Altail HyperWorks automatic batch processing method suitable for conceptual design of stiffened plate

Abstract

The invention relates to the technical field of automation, in particular to an Altail HyperWorks automatic batch processing method suitable for the conceptual design of stiffened plates. The method comprises the following steps that geometric model import, grid division and boundary and load condition setting are achieved through a Tcl script command, parameters are initialized, and Altail HyperMesh pretreatment automation is carried out; after the pretreatment automation is completed, Altail OptiStrat solution automation is carried out, an SMP parallel mode and an in-core memory management mode are adopted to process files in batches, finite element analysis is carried out, and an optimal design variable is found through a loop optimization process; and importing an Altail Optistruct solving result into the HyperView to carry out Altail HyperView post-processing automation, and loading and storing a final density map of the topological structure unit of the stiffened plate. According to the design of the invention, the Altail HyperWorks automation technology is applied to the field of the conceptual design of the stiffened plate for the first time, and the process of pretreatment-solution-post-treatment is realized; and penalty factor optimization and stiffened plate panel optimization operation are realized based on the Tcl script so as to obtain the optimal stiffened plate topological layout.

Description

Technical Field

[0001] The present invention relates to the field of automation technology, and more specifically, to an Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates. Background Art

[0002] Stiffened plate structures are widely used in the fields of offshore platforms, ships, and aerospace due to their characteristics such as light weight and high strength. Topology optimization technology can perform rapid and high-performance design of structures based on mathematical theories. Therefore, it is becoming increasingly popular to conduct conceptual design of stiffened plates based on topology optimization technology. Currently, there are two problems in the research on stiffened plate topology optimization: First, only the optimization of the material layout of the stiffeners is considered while ignoring the optimization of the panel thickness; Second, in the research on stiffened plate topology optimization based on the SIMP method, the most critical penalty factor value is a fixed value, ignoring the necessity of optimizing the penalty factor. Therefore, an Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates is provided. Summary of the Invention

[0003] The purpose of the present invention is to provide an Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates, so as to solve the problems in the above background art that traditional stiffened plate topology optimization research only considers the optimization of the material layout of the stiffeners while ignoring the optimization of the panel thickness and ignoring the necessity of optimizing the penalty factor.

[0004] To achieve the above purpose, the present invention provides an Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates, including the following steps:

[0005] S1. Use Tcl script commands to implement the import of geometric models, mesh generation, setting of boundary and load conditions, and initialize parameters to automate the preprocessing of Altair HyperMesh;

[0006] S2. After the preprocessing automation is completed, perform Altair OptiStruct solution automation, adopt the SMP parallel mode and in-core memory management mode to batch process files, conduct finite element analysis, and find the optimal design variables through a loop optimization process;

[0007] S3. Import the Altair OptiStruct solution results into HyperView to automate the postprocessing of Altair HyperView, load and save the final unit density map of the stiffened plate topology structure.

[0008] As a further improvement of this technical solution, in S1, Tcl script commands are used to implement the import of geometric models, mesh generation, boundary and load condition settings, and initialize parameters for Altair HyperMesh preprocessing automation, including the following steps:

[0009] S1.1. Use the GUI operation of HyperMesh and generate corresponding Tcl script commands according to the Command file;

[0010] S1.2. Use Tcl script commands to import the geometric model of the stiffened plate structure to be analyzed into HyperMesh;

[0011] S1.3. Perform mesh generation operations on the imported geometric model;

[0012] S1.4. Set boundary conditions and apply loads according to the actual working conditions;

[0013] S1.5. Set the design variables, objective function, and constraint conditions required during the optimization process;

[0014] S1.6. Set the initial panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting the optimization process;

[0015] S1.7. Through loop judgment, determine whether the panel thickness h is less than or equal to the maximum allowable value of the panel thickness h max and whether the penalty factor p is less than or equal to the maximum allowable value of the penalty factor p max .

[0016] As a further improvement of this technical solution, in S1.7, through loop judgment, determine whether the panel thickness h is less than or equal to the maximum allowable value of the panel thickness h max and whether the penalty factor p is less than or equal to the maximum allowable value of the penalty factor p max , including the following steps:

[0017] S1.71. Determine whether the panel thickness h is less than or equal to the maximum allowable value of the panel thickness h max , if satisfied, proceed to step S1.72, if not satisfied, end the preprocessing automation process;

[0018] S1.72. Determine whether the penalty factor p is less than or equal to the maximum allowable value of the penalty factor p max , if satisfied, output the ".fem" file supported by Altair OptiStruct based on Tcl commands, name it in the form of h-p, and then proceed to step S1.73, if not satisfied, initialize p = p1 based on Tcl commands and modify h = h + Δh, and then return to step S1.71, where Δh represents the step size of the panel thickness;

[0019] S1.73. Modify p = p + Δp based on the Tcl command and return to step S1.72, where Δp represents the step size of the penalty factor.

