Product integrated modeling method and system based on multi-system and multi-level collaboration

Abstract

The present invention belongs to the technical field of mechanical flexible optimization design, and specifically relates to a product integrated modeling method and system based on multi-system and multi-level collaboration. It aims to solve the technical problems of low precision and low reliability of flexible design for complex system structures in the mechanical flexible design method in the prior art. By generating a preliminary physical model of the product to be modeled, modifying the modifiable feature parameters, updating the physical model of the product to be modeled, constructing a finite element analysis model for static analysis and modal analysis, creating an optimization data model based on model design variables, optimization objectives and constraints, and solving multiple optimization objectives to obtain the optimal design variables, and applying the optimal design variables to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled, the effective optimization of design parameters is achieved, and the design speed and design accuracy are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of server heat dissipation, and in particular relates to a product integration modeling method and system based on multi-system and multi-level collaboration. Background Art

[0002] With the advancement of automation and intelligent manufacturing, the traditional machinery manufacturing industry is undergoing profound industrial transformation and technological upgrades to meet the needs of rapidly responding to and tracking diverse and personalized market demands in real time. The demand for rapid, small-batch, and customized production places higher standards on production quality, efficiency, and cost control. Therefore, integrating information technology and intelligent methods into design and production has become the core of the machinery manufacturing industry's development. For the machinery manufacturing industry, design is the core driving force behind mechanical manufacturing. To meet the concept of customized and personalized design, design methods are shifting from single-line drawing design to a flexible integration of multi-system and multi-level mechanical design processes, including 3D modeling, performance analysis, and structural optimization. Reasonable and efficient flexible design not only reduces processing complexity and production costs during the manufacturing process, improves product quality and consistency, but also reduces material waste, lowers energy consumption, and improves the environmental friendliness of the production process.

[0003] The introduction of innovative technologies such as digitalization, artificial intelligence, and the Internet of Things (IoT) has enhanced the intelligence and flexibility of machinery and equipment. This has led to the recently proposed "digital triad" model, which combines digital twins with AI to enhance machine tools' ability to self-regulate and optimize during production, enabling them to better adapt to complex and changing production demands. Adding sensors to legacy machinery and integrating it with the Industrial Internet of Things (IIoT) offers a low-cost, efficient solution for upgrading traditional equipment to intelligent capabilities, significantly improving its responsiveness and flexibility.

[0004] However, the mechanical flexible design method in the existing technology still needs to improve the accuracy and reliability of flexible design for complex system structures. Therefore, how to improve the flexible transformation of mechanical design in the existing technology, optimize the design parameters, and improve the design speed and design accuracy are technical problems that need to be solved urgently. Summary of the Invention

[0005] The purpose of the present invention is to provide a product integrated modeling method and system based on multi-system and multi-level collaboration, so as to improve the flexible transformation of mechanical design in the existing technology, optimize the design parameters, and improve the design speed and design accuracy.

[0006] In order to solve the above technical problems, the technical solutions adopted by the present invention are as follows:

[0007] First, the product integration modeling method based on multi-system and multi-level collaboration includes the following steps:

[0008] S1: Call the model data file based on the product to be modeled, set the key parameters required for creation, and generate the physical model of the product to be modeled;

[0009] S2: Extract input parameters and parameterize them, generate modifiable feature parameters in the interactive interface, modify the modifiable feature parameters, and update the physical model of the product to be modeled;

[0010] S3: Pre-process the updated solid model of the product to be modeled, build a finite element analysis model for static analysis and modal analysis, and finally complete the solution and export the analysis result report;

[0011] S4: Set model design variables, optimization objectives, and constraints, and create an optimization data model based on the model design variables, optimization objectives, and constraints;

[0012] S5: Based on the optimization goal, a preset multi-objective optimization algorithm is used to search the solution space of the optimization data model, an optimal solution is obtained from the solution space, optimal design variables are obtained, and the optimal design variables are applied to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

[0013] Preferably, in step S1, the specific process of calling the model data file based on the product to be modeled, setting the key parameters required for creation, and generating the physical model of the product to be modeled is as follows:

[0014] S11: Input the specified parameters and start the modeling task;

[0015] S12: Read the file containing the coordinate data of the cross-section outline points of key parts, create features using the secondary development interface function, and add specified constraints to generate a three-dimensional solid model.

[0016] Preferably, the specific process of preprocessing the updated solid model of the product to be modeled in step S3 is: meshing, adding materials, setting contacts, adding loads, adding constraints and setting boundary conditions for the updated solid model of the product to be modeled.

[0017] Preferably, the static analysis of the finite element analysis model in step S3 includes axial stress analysis and deformation analysis, and the modal analysis includes constant frequency analysis and vibration analysis, and a Campbell diagram is generated.

