characterized in that, In step S100, the establishment of the overall structural analysis model includes: The damage factor of the grouting material is calculated as follows: ， ， in, It is a pressure-induced damage factor; It is the tensile damage factor; , For the tensile and compressive stresses of concrete; , This represents the peak stress of concrete under tensile and compressive conditions. , Equivalent stiffness adjustment coefficients for compression and tension, respectively.

5. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 4, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S104: In this invention, since the pipe pile has a circular cross-section, in order to prevent the main surface and the subordinate surface from penetrating each other, the rock body is set as the main surface. At the same time, the normal behavior of the contact surface is set as hard contact, that is, the two surfaces are allowed to separate but not to penetrate each other. The tangential direction adopts the Coulomb friction model. The bolt and the flange are set as hard contact and tangential friction contact. The friction coefficient is usually between 0.2 and 0.

6.

6. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 5, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S105: Initial load refers to the inherent load that the structure bears before it is put into operation and subjected to the action of the external environment. It mainly includes the self-weight of the entire structure (static load) and the bolt preload applied after installation. The self-weight of the structure consists of the sum of the static forces of all components such as the tower, monopile foundation, blades, nacelle, hub and connecting bolts. Before applying gravity loads to the structure, a static equilibrium analysis of the rock mass must be performed. The specific process includes: first, establishing a finite element model of the rock mass and setting material parameters; then, applying gravity loads and calculating the initial stress field under gravity by combining boundary conditions that reflect the actual engineering scenario; extracting the obtained stress field as the initial condition for the main analysis based on the principle of initial stress equilibrium and importing it into the subsequent numerical model; finally, reapplying gravity loads to ensure that the internal stress and external loads are in equilibrium, thereby obtaining a relatively accurate initial stress state without human interference.

7. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 6, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S106: Wind load is one of the most important external loads in the structural design of wind turbines, mainly consisting of tower load and rotor load. The wind load acting on the rotor can be calculated using this formula: ， in, It is the wind load acting on the rotor; It is the air density, and its value is... ; It is the rotor radius; It refers to the wind speed at the wheel hub; It is the thrust coefficient; Depending on the operating status of the wind turbine, its value is usually between 0.2 and 0.

8.

8. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 7, characterized in that, In step S100, the establishment of the overall structural analysis model includes: When the wind speed exceeds the shutdown wind speed of the offshore wind turbine, the wind turbine stops rotating. At this time, the thrust coefficient approaches zero, and the horizontal thrust of the rotor in the shutdown state can be calculated using this equation: ， in, It is the thrust acting on the rotor under extreme loads; The thrust coefficient under extreme operating conditions; Number of leaves; air density; This represents the projected area of ​​a single blade. This refers to wind speeds under extreme operating conditions.

9. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 8, characterized in that, In step S100, the establishment of the overall structural analysis model includes: The wind load acting on the tower is calculated using the following formula. The tower is divided into 10 sections, and the wind load on each section is applied as a concentrated force. Through coupling constraints, the concentrated force is applied to the center of the windward side of the corresponding tower section. The expression is as follows: ， in, Height of wind load application; The projected area of ​​each segment; air density; This is the tower shape factor (value 0.5). Let be a function of wind speed as a function of altitude, and its expression is as follows: ， in, For wheel hub height, For the height of the tower, The power-law exponent. The wind speed at the wheel hub.

10. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 9, characterized in that, In step S100, the establishment of the overall structural analysis model includes: S107: The wave load calculation for slender structures uses the Morrison equation, which assumes that the total wave force acting on the pile foundation consists of the sum of drag and inertial forces. Its expression is as follows: ， in, This is the expression for inertial force; Here is the expression for the drag force; For water depth; This is the quality coefficient; This is the drag coefficient; The density of seawater; The outer diameter of a single pile foundation; For horizontal wave induced velocity; For horizontal wave-induced acceleration; The surface wave distribution function is expressed as follows: ， ， ， in, Wave height; Wave number; The frequency of the wave; For wave period; This invention uses the power-law distribution method to calculate the flow load, and its expression is as follows: ， in, The horizontal flow resistance per unit length, This represents the local flow velocity.