[0020] As a further improvement of this technical solution, in S2, after the pre-processing is completed automatically, the Altair OptiStruct solution is automated, and the SMP parallel mode and in-core memory management mode are used to batch process files, including the following steps:

[0021] S2.1. Call the solver to batch input all ".fem" files and use the SMP parallel mode and in-core memory management mode;

[0022] S2.2. Set the initial panel thickness h = h1 and the penalty factor value p = p1;

[0023] S2.3. Determine whether the panel thickness h is less than or equal to the maximum allowable value h of the panel thickness max , if satisfied, continue to the next step, otherwise, end the solution automation process;

[0024] S2.4. Select the corresponding ".fem" file for analysis according to the current panel thickness h and penalty factor p.

[0025] As a further improvement of this technical solution, in S2, finite element analysis is performed and the optimal design variables are found through a cyclic optimization process, including the following steps:

[0026] S2.5. Set the design variables, objective function, and constraint conditions for the finite element analysis;

[0027] S2.6. Run the finite element analysis to evaluate the structural performance under the current design variables;

[0028] S2.7. If the analysis result converges, perform step S2.12, if not, continue the optimization process;

[0029] S2.8. Perform a sensitivity analysis on the design variables to understand the influence degree of each design variable on the final result;

[0030] S2.9. Based on the results of the sensitivity analysis, construct an approximation model to predict the performance of different design schemes;

[0031] S2.10. Optimize according to the approximation model to find the combination of design variables that makes the objective function reach the optimum;

[0032] S2.11. Update the design variables and return to step S2.6 to re-perform the finite element analysis until the optimal design variables are found;

[0033] S2.12. After each iteration, check whether both the panel thickness h and the penalty factor p have reached their maximum values. If the upper limits are not reached, adjust the panel thickness h and the penalty factor p, and repeat the above process until the analysis within all preset parameter ranges is completed.

[0034] As a further improvement of this technical solution, in S2.8, perform a sensitivity analysis on the design variables to understand the influence degree of each variable on the final result, including the following steps:

[0035] S2.81. Select the adjoint variable algorithm as the sensitivity analysis method, and in view of the problem that the influence of design variables on the objective function is unbalanced, introduce a weight coefficient α in the adjoint variable algorithm to dynamically adjust the sensitivities of different design variables;

[0036] S2.82. Use the built-in function of OptiStruct to start the sensitivity analysis;

[0037] S2.83. After the analysis is completed, extract the sensitivity data from the output file generated by OptiStruct;

[0038] S2.84. Analyze and understand the sensitivity values of each design variable.

[0039] As a further improvement of this technical solution, in S2.81, the adjoint variable algorithm is:

[0040]

[0041] where C represents the constraint condition; x e represents the design variable; U represents the adjoint vector; K represents the stiffness matrix; represents the partial derivative of the stiffness matrix K with respect to the design variable x e ; T represents the transpose operation;

[0042] In view of the problem that the influence of design variables on the objective function is unbalanced, introduce a weight coefficient α in the adjoint variable algorithm to dynamically adjust the sensitivities of different design variables:

[0043]

[0044] As a further improvement of this technical solution, in S2.9, based on the results of the sensitivity analysis, construct an approximate model for quickly predicting the performance of different design schemes, including the following steps:

[0045] S2.91. After performing the sensitivity analysis, use the results of the sensitivity analysis to construct an approximate model;

[0046] S2.92. Define the optimization objectives and constraint conditions, and determine the design variables and the boundary conditions of the design variables;

[0047] S2.93. Select the gradient descent method as the optimization algorithm and configure the parameters of the gradient descent method;

[0048] S2.94. Use the gradient descent method to start iterative search for the best combination of design variables based on the approximate model. For each iteration, use the approximate model to predict the objective function value according to the current combination of design variables.