[0018] Preferably, the product to be modeled is a long blade of the last stage of a steam turbine, the design variables include the tie bar height, the tie bar angle, and the shroud thickness, the constraints include the equivalent stress and the maximum deformation value, and the formula of the established optimization data model is as follows, with each triple resonance frequency as the objective function:

[0019] ;

[0020] in, X is the design variable, P 1 represents the stretching height, P 2 represents the stretching angle of the tendons. P 3 represents the thickness of the belt, f 1( X ) represents the second-order, third-pitch vibration frequency of the blade over the entire cycle at rated speed. f 2( X ) is the 3rd order 7th pitch vibration frequency of the blade at rated speed. 、 Represent the weights of the three optimization objective functions, P i,min and P i,max represent the lower and upper limits of the design variables, respectively. G 1( X ) represents the 4th order 8th pitch frequency of the blade at rated speed. In order to avoid the resonance triple point generated by the intersection of the 4th order 8th pitch frequency curve and the K=8th frequency line, G 1( X )≥206Hz, G 2( X ) represents the maximum equivalent stress of the blade, G 2( X )<850MPa, G 3( X ) represents the maximum equivalent stress of the impeller, G 3( X )<760MPa.

[0021] Preferably, the optimization target is set as follows:

[0022] The optimization target of the 2nd order 3rd node frequency f1(X) is set to be maximized, and the optimization target of the 3rd order 7th node frequency f2(X) is set to be minimized.

[0023] In a second aspect, a product integrated modeling system based on multi-system and multi-level collaboration is provided, which is used to implement any one of the product integrated modeling methods based on multi-system and multi-level collaboration, including a design resource database, a product model generation module, a parameterization and simulation analysis module, and a parameter optimization model, wherein the parameterization and simulation analysis module includes a preprocessing module and an automatic analysis module, the design resource database is connected to the product model generation module, the product model generation module is connected to the parameterization and simulation analysis module, and the parameterization and simulation analysis module is connected to the parameter optimization model;

[0024] The product model generation module is used to retrieve the model data file from the design resource database, set the key parameters required for creation, and generate the physical model of the product to be modeled;

[0025] The product model generation module is used to extract input parameters and parameterize them, generate modifiable feature parameters in the interactive interface, modify the modifiable feature parameters, and update the physical model of the product to be modeled;

[0026] The preprocessing module is used to preprocess the updated entity model of the product to be modeled;

[0027] The automatic analysis module is used to construct a finite element analysis model to perform static analysis and modal analysis, and finally complete the solution and export the analysis result report;

[0028] The parameter optimization model is used to set model design variables, optimization objectives and constraints, create an optimization data model based on the model design variables, optimization objectives and constraints, and based on the optimization objectives, preset a multi-objective optimization algorithm to search the solution space of the optimization data model, obtain the optimal solution from the solution space, obtain the optimal design variables, and apply the optimal design variables to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

[0029] Preferably, the parameter optimization model includes an experimental design module, an acquisition response module, an agent module and a multi-objective optimization module, wherein the experimental design module is connected to the acquisition response module, the acquisition response module is connected to the agent module, and the agent module is connected to the multi-objective optimization module;

[0030] The experimental design module is used to select parameters, design methods and generate experimental points;

[0031] The response acquisition module is used to update the model and obtain simulation results;

[0032] The agent module is used to generate a response surface and evaluate effectiveness;

[0033] The multi-objective optimization module is used to obtain the best design parameter combination by optimizing parameters and optimizing algorithm settings.

[0034] The beneficial effects of the present invention include:

[0035] The product integrated modeling method and system based on multi-system and multi-level collaboration provided by the present invention address the technical problems of low precision and low reliability of flexible design of complex system structures in the mechanical flexible design method in the existing technology. By generating a preliminary physical model of the product to be modeled, modifying the modifiable feature parameters, updating the physical model of the product to be modeled, constructing a finite element analysis model for static analysis and modal analysis, creating an optimization data model based on model design variables, optimization objectives and constraints, and solving multiple optimization objectives to obtain the optimal design variables, the optimal design variables are applied to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled, thereby achieving effective optimization of design parameters and improving design speed and design accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a flow chart of the product integration modeling method based on multi-system and multi-level collaboration of the present invention.

[0037] Figure 2 Schematic diagram of the overall architecture of the product integrated modeling system of the present invention.

[0038] Figure 3 Schematic diagram of the functional architecture of the integrated modeling system of the present invention.

[0039] Figure 4 Schematic diagram of the blade-impeller model of the present invention.

[0040] Figure 5 A schematic diagram of the blade body of the present invention is drawn.

[0041] Figure 6 Schematic diagram of the blade root cross section of the present invention, wherein (a) is a schematic diagram of the blade root tooth profile, and (b) is a schematic diagram of the closed blade root cross section curve.

[0042] Figure 7 Schematic diagram of drawing the cross-sectional profile of the belt of the present invention, (a) is a schematic diagram of creating the coordinate line of the belt profile, and (b) is a schematic diagram of the fitting effect of the belt profile.

[0043] Figure 8 Schematic diagrams are created for the blade root and shroud features of the present invention, (a) is a schematic diagram of the effect after the blade root cross section is stretched, and (b) is a schematic diagram of the effect after the shroud cross section is stretched.