11. The integrated analysis method for multi-scale damage of offshore wind power structures according to any one of claims 1-10, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S201: Create detailed models of the flange and bolts respectively; S202: Assign appropriate material properties to flanges and bolts respectively; S203: Assemble the completed models together and set their interactions; S204: Mesh the bolts and flanges separately, with the mesh type being an eight-node hexahedral linear reduced integral element.

12. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 11, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S205: As a vulnerable and critical component, bolts require a suitable damage model to accurately characterize their damage evolution in complex marine environments. This invention uses the GTN damage criterion to characterize the damage evolution process of bolts. The GTN damage criterion parameters for high-strength bolts are as follows: , , , , , , , , .

13. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 12, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S206: The stress-strain curve of a bolt can be divided into three stages: the elastic stage, the necking stage, and the post-necking stage. Numerical simulation requires the input of the material's actual stress-strain curve, which can be obtained by converting the engineering stress-strain curve. The calculation formula for the actual stress-strain curve is as follows: ， in, Bolt stress; It is the elastic modulus of the bolt; It is the strain of the bolt; The true stress-strain curve during the necking stage is represented by the following equation: ， ， in, , For true stress and true strain; , To test the obtained engineering stress and engineering strain; The starting point of the necking stage is usually the ultimate stress point in the material's tensile test. ， ， ， in, This is the weighting factor required for calibration, with a value between 0 and 1; other parameters must meet the continuity requirements of stress and strain. For linear strengthening modulus, ; The strain hardening index is... ; The stress intercept of the linear model. ; The strength coefficient, , ; , It represents the actual stress and strain at the material's ultimate stress point.

14. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 13, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S207: In establishing the equivalent model of the flange bolt connection, due to the significant differences in stress response in the three principal stress directions, the overall connection structure exhibits obvious anisotropic mechanical behavior. Although the Mises yield criterion is simple in form, its isotropic assumption makes it difficult to accurately describe the yield characteristics of such anisotropic materials. Therefore, this invention adopts the Hill anisotropic yield criterion for equivalent simulation. By introducing multi-directional yield strength parameters, this criterion can more effectively characterize the plastic deformation behavior of materials in different directions, thereby more accurately reflecting the actual mechanical response of the flange bolt connection under complex stress states. The specific expression of the Hill anisotropic yield criterion is as follows: ， ， ， ， in, The yield stress; These are the yield stress values ​​in various directions, measured along the three principal stress directions. , , It is normal stress. , , For shear stress components; F, G, and H are anisotropic yield parameters; It is the user-defined yield stress; It is the anisotropic yield stress ratio; ; The flow rule for the Hill anisotropic yield criterion is as follows: ， , in, This is the plastic strain increment vector; These are plastic multiplier vectors; is the normal vector of the yield surface in stress space; b is the plastic flow direction vector; F, G, H are the normal stress anisotropy weighting parameters, and N is the shear stress anisotropy parameter.

15. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 14, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S208: Based on the mechanical response characteristics of bolt and flange connection structures under external loads, this invention selects a representative volume element (RVE) as the basic unit for equivalent analysis. The RVE consists of a single bolt and its surrounding flange region. Its design fully reflects the key geometric features of bolted connections and can effectively capture local stress concentration phenomena and the anisotropic response characteristics of materials. At the same time, a representative volume element is established without considering the interaction between the flange and the bolt, that is, the bolt modeling is ignored and only the flange is established (i.e., a homogenized model).

16. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 15, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S209: Bolt preload is applied to the bolt by a cooling method, and the calculation formula is as follows: Bolt deformation under preload : ， ， Where F is the bolt preload; For bolt rod stiffness; It is the elastic modulus of the bolt; It is the length of the bolt rod; It is the cross-sectional area of ​​the bolt rod; It is the yield point of the bolt material (the yield point of a 10.9 grade bolt is 900 MPa). This is the bolt preload, which is typically [value missing]. ; Deformation of the connector under preload : ， This expression can be further written as: ， in, It is the coefficient of linear expansion of the bolt material. For the stiffness of the connection part, This is the initial temperature of the bolt, with a value of 0. The applied temperature is determined based on the magnitude of the bolt preload.

17. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 16, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S210: To clarify the influence of bolted connections on yielding behavior, this invention first compares the differences in yield points in the plane between the refined model and the model that ignores bolts, and obtains a comparison of the mechanical properties of the refined model and the model that ignores bolts. The calculation process is as follows: yield data of representative volume elements (RVE) under different stress states were obtained through numerical simulation; first, uniaxial tensile and compressive loads were applied to the three principal stress directions of the bolt-flange model and the traditional simplified model respectively to obtain their uniaxial yield strength; then, the influence of biaxial stress state on yield behavior was characterized by applying composite biaxial loads along different principal stress directions; finally, based on the above series of numerical test results, the yield surface of RVE in the principal stress plane was obtained.

18. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 17, characterized in that, In step S200, the establishment of the numerical model of the flange-bolt micro-multiscale includes: S211: Substitute the material parameters calculated by the refined model in the three principal stress directions into the homogenized model based on the Hill criterion.

19. The integrated analysis method for multi-scale damage of offshore wind power structures according to any one of claims 1-10, characterized in that, The steps for substituting the homogenized equivalent model established in step S200 into the overall structure and analyzing the dynamic response of the overall structure under various loads such as wind, waves, and earthquakes include: S301: Substitute the established homogeneous flange-bolt connection model into the overall structure; S302: First, perform ground stress balance, then apply gravity load to the entire model; S303: Apply wave load to the jacket foundation, wind load to the tower and rotor, and then apply seismic load to the overall model; S304: Analyze the dynamic response of the overall model under wind, wave, and seismic loads.

20. The integrated analysis method for multi-scale damage of offshore wind power structures according to claim 19, characterized in that, Based on the results of the overall structural calculations in S300, the flange-bolted connection structure is evaluated using sub-modeling technology, including: S401: Establish a detailed model of the flange and bolts; S402: Assign material parameters to the flange and bolt respectively; S403: Assemble the flange and bolt models together, mesh them, and set the corresponding contact relationships; S404: Extract the stress state at the flange-bolt connection in the overall model under wind and wave loads, substitute the extracted stress state at the flange-bolt connection into the sub-model, and analyze the established refined flange-bolt model.

21. An integrated analysis system for multi-scale damage of offshore wind power structures, employing the integrated analysis method for multi-scale damage of offshore wind power structures as described in any one of claims 1-20, characterized in that, include: The data acquisition and preprocessing module collects multi-field load data and structural geometry and material parameters of offshore wind power structures, and completes filtering, noise reduction and standardization preprocessing. The multi-scale model building module establishes an overall structural model of the offshore wind turbine, constructs a flange-bolt micro-multi-scale numerical model based on the GTN damage model, establishes a homogenized flange model through RVE homogenization and Hill anisotropic yield criterion, and builds a refined flange-bolt sub-model. The multi-field coupling analysis module embeds a homogenized equivalent model into the overall structure to analyze the overall dynamic response under multiple loads such as wind, waves, and earthquakes. It imports the flange-bolt stress state of the overall structure through sub-model technology to achieve multi-scale coupling calculation. The damage co-assessment module combines the overall structural dynamic response with the analysis results of local sub-models to characterize the stress distribution, plastic development, and damage evolution of the flange-bolted connection, and completes the co-assessment of global response and local damage. The results output and optimization module outputs structural damage assessment reports and remaining life prediction results. It optimizes the accuracy of the model through parameterized calibration and simultaneously provides graded early warning and operation and maintenance decision suggestions.

Images