A method for analyzing blade failure of an air compressor

Abstract

This invention discloses a method for blade failure analysis of air compressors, belonging to the field of blade failure analysis. The method includes: constructing a three-dimensional finite element analysis model based on the air compressor blade design parameters; dividing the computational domain into a fluid domain and a solid domain; correcting the fluid domain range until convergence; obtaining the strain of mesh elements under circulating wind load to obtain the macroscopic damage of the mesh elements; screening excessively damaged mesh elements, determining the authenticity of excessively damaged mesh elements, and interpolating and correcting the macroscopic damage; forming excessively damaged regions and calculating the weight of the excessively damaged regions based on the nearest distance; calculating the average macroscopic damage of the excessively damaged regions based on the macroscopic damage of each excessively damaged mesh element; calculating the overall damage intensity of the blade; and assessing the failure risk of the blade. This invention considers the degree of excessive loss at different locations and the weight of each location to comprehensively evaluate the overall failure situation of the blade, effectively improving the rationality of the failure analysis.

Description

Technical Field

[0001] This invention relates to the field of blade failure analysis, and more specifically to a method for blade failure analysis of an air compressor. Background Technology

[0002] As a core power equipment in industrial production, air compressor blades operate under harsh conditions of high temperature, high pressure, and high-speed rotation for extended periods. They are subjected to the combined effects of aerodynamic, centrifugal, and thermal loads, making them highly susceptible to failures such as fatigue fracture, corrosion wear, and creep deformation. Statistics show that blade failure accounts for more than 30% of all air compressor malfunctions. Once a blade breaks, it can lead to compressor shutdown or even complete failure, causing significant economic losses and safety hazards.

[0003] Before an air compressor is put into operation, failure analysis and testing of blade defects are crucial for extending its service life and ensuring its quality. Current technologies mostly employ two methods for blade failure analysis: one involves placing the blades in a real indoor environment for testing and collecting test data for failure analysis. This method is time-consuming, costly, and cannot quickly yield conclusions, nor does it reveal effective data analysis methods for reasonably analyzing whether failure has occurred. The other method is finite element analysis, which simulates real physical scenarios. However, current technologies mostly perform single-physics field analysis, and the applied fluid and aerodynamic fields have significant errors, leading to inaccurate stress, strain, and other related data, thus resulting in inaccurate failure analysis. Summary of the Invention

[0004] To address the aforementioned shortcomings in the prior art, this invention provides a method for analyzing the failure of air compressor blades. By planning a precise fluid domain, it enables accurate analysis of the failure process of air compressor blades.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A method for blade failure analysis of an air compressor is provided, comprising: Step S1: Construct a three-dimensional finite element analysis model based on the blade design parameters of the air compressor, and perform mesh generation. Divide the computational domain into a fluid domain and a solid domain. Apply preset aerodynamic parameters to the fluid domain. Calculate the displacement and temperature of the blade mesh elements in the solid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. Correct the fluid domain range until the fluid domain converges. Step S2: Obtain the strain of the blade grid cells under the set cyclic wind load after the fluid domain converges, construct a damage evolution model, calculate the damage evolution rate of each grid cell in the blade, and obtain the macroscopic damage of the grid cells. Step S3: Filter excessively damaged grid cells based on the macroscopic damage of the grid cells, and calculate the neighborhood distribution density of each excessively damaged grid cell based on the distance between them; determine the authenticity of the excessively damaged grid cells, and perform interpolation correction on the macroscopic damage of falsely excessively damaged grid cells. Step S4: Merge adjacent over-damaged mesh cells to form an over-damaged region, and calculate the nearest distance between the geometric center of the over-damaged region and the blade surface. Calculate the weight of the over-damaged region based on the nearest distance. Step S5: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged grid cell, and calculate the overall damage intensity of the blade by combining the weights of the over-damaged regions, and assess the failure risk of the blade.