[0044] Figure 9 It is a schematic diagram of the blade root feature removal operation of the present invention.

[0045] Figure 10 Schematic diagram of the key dimensions of the reinforcement cross section of the present invention.

[0046] Figure 11It is a schematic diagram of the reinforcement entity structure of the present invention.

[0047] Figure 12 This is a sketch of the impeller outline of the present invention.

[0048] Figure 13 It is a schematic diagram of the impeller structure of the present invention.

[0049] Figure 14 It is a schematic diagram of the contact arrangement of the blade part of the present invention.

[0050] Figure 15 Schematic diagram of the mesh division of the blade as a whole and the contact parts of the present invention, where (a) is a schematic diagram of a mountain area of ​​the blade, (b) is a schematic diagram of the contact between the ribs, (c) is a schematic diagram of the contact between the shrouds, and (d) is a schematic diagram of the contact between the blade root and the impeller groove.

[0051] Figure 16 It is the total deformation cloud map.

[0052] Figure 17 It is the X-axis deformation cloud map.

[0053] Figure 18 It is the Y-axis deformation cloud map.

[0054] Figure 19 It is the Z-axis deformation cloud map.

[0055] Figure 20 This is a pitch frequency diagram of a full circle of blades of the present invention.

[0056] Figure 21 This is the response surface model of the design variables and the 2nd order and 3rd nodal frequency of the present invention.

[0057] Figure 22 This is the response surface model of the design variables of the present invention and the 3rd order and 7th nodal frequency.

[0058] Figure 23 This is the Campbell diagram of the 2nd order 0-8 pitch diameter vibration of the blade throughout the entire cycle after optimization of the present invention.

[0059] Figure 24 This is the Campbell diagram of the 2nd order 0-8 pitch diameter vibration of the blade throughout the entire cycle after optimization of the present invention. DETAILED DESCRIPTION

[0060] The following is combined with Figures 1-21 The present invention is described in further detail:

[0061] Example 1

[0062] See attached Figure 1 As shown in the figure, the product integration modeling method based on multi-system and multi-level collaboration includes the following steps:

[0063] S1: Call the model data file based on the product to be modeled, set the key parameters required for creation, and generate the physical model of the product to be modeled;

[0064] S2: Extract input parameters and parameterize them, generate modifiable feature parameters in the interactive interface, modify the modifiable feature parameters, and update the physical model of the product to be modeled;

[0065] S3: Pre-process the updated solid model of the product to be modeled, build a finite element analysis model for static analysis and modal analysis, and finally complete the solution and export the analysis result report;

[0066] S4: Set model design variables, optimization objectives, and constraints, and create an optimization data model based on the model design variables, optimization objectives, and constraints;

[0067] S5: Based on the optimization goal, a preset multi-objective optimization algorithm is used to search the solution space of the optimization data model, an optimal solution is obtained from the solution space, optimal design variables are obtained, and the optimal design variables are applied to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

[0068] The product integration modeling method based on multi-system and multi-level collaboration of the present invention includes flexible parametric modeling, flexible simulation analysis, and flexible multi-objective parameter optimization. Flexible parametric modeling enables rapid generation and modification of models by defining key geometric parameters (such as blade length, thickness, inclination, etc.) and their associated relationships. This modeling method can flexibly adjust the model according to different working conditions and can complete the exploration and comparison of multiple design schemes in a short period of time. For example, when the design requirements change, only a few key parameters need to be adjusted, and the system can automatically update the entire model without the need to re-model from scratch. This flexibility not only shortens the design cycle, but also improves the diversity and adaptability of design schemes.

[0069] Flexible simulation analysis, based on parametric models, allows real-time parameter adjustments in a virtual environment to observe their impact on performance. This allows for early identification of potential issues during the design phase, avoiding the waste of physical testing. By adjusting the blade's geometric parameters, the system rapidly generates a new finite element model and performs simulation analysis, providing real-time feedback on key performance indicators such as stress distribution, deformation characteristics, and vibration modes. This real-time feedback and rapid iteration capabilities enable flexible response to complex design constraints and multi-physics coupling effects, thereby optimizing product structural parameters and improving performance.

[0070] Flexible multi-objective parameter optimization can be achieved by combining intelligent optimization algorithms with performance analysis and intelligent prediction models to construct a multi-parameter collaborative optimization model and propose a comprehensive optimal solution. This allows for simultaneous consideration of multiple performance indicators (such as structural strength and vibration characteristics) and the identification of optimal design solutions under complex constraints. For example, the system can dynamically adjust optimization objectives and constraints based on design requirements, traversing data to obtain multi-objective optimization results. It can adjust optimization strategies based on actual needs at different design stages to obtain a comprehensive optimal solution that meets multiple performance requirements. This allows for rapid iteration and optimization, significantly improving product mechanical design efficiency and performance while reducing development costs and time.