[0006] Further, step S1 includes: Step S11: Construct a three-dimensional geometric model based on the blade design parameters of the air compressor, and mesh the three-dimensional geometric model using tetrahedral elements. Input the mesh into the finite element analysis software to obtain the three-dimensional finite element analysis model of the blade. Step S12: Construct the aerodynamic load field, centrifugal load field, and thermal load field of the three-dimensional finite element analysis model, divide the computational domain of the finite element analysis into the fluid domain and the solid domain, apply the preset aerodynamic parameters to the fluid domain, and solve the pressure distribution and temperature distribution of the fluid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. Step S13: Transfer the pressure and temperature distributions from the fluid domain to the solid domain, solve the heat conduction control equations and the elasticity equilibrium equations, and output the temperature field inside the blade. and the displacement field of the blade , These are the three-dimensional coordinates of the mesh cells; Step S14: Based on the displacement field Obtain the displacement of each grid cell within the blade, and filter the grid cells with the largest displacement within the blade. Based on the temperature field inside the blade Obtain the solid temperature of each grid cell within the blade, and then filter the grid cells with the highest solid temperature within the blade. ; Step S15: Set the blade deformation threshold and the temperature threshold that can be withstood ,like and If the fluid domain converges, then the fluid domain is determined to be converged; otherwise, the range of the fluid domain is adjusted, and the process returns to step S12. Furthermore, the method for adjusting the fluid domain range is as follows: based on the displacement field average Adjust the position of each grid cell within the blade, then return to step S2 until the set fluid domain converges. When the fluid domain converges, the element grid within the blade satisfies... and .

[0007] Furthermore, the centrifugal load field is represented by the distribution of centrifugal loads generated during blade rotation.

[0008] Furthermore, the thermal load field is generated by the combined effects of convective heat transfer and thermal conduction.

[0009] Further, step S2 includes: Step S21: Obtain the strain of the blade mesh element under the set cyclic wind load after the fluid domain converges, construct a damage evolution model based on the blade's damage initiation equivalent plastic strain, and calculate the damage evolution rate of each mesh element in the blade. Step S22: Integrate the damage evolution rate during the cyclic wind load process to obtain the macroscopic damage of each grid cell within the blade.

[0010] Further, step S3 includes: Step S31: Set the macroscopic damage threshold for the internal grid cell under a set duration of cyclic wind load. If the macroscopic damage is greater than the macroscopic damage threshold, the grid cell is determined to have excessive damage and is set as an excessively damaged grid cell; otherwise, the grid cell is determined not to have excessive damage. Step S32: Obtain all excessively damaged grid cells within the blade, and calculate the distance between one of the excessively damaged grid cells and the surrounding excessively damaged grid cells, and then calculate the neighborhood distribution density of the excessively damaged grid cells. Step S33: Set the neighbor distribution density threshold. If the neighbor distribution density is greater than the neighbor distribution density threshold, the excessively damaged mesh cell is determined to be a false excessively damaged mesh cell, and step S34 is executed. Otherwise, the excessively damaged mesh cell is determined to be a true excessively damaged mesh cell. Step S34: Interpolate and correct the macroscopic damage corresponding to the pseudo-over-damage mesh element, and update the macroscopic damage corresponding to the pseudo-over-damage mesh element.

[0011] Further, step S4 includes: Step S41: After the macroscopic damage update is completed for all pseudo-excessive damage mesh cells within the blade, adjacent excessive damage mesh cells are merged to form an excessive damage region, and the geometric center coordinates of the excessive damage region are calculated. Step S42: Calculate the distance between the geometric center and the mesh cells located on the blade surface based on the coordinates of the geometric center, and filter for the closest distance between the geometric center and the blade surface. ; Step S43: Based on the closest distance between the geometric center and the blade surface Calculate the weights of the excessively damaged regions.

[0012] Further, step S5 includes: Step S51: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged grid cell, and calculate the overall damage intensity of the blade by combining the weights of the over-damaged regions and the average macroscopic damage. ; Step S52: Utilize the overall damage strength of the blade Assess the failure risk and damage intensity under cyclic wind loads. The higher the value, the higher the risk of failure, and vice versa.

[0013] The beneficial effects of this invention are as follows: This invention performs finite element analysis on the blades of an air compressor. By dividing the computational domain into a fluid domain and a solid domain, and correcting the fluid domain based on data feedback from the solid domain, it ensures that an accurate fluid domain is formed during the finite element analysis process, making the aerodynamic and temperature fields experienced by the blades more accurate and reasonable.

[0014] Meanwhile, considering the coupled superposition effect of aerodynamic load field, centrifugal load field and thermal load field, and taking into account the interaction between fluid and solid, the deformation of the blade will affect the fluid flow, and the change of fluid flow will in turn affect the load and deformation of the blade, thus enabling accurate calculation of the blade load distribution under the coupling effect of multiple physics fields.

[0015] By applying long-term cyclic wind loads, the damage evolution process of the blades was characterized using macroscopic phenomenological damage analysis. The damage distribution of the macroscopic structure is calculated to provide data for subsequent failure analysis. Furthermore, this invention considers the degree of excessive loss at different locations and the weighting of these locations to comprehensively evaluate the overall failure status of the blade, effectively improving the rationality of the failure analysis. Attached Figure Description

[0016] Figure 1 This is a flowchart of a blade failure analysis method for air compressors. Detailed Implementation

[0017] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0018] like Figure 1As shown, a method for blade failure analysis of an air compressor includes: Step S1: Construct a three-dimensional finite element analysis model based on the air compressor blade design parameters, and perform mesh generation. Divide the computational domain into a fluid domain and a solid domain. Apply preset aerodynamic parameters to the fluid domain. Calculate the displacement and temperature of the blade mesh elements in the solid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. Correct the fluid domain range until the fluid domain converges.

[0019] Step S1 specifically includes the following steps: Step S11: Construct a three-dimensional geometric model based on the blade design parameters of the air compressor, and mesh the three-dimensional geometric model using tetrahedral elements. Input the mesh into the finite element analysis software to obtain the three-dimensional finite element analysis model of the blade. In this embodiment, the surface mesh size of the 3D geometric model is 0.5mm, and the internal mesh size is 2mm. The 3D geometric model is input into the finite element analysis software to obtain the 3D finite element analysis model. During the finite element analysis, the computational domain is divided into a fluid domain and a solid domain. The fluid domain inlet extends to 5 times the chord length of the blade leading edge, and the outlet extends to 10 times the chord length of the blade trailing edge. The solid domain is fixed based on the blade design parameters, while the fluid domain is adaptively adjusted according to the simulation requirements to ensure that the load applied to the solid domain meets the requirements.

[0020] Step S12: Construct the aerodynamic load field, centrifugal load field, and thermal load field of the three-dimensional finite element analysis model. Divide the computational domain of the finite element analysis into a fluid domain and a solid domain. Apply preset aerodynamic parameters to the fluid domain. Solve for the pressure distribution in the fluid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. and temperature distribution , These are the three-dimensional coordinates of the mesh cells; The governing equations for the aerodynamic load field are: ; in, For gas density, t For time, They are respectively x , y , z Gas velocity components along the coordinate direction, p For gas pressure, For gas dynamic viscosity, Let Kronecker function be used when hour, ,otherwise , They are respectively x , y Reynolds stress terms in the coordinate directions; The Reynolds stress term is closed using the SSTk-ω turbulence model: ;in, k For turbulent kinetic energy, For specific dissipation rate, For turbulent viscosity, For turbulent kinetic energy generation, F 1 is a mixed function. S The magnitude of the strain rate tensor These are model constants. The values ​​are 0.556, 0.075, 0.09, 1.176, 2.0, and 1.168.

[0021] The centrifugal load distribution generated by the rotation of the blade in the centrifugal load field is as follows: ; in, The radial centrifugal load density of the grid cells. r Distance from the grid cell to the axis of rotation Leave, For material density, The set blade rotational angular velocity; The mesh cells here include mesh cells in the fluid domain and the solid domain. In the fluid domain, the gas is driven to rotate by the rotation of the blades, which also generates radial centrifugal loads. The distribution of centrifugal loads in both the fluid domain and the solid domain is calculated using the above formula.

[0022] The thermal load field is generated by the combined effects of convective heat transfer and heat conduction. The governing equation for heat conduction is: ; in, The specific heat capacity at constant pressure of the blade material. T For temperature, The thermal conductivity of the blade material, Intensity of the internal heat source; The convective heat transfer boundary conditions on the blade surface are: ; in, n The surface normal vector of the blade surface mesh element. h The convective heat transfer coefficient is... The surface temperature of the blade. The gas temperature; Step S13: Distribute the pressure in the fluid domain and temperature distribution The heat transfer equations and elasticity equilibrium equations are solved in the solid domain to output the temperature field inside the blade. and the displacement field of the blade ; The elasticity equilibrium equation in this embodiment adopts the classical elasticity equilibrium equation, which is used to describe the balance relationship between the internal stress and external force of the blade. It is the fundamental equation for solving the stress and displacement of the internal grid element (elastic body) of the blade.

[0023] Step S14: Based on the displacement field Obtain the displacement of each grid cell within the blade, and filter the grid cells with the largest displacement within the blade. Based on the temperature field inside the blade Obtain the solid temperature of each grid cell within the blade, and then filter the grid cells with the highest solid temperature within the blade. ; Step S15: Set the blade deformation threshold and the temperature threshold that can be withstood ,like and If the fluid domain converges, then the fluid domain is determined to be converged; otherwise, the range of the fluid domain is adjusted, and the process returns to step S12. The method for adjusting the fluid domain range is as follows: based on the displacement field average Adjust the position of each grid cell within the blade, then return to step S2 until the set fluid domain converges. When the fluid domain converges, the element grid within the blade satisfies... and .