[0071] In this embodiment, the specific process of calling the model data file based on the product to be modeled, setting the key parameters required for creation, and generating the physical model of the product to be modeled in step S1 is as follows:

[0072] S11: Input the specified parameters and start the modeling task;

[0073] S12: Read the file containing the coordinate data of the cross-section outline points of key parts, create features using the secondary development interface function, and add specified constraints to generate a three-dimensional solid model.

[0074] Example 2

[0075] Based on Example 1, the specific process of preprocessing the updated solid model of the product to be modeled in step S3 includes meshing, adding materials, setting contacts, adding loads, adding constraints, and setting boundary conditions. The static analysis of the finite element analysis model in step S3 includes axial stress analysis and deformation analysis, and the modal analysis includes constant frequency analysis and vibration analysis, generating a Campbell diagram.

[0076] The product to be modeled is a long blade at the last stage of a steam turbine. The design variables include the tie bar height, tie bar angle, and shroud thickness. The constraints include the equivalent stress and maximum deformation value. The formula for the established optimization data model is as follows, with each triple resonance frequency as the objective function:

[0077] ;

[0078] in, X is the design variable, P 1 represents the stretching height, P 2 represents the stretching angle of the tendons. P 3 represents the thickness of the belt, f 1( X ) represents the second-order, third-pitch vibration frequency of the blade over the entire cycle at rated speed. f 2( X) is the 3rd order 7th pitch vibration frequency of the blade at rated speed. 、 Represent the weights of the three optimization objective functions, P i,min and P i,max represent the lower and upper limits of the design variables, respectively. G 1( X ) represents the 4th order 8th pitch frequency of the blade at rated speed. In order to avoid the resonance triple point generated by the intersection of the 4th order 8th pitch frequency curve and the K=8th frequency line, G 1( X )≥206Hz, G 2( X ) represents the maximum equivalent stress of the blade, G 2( X )<850MPa, G 3( X ) represents the maximum equivalent stress of the impeller, G 3( X The optimization objectives are set as follows: the optimization objective of the 2nd order 3rd node frequency f1(X) is set to be maximized, and the optimization objective of the 3rd order 7th node frequency f2(X) is set to be minimized.

[0079] A product integrated modeling system based on multi-system, multi-level collaboration, for implementing any of the aforementioned methods for multi-system, multi-level collaboration, comprises a design resource database, a product model generation module, a parameterization and simulation analysis module, and a parameter optimization model. The parameterization and simulation analysis module comprises a preprocessing module and an automatic analysis module. The design resource database is connected to the product model generation module, which is then connected to the parameterization and simulation analysis module, which is then connected to the parameter optimization model. The product model generation module is configured to retrieve model data files from the design resource database, set key parameters required for creation, and generate a physical model of the product to be modeled. The product model generation module is configured to extract input parameters and parameterize them, generate modifiable feature parameters in an interactive interface, modify these modifiable feature parameters, and update the physical model of the product to be modeled. The preprocessing module is configured to preprocess the updated physical model of the product to be modeled. The automatic analysis module is configured to construct a finite element analysis model, perform static and modal analysis, and ultimately generate and export an analysis result report. The parameter optimization model is used to set model design variables, optimization objectives and constraints, create an optimization data model based on the model design variables, optimization objectives and constraints, and based on the optimization objectives, preset a multi-objective optimization algorithm to search the solution space of the optimization data model, obtain the optimal solution from the solution space, obtain the optimal design variables, and apply the optimal design variables to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

[0080] The parameter optimization model includes an experimental design module, a response acquisition module, an agent module, and a multi-objective optimization module. The experimental design module is connected to the response acquisition module, which is connected to the agent module, and the agent module is connected to the multi-objective optimization module. The experimental design module is used to select parameters, design methods, and generate experimental points. The response acquisition module is used to update the model and obtain simulation results. The agent module is used to generate a response surface and evaluate effectiveness. The multi-objective optimization module is used to obtain the optimal design parameter combination by optimizing parameters and setting the optimization algorithm.

[0081] See also Figure 2 The modeling system of the present invention is provided with a user interface layer, a functional module layer, a data layer and a system layer to build an integrated flexible design system. The system architecture is as follows Figure 3 By integrating different functional modules, calling multiple system resources, and conducting multi-level collaborative improvements in the design process, the flexibility and automatic maintenance of the entire mechanical design process are achieved, and multi-objective collaborative optimization of design parameters for different design requirements is achieved.

[0082] See also Figure 3-Figure 16Taking the long blades of the last stage of a steam turbine as an example, integrated flexible modeling was conducted, focusing on verifying flexible parameter modeling, flexible simulation analysis, and flexible multi-objective parameter optimization. As the core component of the steam turbine, the last-stage blade is located in the last stage of the turbine and is in the final stage of steam expansion. The steam pressure and temperature have been significantly reduced. Therefore, its design and optimization are key to improving turbine efficiency. In addition, because the last-stage blade has the longest tip diameter and low rigidity, it is prone to excessive deformation and resonance under the action of centrifugal force and unstable airflow force. Therefore, a comprehensive analysis of the stress, deformation, and vibration characteristics of the blade under different operating conditions is carried out to verify whether its structural strength, rigidity, and stability meet the requirements. At the same time, a comprehensive optimization scheme is proposed based on these analysis results, which is of great significance for the design and optimization of the blade.