[0024] Step S2: Obtain the strain of the blade grid cells under the set cyclic wind load after the fluid domain converges, construct a damage evolution model, calculate the damage evolution rate of each grid cell in the blade, and obtain the macroscopic damage of the grid cells.

[0025] Step S2 specifically includes: Step S21: Obtain the strain of the blade mesh elements under the set cyclic wind load after the fluid domain converges. Based on the equivalent plastic strain of the blade damage initiation A damage evolution model was constructed to calculate the damage evolution rate of each grid cell within the blade. ; ; in, To reduce the rate of energy release from damage, Damage toughness represents the critical energy density at which a material resists damage development. Blades are typically made of metal, and the value is taken as... , The plastic strain rate of the internal grid elements of the blade under cyclic wind load. For the Heaviside step function, only if the cumulative plastic strain exceeds a threshold ( Damage only begins to evolve during the elastic phase (to avoid creating false damage during the elastic phase). The equivalent stress of the grid elements within the blade. Let be the triaxial stress function of the mesh element within the blade. E For elastic modulus, D For macroscopic scalar damage variables, v Poisson's ratio, The average stress experienced by the grid cells within the blade. s These are material constants; Step S22: Damage evolution rate during cyclic wind loading. Integrating, we obtain the macroscopic damage of each grid cell within the blade. ; ; in, These represent the start and end times of the cyclic wind load, respectively.

[0026] Step S3: Screen over-damaged grid cells based on the macroscopic damage of the grid cells, calculate the neighbor distribution density of each over-damaged grid cell based on the distance between the over-damaged grid cells, determine the authenticity of the over-damaged grid cells, and perform interpolation correction on the macroscopic damage of false over-damaged grid cells.

[0027] Step S3 specifically includes: Step S31: Set the internal mesh cells of the blade for the specified duration. Macroscopic damage threshold under cyclic wind load ,like If the damage is excessive, the mesh cell is determined to be an excessively damaged mesh cell; otherwise, the mesh cell is determined not to be excessively damaged. The circulating air load applied in this embodiment mainly includes: compressor inlet total pressure, incoming static pressure, incoming airflow angle of attack, load cycle amplitude, load cycle frequency, and aerodynamic load pulsation coefficient.

[0028] Step S32: Obtain all excessively damaged mesh elements within the blade. and with one of the excessively damaged mesh cells Calculate the distance between the center and the surrounding excessively damaged mesh cells. Then calculate the excessively damaged mesh element. Neighbor distribution density ; ; in, J For excessively damaged mesh elements The number of excessively damaged mesh cells in the surrounding area, jFor excessively damaged mesh elements Surrounding excessively damaged mesh cell numbers, For excessively damaged mesh elements Grid coordinates, For excessively damaged mesh elements Grid coordinates, To filter out excessively damaged mesh cells in the surrounding area; To filter out excessively damaged mesh elements, only mesh elements that are sufficiently close together can participate in the calculation. The distance threshold can be reasonably set based on the blade's operating scenario and the accuracy of the failure analysis. Generally, the distance threshold is set to the length of 10-100 cell grids. The discreteness of excessively damaged mesh elements is determined by calculating their neighborhood density, which in turn determines the authenticity of excessive damage. Generally, excessively damaged mesh elements cluster at the location of excessive damage, forming an excessively damaged region. Within this region, there are multiple consecutive excessively damaged mesh elements. The neighborhood density of excessively damaged mesh elements within the region is low, while the neighborhood density of those outside the region is high. This allows for the screening out of falsely excessively damaged mesh elements.

[0029] Step S33: Set the neighbor distribution density threshold ,like This will result in excessive damage to the mesh cells. If the mesh element is determined to be pseudo-over-damaged, proceed to step S34; otherwise, determine it as over-damaged. For true over-damaged mesh elements; Step S34: Macroscopic damage corresponding to pseudo-excessive damage mesh elements. Interpolation correction is performed to update the macroscopic damage corresponding to the pseudo-over-damage mesh cells; ; in, For the macroscopic damage corresponding to the updated pseudo-over-damaged mesh cells, Macroscopic damage to mesh cells surrounding pseudo-excessively damaged mesh cells. e Number the mesh cells surrounding the pseudo-excessively damaged mesh cell. E This represents the number of mesh cells surrounding a pseudo-excessively damaged mesh cell.

[0030] Since the mesh elements are tetrahedral, the average macroscopic damage of the two layers of mesh elements (a total of 8) surrounding the pseudo-excessively damaged mesh element is interpolated to correct the macroscopic damage corresponding to the pseudo-excessively damaged mesh element. During blade failure, the mesh elements, as micro-units, exhibit continuity in damage between them, without accidental abrupt changes. Therefore, taking the central average can effectively correct the macroscopic damage of the pseudo-excessively damaged mesh element, reducing the error in blade failure analysis.

[0031] Step S4: Merge adjacent over-damaged mesh cells to form an over-damaged region, calculate the nearest distance between the geometric center of the over-damaged region and the blade surface, and calculate the weight of the over-damaged region based on the nearest distance.

[0032] Step S4 specifically includes: Step S41: After the macroscopic damage update is completed for all pseudo-excessive damage mesh elements within the blade, adjacent excessive damage mesh elements are merged to form an excessive damage region, and the geometric center coordinates of the excessive damage region are calculated. ; ; in, w The numbering of the excessively damaged areas, A This represents the number of excessively damaged mesh cells in the excessively damaged region. Step S42: Based on the coordinates of the geometric center Calculate the distances between the blade and the mesh cells located on the blade surface, and filter for the closest distance between the geometric center and the blade surface. ; ; in, f Number the grid cells on the blade surface. For the blade surface, it is a set of grid cells. Mesh cell coordinates on the blade surface g Numbering of excessively damaged areas within the blade; Step S43: Based on the closest distance between the geometric center and the blade surface Calculate the weight of the excessively damaged region ; ; in, The thickness of the area where the excessive damage is located.

[0033] Step S5: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged grid cell, and calculate the overall damage intensity of the blade by combining the weights of the over-damaged regions, and assess the failure risk of the blade.

[0034] Step S5 specifically includes: Step S51: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged mesh element. Weights of the combined excessively damaged region and average macroscopic damage Calculate the overall damage intensity of the blade. ; ; Step S52: Utilize the overall damage strength of the blade Assess the failure risk and damage intensity under cyclic wind loads. The higher the value, the higher the risk of failure, and vice versa.

[0035] Weight of excessively damaged areas Weights are used to characterize the importance of the excessively damaged area during failure risk analysis. The larger the value, the closer the excessively damaged area is to the blade surface, the greater the risk of cracking, and the greater its impact on blade failure. Overall damage intensity. It characterizes the overall failure risk of the blade under cyclic wind load. The larger the value, the higher the risk of failure, and vice versa.

[0036] In this embodiment, during the finite element analysis, different excessive damage areas are marked based on their weights at different locations on the blade. Excess damage areas with higher weights are highlighted to draw attention to the failure risk of these areas.

Claims

1. A method for blade failure analysis of an air compressor, characterized in that, include: Step S1: Construct a three-dimensional finite element analysis model based on the blade design parameters of the air compressor, and perform mesh generation. Divide the computational domain into a fluid domain and a solid domain. Apply preset aerodynamic parameters to the fluid domain. Calculate the displacement and temperature of the blade mesh elements in the solid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. Correct the fluid domain range until the fluid domain converges. Step S2: Obtain the strain of the blade grid cells under the set cyclic wind load after the fluid domain converges, construct a damage evolution model, calculate the damage evolution rate of each grid cell in the blade, and obtain the macroscopic damage of the grid cells. Step S3: Filter excessively damaged grid cells based on the macroscopic damage of the grid cells, and calculate the neighborhood distribution density of each excessively damaged grid cell based on the distance between them; determine the authenticity of the excessively damaged grid cells, and perform interpolation correction on the macroscopic damage of falsely excessively damaged grid cells. Step S4: Merge adjacent over-damaged mesh cells to form an over-damaged region, and calculate the nearest distance between the geometric center of the over-damaged region and the blade surface. Calculate the weight of the over-damaged region based on the nearest distance. Step S5: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged grid cell, and calculate the overall damage intensity of the blade by combining the weights of the over-damaged regions, and assess the failure risk of the blade.