[0083] Single steam turbine last stage blade-impeller model, such as Figure 4 As shown, the blade consists of a straight-tooth fir-tree root blade and its associated impeller. The main features of the blade from top to bottom include the shroud, blade body, tie bars, and straight-tooth fir-tree root. The impeller consists of a wedge-shaped plate with a fir-tree-shaped tongue and groove connected to the root. This paper uses feature modeling technology to parametrically construct the structure of a typical straight-tooth fir-tree root steam turbine last-stage long blade (including the blade profile, blade root, shroud, tie bars, etc.). By modifying key parameters through a compiled program, the blade structure can be precisely and quickly adjusted.

[0084] The blade profile set for the final long blade consists of 26 closed smooth spline curves. A program reads a data file of aerodynamically designed cross-section profile coordinate points, sequentially generating X, Y, and Z coordinate values. The function UF_CURVE_create_point is then used to create all profile coordinate points. Subsequently, a smooth closed profile is generated using the curve fitting function. Because the curvature of the blade back and basin differs significantly from that of the leading and trailing edges, this paper defines NXOpen::Features::FitCurveBuilder to fit four curves, one for the leading edge, one for the trailing edge, one for the basin, and one for the blade back. These point objects are then added to the fitted point set using the Add function, ultimately generating a complete closed profile.

[0085] After obtaining the 26 cross-sectional profiles of the blade, similar to the process of fitting a curve from points, after defining the basic settings of ThroughCurvesBuilder, use the statement<NXOpen::Features::FitCurve*> Define an array to store curve objects, then use the FindObject function to find the generated profile curve and add it to the defined curve array. Execute the "Through Curve Group" function to complete the creation of the blade body entity. The resulting entity model is as follows Figure 5 shown.

[0086] The blade root is an important part connecting the blade and the impeller disk. First, draw the blade root cross-section profile in the specified plane, then fit the arc segment into a smooth quadratic spline curve and use the NXOpen C API to generate straight line segments, and finally form a closed cross-section profile consisting of arc segments and straight line segments, as shown in the following figure: Figure 6 The blade root width is determined by the number of blade periods and can be adjusted later by stretching and cutting operations.

[0087] The drawing method of the cross-section profile of the shroud is similar to that of the blade root. The coordinate point data of the profile curve of the shroud is read from the outside through the command, and then the profile coordinate points are created accordingly using the point feature creation function, such as Figure 6 As shown, the coordinate points are connected into a complete closed curve with the help of the straight line creation function and the curve fitting function, as shown in Figure 7 shown.

[0088] The blade root and shroud entities are both formed by stretching closed lines. Use the UF_MODL_create_extruded function to implement the stretching operation. Before using this function, you must first define a linked list LineList and the stretching direction vector Direction as well as the start and end coordinates of the stretching. Then set the Boolean operation type Sign of the stretching generated entity and the blade body entity to UF_UNSIGNED, which means that the part intersecting with the target entity is obtained. The effect of stretching the shroud and blade root cross-section lines is as follows Figure 8 As shown. After the blade root section is stretched, a cutting operation is performed to determine the blade root width of a single blade model according to the number of cycles divided by the entire circle of the turbine last stage blade. After the cutting operation is completed, the blade root structure is as follows Figure 9 shown.

[0089] The tie bars are located on both sides of the blade. Their cross-sectional profiles are created differently from the blade, root, and shroud. Instead, they are constructed directly based on externally imported profile coordinate point data and created directly in the sketch according to the specified dimensions and constraints. The cross-sectional shape of the tie bar is elliptical, and its key dimensional parameters are as follows: Figure 10 As shown. In addition to the functions of stretching and cutting, the construction of the reinforcement entity also includes the drafting operation realized by the function UF_MODL_create_taper_from_faces() and the rounding operation realized by the function UF_MODL_create_blend(). The completed reinforcement entity structure is as follows Figure 11 shown.

[0090] The impeller is mainly Figure 12The impeller outline shown is generated by rotating it at a certain angle. This angle is determined by the number of cycles divided by the entire blade. Then, using the groove profile that matches the blade root, a portion of the entity is cut off at the specified location. Finally, filleting is completed at some locations to create a 3D solid impeller, as shown in the following example. Figure 13 shown.

[0091] The entire turbine blade stage model is quite large. Performing modal analysis directly on the entire stage would significantly increase computational time and cost. Therefore, leveraging the cyclic symmetry of the entire turbine blade stage, a single cycle of the blade model is used as a sector for finite element simulation analysis. Using the cyclic symmetry settings in ANSYS, the program generates additional repeating sectors along the specified direction during the solution.

[0092] A single sector was selected as the assembly model of the blade and impeller. Its total length is 2755 mm. The data of the impeller and blade materials are shown in Table 1 below. The cyclic symmetry plane on the blade is selected inside the blade. Part of the structure on one side of the shroud, ribs, and blade root is cut out. The cut part is rotated through a period around the cyclic coordinate system so that it matches the shroud, ribs, and blade root on the other side. Figure 14 shown.

[0093] Table 1 Blade and impeller material parameters

[0094] Since the blade structure is relatively complex and contains multiple curved surfaces, the mesh quality obtained by the global automatic meshing method is poor, so the partitioning method is adopted. For the impeller part with a more regular shape, hexahedron meshing is used for division, the blade body and the shroud part are swept with hexahedron meshing, and the ribs and blade roots are divided with tetrahedron meshing. The mesh of the contact surface between the blade root and the impeller tongue and groove is refined. After weighing and comparing, the meshing results of each key part and the overall meshing are as follows: Figure 15 The number and proportion of grids in each part are shown in Table 2.

[0095] Table 2 Type and quantity of blade model units

[0096]

[0097] Total deformation cloud from turbine blades Figure 16 As shown in the figure, the maximum total deformation is located at the crown edge, with a value of 3.3178 mm. The nephogram of the blade's deformation in all directions shows that, among the three directions (X, Y, and Z), the maximum deformation in the X direction is also located at the crown edge, with a value of 3.1334 mm. The maximum deformation in the Y and Z directions is located at the trailing edge of the lower-middle part of the blade, with maximum deformation values ​​of 1.139 mm and 1.0803 mm, respectively.

[0098] Different from the vibration characteristics of a single blade, the natural vibration of the entire blade circle exhibits characteristics such as pitch circle vibration and pitch diameter vibration. The vibration performance of the last long blade was solved by the cyclic symmetry modal analysis method, and the first six natural frequencies of the turbine blades from 0 to 46 pitch diameters under the high-speed rotation condition of 1650 rpm were finally obtained. The curve of the change of frequency with the number of pitch diameters shows that, as shown in the figure below, Figure 17 The larger the number of node diameters, the higher the natural frequency. As the number of node diameters gradually increases, the frequency change becomes smaller and smaller and gradually tends to be stable.

[0099] The frequency modulation optimization design for the long blades of the last stage of a steam turbine is a typical form and size optimization. Since the flow path dimensions, air flow rate, and pressure ratio of each blade stage are essentially determined after the overall turbine design is completed, the aerodynamic shape parameters of the blades are often not modified during frequency modulation optimization. Instead, characteristic parameters of the tie bars and shrouds of the long blades of the last stage are selected as design variables for frequency modulation optimization of the entire turbine blade circumference. The specific optimization design variables selected are shown in Table 3 below.

[0100] Table 3 Optimization design variables for frequency regulation of the last stage long blades of the steam turbine

[0101]

[0102] The ultimate goal of frequency modulation optimization for the long last-stage blades of a steam turbine is to minimize the likelihood of "triple-point" resonance within the critical speed range while meeting the full-circle blade strength and stiffness design requirements. Therefore, when conducting multi-objective optimization of the long last-stage blades of a steam turbine, to ensure safe operation, comprehensive requirements for stress, deformation, and frequency must be met simultaneously. In this multi-objective frequency modulation optimization process for the long last-stage blades of a steam turbine, equivalent stress and maximum deformation values ​​are used as constraints, and the resonance frequencies of the "triple-point" resonances are used as objective functions to obtain the optimal combination of design variable values.

[0103] Based on the selection and analysis of the above-mentioned design variables, constraints and optimization objectives, a multi-objective frequency regulation optimization mathematical model for the last-stage long blades of the steam turbine is established, as shown in the following formula.

[0104] ;

[0105] in, X is the design variable, P 1 represents the stretching height, P 2 represents the stretching angle of the tendons. P 3 represents the thickness of the belt, f 1( X ) represents the second-order, third-pitch vibration frequency of the blade over the entire cycle at rated speed. f 2( X) is the 3rd order 7th pitch vibration frequency of the blade at rated speed. 、 Represent the weights of the three optimization objective functions, P i,min and P i,max represent the lower and upper limits of the design variables, respectively. G 1( X ) represents the 4th order 8th pitch frequency of the blade at rated speed. In order to avoid the resonance triple point generated by the intersection of the 4th order 8th pitch frequency curve and the K=8th frequency line, G 1( X )≥206Hz, G 2( X ) represents the maximum equivalent stress of the blade, G 2( X )<850MPa, G 3( X ) represents the maximum equivalent stress of the impeller, G 3( X )<760MPa. The optimization target is set to: set the 2nd order 3rd node frequency f 1( X ) is optimized to maximize the frequency of the 3rd order and 7th node diameter. f 2( X ) is optimized to minimize.

[0106] The response surface model between the three optimization objectives and the design variables is as follows: Figure 18 and Figure 19 As shown in the response surface model, it can be seen intuitively that among the three design variables of tie height, tie angle and belt thickness, the belt thickness has the greatest impact on the three frequency responses. As the belt thickness gradually increases, the second-order third-node frequency f 1( X ) gradually decreases, and the 3rd order 7th pitch frequency f 2( X ) gradually increases. As the height of the reinforcement increases, the frequency f 1( X ) shows a trend of increasing first and then decreasing, and the frequency f 2( X ) decreases slightly. As the tensioning angle increases, the frequency f 1( X ) gradually increases, f 2( X ) shows a trend of first decreasing and then increasing.