2. The method for blade failure analysis of an air compressor according to claim 1, characterized in that, Step S1 includes: Step S11: Construct a three-dimensional geometric model based on the blade design parameters of the air compressor, and mesh the three-dimensional geometric model using tetrahedral elements. Input the mesh into the finite element analysis software to obtain the three-dimensional finite element analysis model of the blade. Step S12: Construct the aerodynamic load field, centrifugal load field, and thermal load field of the three-dimensional finite element analysis model, divide the computational domain of the finite element analysis into the fluid domain and the solid domain, apply the preset aerodynamic parameters to the fluid domain, and solve the pressure distribution and temperature distribution of the fluid domain based on the aerodynamic load field, centrifugal load field, and thermal load field. Step S13: Transfer the pressure and temperature distributions from the fluid domain to the solid domain, solve the heat conduction control equations and the elasticity equilibrium equations, and output the temperature field inside the blade. and the displacement field of the blade , These are the three-dimensional coordinates of the mesh cells; Step S14: Based on the displacement field Obtain the displacement of each grid cell within the blade, and filter the grid cells with the largest displacement within the blade. Based on the temperature field inside the blade Obtain the solid temperature of each grid cell within the blade, and then filter the grid cells with the highest solid temperature within the blade. ; Step S15: Set the blade deformation threshold and the temperature threshold that can be withstood ,like and If the fluid domain converges, the fluid domain is determined to be converged; otherwise, the range of the fluid domain is adjusted, and the process returns to step S12.

3. The method for blade failure analysis of an air compressor according to claim 2, characterized in that, The method for adjusting the fluid domain range is as follows: based on the displacement field average Adjust the position of each grid cell within the blade, then return to step S2 until the set fluid domain converges. When the fluid domain converges, the element grid within the blade satisfies... and .

4. The method for blade failure analysis of an air compressor according to claim 2, characterized in that, The centrifugal load field is represented by the distribution of centrifugal loads generated when the blades rotate.

5. The method for blade failure analysis of an air compressor according to claim 2, characterized in that, The thermal load field is generated by the combined effects of convective heat transfer and thermal conduction.

6. The method for blade failure analysis of an air compressor according to claim 2, characterized in that, Step S2 includes: Step S21: Obtain the strain of the blade mesh element under the set cyclic wind load after the fluid domain converges, construct a damage evolution model based on the blade's damage initiation equivalent plastic strain, and calculate the damage evolution rate of each mesh element in the blade. Step S22: Integrate the damage evolution rate during the cyclic wind load process to obtain the macroscopic damage of each grid cell within the blade.

7. The method for blade failure analysis of an air compressor according to claim 6, characterized in that, Step S3 includes: Step S31: Set the macroscopic damage threshold for the internal grid cell under a set duration of cyclic wind load. If the macroscopic damage is greater than the macroscopic damage threshold, the grid cell is determined to have excessive damage and is set as an excessively damaged grid cell; otherwise, the grid cell is determined not to have excessive damage. Step S32: Obtain all excessively damaged grid cells within the blade, and calculate the distance between one of the excessively damaged grid cells and the surrounding excessively damaged grid cells, and then calculate the neighborhood distribution density of the excessively damaged grid cells. Step S33: Set the neighbor distribution density threshold. If the neighbor distribution density is greater than the neighbor distribution density threshold, the excessively damaged mesh cell is determined to be a false excessively damaged mesh cell, and step S34 is executed. Otherwise, the excessively damaged mesh cell is determined to be a true excessively damaged mesh cell. Step S34: Interpolate and correct the macroscopic damage corresponding to the pseudo-over-damage mesh element, and update the macroscopic damage corresponding to the pseudo-over-damage mesh element.

8. The method for blade failure analysis of an air compressor according to claim 7, characterized in that, Step S4 includes: Step S41: After the macroscopic damage update is completed for all pseudo-excessive damage mesh cells within the blade, adjacent excessive damage mesh cells are merged to form an excessive damage region, and the geometric center coordinates of the excessive damage region are calculated. Step S42: Calculate the distance between the geometric center and the mesh cells located on the blade surface based on the coordinates of the geometric center, and filter for the closest distance between the geometric center and the blade surface. ; Step S43: Based on the closest distance between the geometric center and the blade surface Calculate the weights of the excessively damaged regions.

9. The method for blade failure analysis of an air compressor according to claim 8, characterized in that, Step S5 includes: Step S51: Calculate the average macroscopic damage of the over-damaged region based on the macroscopic damage of each over-damaged grid cell, and calculate the overall damage intensity of the blade by combining the weights of the over-damaged regions and the average macroscopic damage. ; Step S52: Utilize the overall damage strength of the blade Assess the failure risk and damage intensity under cyclic wind loads. The higher the value, the higher the risk of failure, and vice versa.

Images