[0107] By simulating natural selection and genetic variation, the solution space is searched and the optimal solution is selected according to the Pareto optimality principle. The optimization target is set to MOGA, and the 2nd order and 3rd node frequency are set. f 1( X ) is optimized to maximize the frequency of the 3rd order and 7th node diameter. f 2( X ) is optimized to minimize f 1( X )>77.721, f 2( X )<171.9, set the constraint 4th order 8th diameter frequency G 1≥206. Set the maximum value of blade equivalent stress G 2<850, set the maximum value of impeller equivalent stress G 3<760. The optimal design parameter value combination obtained through iterative solution is shown in Table 4.

[0108] Table 4 Comparison of design variables before and after optimization

[0109]

[0110] According to the optimized design variables, the parametric model of the blade is changed, and the prestressed modal analysis is automatically performed to obtain the two frequencies of the optimized blade. f 1( X )and f 2( X ) are compared with the predicted values ​​given by multi-objective optimization and the frequency values ​​before optimization, as shown in Table 5.

[0111] Table 5 Comparison of target values ​​and predicted values ​​before and after optimization

[0112]

[0113] Redraw the Campbell diagram of the second and third order natural frequencies of the optimized model at 0~8 pitch diameters as a function of speed, as shown in Figure 20 and 21 As shown. Figure 20 Information calculation shows that the intersection of the optimized frequency multiplication curve K=3 and the 3-pitch diameter curve, i.e. the “triple point”, has a 10% avoidance rate with the rated speed of 1500 rpm, which is significantly improved compared to 6.6% before optimization. Figure 21 Information calculation shows that the avoidance rate of the speed corresponding to the intersection of the optimized frequency-multiplication curve K=7 and the 7-pitch diameter curve, namely the "triple point", and the rated speed of 1500 rpm increased from -2.67% to -6.1%, successfully avoiding the dangerous area.

[0114] In building a multi-objective parameter optimization system, we focused on implementing optimization capabilities based on design of experiments (DOE) and response surface methodology, combining them with MOGA (multi-objective genetic algorithm) to achieve efficient multi-objective optimization. The system interface primarily consists of a DOE experimental design area, a surrogate model construction area, an optimization objective and constraint area, a multi-objective optimization module, and an optimization results area. In the DOE experimental design area, users can set upper and lower bounds for design variables and select optimization objectives. After the experimental design is completed, the system automatically generates a response surface model and displays a response surface plot and a fit evaluation chart in the surrogate model construction area to help users evaluate the accuracy of the surrogate model.

[0115] In the Optimization Objectives and Constraints area, users can set optimization objectives and constraints, such as minimizing stress and maximizing natural frequency. The multi-objective optimization module allows users to set control parameters such as the number of iterations and population size for the optimization algorithm. After optimization is complete, the interface displays the optimal design parameters, along with their corresponding model-predicted values ​​and actual solution response values, in the Optimization Results area. MOGA was selected as the multi-objective optimization algorithm due to its high efficiency and robustness in handling multi-objective optimization problems. By constructing a response surface model, the system reduces the number of direct simulation analysis calls, significantly improving optimization efficiency. Optimization results can be exported as data files for subsequent analysis and application. Through scripting and API calls, the system automates the entire process from experimental design to optimization solution, ensuring the efficiency and accuracy of the optimization process. After optimization is complete, the system compares the optimization results with the actual simulation results to verify the optimization effect and ensure the reliability and practicality of the optimal design parameters. The system's automated process and efficient performance make it valuable for the multi-objective optimization design of long last-stage steam turbine blades.

[0116] In summary, the product integrated modeling method and system based on multi-system and multi-level collaboration provided by the present invention aims to solve the technical problems of low precision and low reliability of flexible design of complex system structures in the mechanical flexible design method in the prior art. By generating a preliminary physical model of the product to be modeled, modifying the modifiable feature parameters, updating the physical model of the product to be modeled, constructing a finite element analysis model for static analysis and modal analysis, creating an optimization data model based on model design variables, optimization objectives and constraints, and solving multiple optimization objectives to obtain the optimal design variables, the optimal design variables are applied to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled, thereby achieving effective optimization of design parameters and improving design speed and design accuracy.

Claims

1. A product integration modeling method based on multi-system and multi-level collaboration, characterized by: The following steps are involved: S1: Call the model data file based on the product to be modeled, set the key parameters required for creation, and generate the physical model of the product to be modeled; S2: Extract input parameters and parameterize them, generate modifiable feature parameters in the interactive interface, modify the modifiable feature parameters, and update the physical model of the product to be modeled; S3: Pre-process the updated solid model of the product to be modeled, build a finite element analysis model for static analysis and modal analysis, and finally complete the solution and export the analysis result report; S4: Set model design variables, optimization objectives, and constraints, and create an optimization data model based on the model design variables, optimization objectives, and constraints; S5: Based on the optimization goal, a preset multi-objective optimization algorithm is used to search the solution space of the optimization data model, an optimal solution is obtained from the solution space, optimal design variables are obtained, and the optimal design variables are applied to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

2. The product integrated modeling method based on multi-system and multi-level collaboration according to claim 1 is characterized in that: In step S1, the specific process of calling the model data file based on the product to be modeled, setting the key parameters required for creation, and generating the physical model of the product to be modeled is as follows: S11: Input the specified parameters and start the modeling task; S12: Read the file containing the coordinate data of the cross-section outline points of key parts, create features using the secondary development interface function, and add specified constraints to generate a three-dimensional solid model.

3. The product integrated modeling method based on multi-system and multi-level collaboration according to claim 1 is characterized in that: The specific process of preprocessing the updated solid model of the product to be modeled in step S3 is: meshing, adding materials, setting contacts, adding loads, adding constraints, and setting boundary conditions for the updated solid model of the product to be modeled.

4. The product integrated modeling method based on multi-system and multi-level collaboration according to claim 3 is characterized in that: In step S3, the static analysis of the finite element analysis model includes axial stress analysis and deformation analysis, and the modal analysis includes constant frequency analysis and vibration analysis, and a Campbell diagram is generated.

5. The product integrated modeling method based on multi-system and multi-level collaboration according to claim 1 is characterized in that: The product to be modeled is a long blade at the last stage of a steam turbine. The design variables include the tie bar height, tie bar angle, and shroud thickness. The constraints include the equivalent stress and maximum deformation value. The formula for the established optimization data model is as follows, with each triple resonance frequency as the objective function: ； in, X is the design variable, P 1 represents the stretching height, P 2 represents the stretching angle of the tendons. P 3 represents the thickness of the belt, f 1( X ) represents the second-order, third-pitch vibration frequency of the blade over the entire cycle at rated speed. f 2( X ) is the 3rd order 7th pitch vibration frequency of the blade at rated speed. 、 Represent the weights of the three optimization objective functions, P i,min and P i,max represent the lower and upper limits of the design variables, respectively. G 1( X ) represents the 4th order 8th pitch frequency of the blade at rated speed. In order to avoid the resonance triple point generated by the intersection of the 4th order 8th pitch frequency curve and the K=8th frequency line, G 1( X )≥206Hz, G 2( X ) represents the maximum equivalent stress of the blade, G 2( X )<850MPa, G 3( X ) represents the maximum equivalent stress of the impeller, G 3( X )<760MPa.

6. The product integrated modeling method based on multi-system and multi-level collaboration according to claim 5 is characterized in that: The optimization objectives are set as follows: Set the 2nd order 3rd pitch frequency f 1( X ) is optimized to maximize the frequency of the 3rd order and 7th node diameter. f 2( X ) is optimized to minimize.

7. A product integrated modeling system based on multi-system and multi-level collaboration, used to implement the product integrated modeling method based on multi-system and multi-level collaboration according to any one of claims 1 to 6, characterized in that: It includes a design resource database, a product model generation module, a parameterization and simulation analysis module, and a parameter optimization model. The parameterization and simulation analysis module includes a preprocessing module and an automatic analysis module. The design resource database is connected to the product model generation module, the product model generation module is connected to the parameterization and simulation analysis module, and the parameterization and simulation analysis module is connected to the parameter optimization model. The product model generation module is used to retrieve the model data file from the design resource database, set the key parameters required for creation, and generate the physical model of the product to be modeled; The product model generation module is used to extract input parameters and parameterize them, generate modifiable feature parameters in the interactive interface, modify the modifiable feature parameters, and update the physical model of the product to be modeled; The preprocessing module is used to preprocess the updated entity model of the product to be modeled; The automatic analysis module is used to construct a finite element analysis model to perform static analysis and modal analysis, and finally complete the solution and export the analysis result report; The parameter optimization model is used to set model design variables, optimization objectives and constraints, create an optimization data model based on the model design variables, optimization objectives and constraints, and based on the optimization objectives, preset a multi-objective optimization algorithm to search the solution space of the optimization data model, obtain the optimal solution from the solution space, obtain the optimal design variables, and apply the optimal design variables to the physical model of the product to be modeled to obtain the final physical model of the product to be modeled.

8. The product integrated modeling system based on multi-system and multi-level collaboration according to claim 7 is characterized in that: The parameter optimization model includes an experimental design module, an acquisition response module, an agent module and a multi-objective optimization module, wherein the experimental design module is connected to the acquisition response module, the acquisition response module is connected to the agent module, and the agent module is connected to the multi-objective optimization module; The experimental design module is used to select parameters, design methods and generate experimental points; The response acquisition module is used to update the model and obtain simulation results; The agent module is used to generate a response surface and evaluate effectiveness; The multi-objective optimization module is used to obtain the best design parameter combination by optimizing parameters and setting optimization algorithms.

Images