[0049] As a further improvement of this technical solution, in S3, import the Altair OptiStruct solution results into HyperView for Altair HyperView post-processing automation, including the following steps:

[0050] S3.1. Initialize the panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting post-processing;

[0051] S3.2. Determine whether the panel thickness h is less than or equal to the maximum allowable value h max of the panel thickness. If satisfied, continue with step S3.3; otherwise, end the post-processing automation process;

[0052] S3.3. Release all previously existing object handles.

[0053] As a further improvement of this technical solution, in S3, load and save the final ribbed plate topology unit density map, including the following steps:

[0054] S3.4. Use Tcl script commands to load the analysis result file generated by Altair OptiStruct and the corresponding unit density file;

[0055] S3.5. Set the model currently being processed through Tcl commands;

[0056] S3.6. Select the working condition according to actual needs;

[0057] S3.7. Obtain and record the total number of iterations performed throughout the optimization process;

[0058] S3.8. Display the unit density distribution map of the ribbed plate structure obtained in the last iteration;

[0059] S3.9. Take a screenshot of the finally displayed unit density map and name and save it in the format of "panel thickness - penalty factor".

[0060] Compared with the prior art, the beneficial effects of the present invention:

[0061] In the Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates, the Altair HyperWorks automation technology is applied to the field of conceptual design of stiffened plates for the first time, realizing the process of preprocessing - solving - postprocessing; based on Tcl scripts, the optimization of penalty factors and the optimization of the stiffened plate panel are realized to obtain the best topological layout of the stiffened plate, while saving a large amount of manpower and material resources and avoiding the human error problems caused by a large number of repeated operations; the solution adopts the SMP parallel mode and the in-core memory management mode, thus greatly improving the calculation efficiency. Brief Description of the Drawings

[0062] Figure 1 It is the overall method flow chart of the present invention;

[0063] Figure 2 It is the schematic diagram of the Altair HyperMesh preprocessing automation process of the present invention;

[0064] Figure 3 It is the schematic diagram of the Altair OptiStruct solution automation process of the present invention;

[0065] Figure 4 It is the schematic diagram of the Altair HyperView postprocessing automation process of the present invention;

[0066] Figure 5 It is the Altair HyperView function structure tree of the present invention;

[0067] Figure 6 The geometric model of the stiffened plate for verifying the effectiveness of the present invention;

[0068] Figure 7 It is the schematic diagram of the optimization of the penalty factor and the panel thickness of the present invention;

[0069] Figure 8 It is the schematic diagram of the topological layout of the stiffened plate under the best panel thickness of the present invention. Detailed Embodiments

[0070] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0071] Embodiment: Please refer to Figure 1-8 As shown, this embodiment provides an Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates, including the following steps:

[0072] S1. Use Tcl script commands to achieve geometric model import, mesh generation, boundary and load condition setting, and initialize parameters such as panel thickness and penalty factor for Altair HyperMesh pre - processing automation (the Altair HyperMesh pre - processing automation process is as Figure 2 shown);

[0073] The high - performance finite - element pre - processor Altair HyperMesh provides a very rich set of API functions in the form of Tcl, which can be roughly divided into Tcl GUI functions, Tcl Modify functions, Tcl Query functions, and macro menu functions. Altair HyperMesh writes to the Command file in real - time during operation, recording the GUI operation history. This file is a standard ASCII - code file and is automatically generated in the starting directory of Altair HyperMesh; Altair HyperWorks is widely used in the field of structural optimization design due to its comprehensive and open architecture. The Altair HyperWorks software package includes Altair HyperMesh, Altair OptiStruct, and Altair HyperView, corresponding to pre - processing, solving, and post - processing respectively. Conducting topology optimization research on stiffened - plate structures based on Altair HyperWorks and optimizing panel thickness and penalty factor will involve a large number of cases. And for each case of topology optimization research, pre - processing, solving, and post - processing processes are required. A complete design work may need to go through this process hundreds or even thousands of times. Manual repetitive operations are not only time - consuming and laborious but also prone to human errors. The automated batch - processing technology is a knowledge - based engineering platform that greatly improves productivity, improves quality, and creates a sustainable competitive advantage by automating the structural design process. In addition, this technology allows engineering designers to focus on more important issues related to structural topology optimization, such as topology structure performance, topology structure "0 - 1" distribution, etc.

[0074] The development of the Altair HyperWorks automation system can rely on a variety of tools and languages, including ProcessManager, Process Studio, the Tcl / Tk language, the HTML language, and the functional functions provided by HyperWorks. The secondary development of Altair HyperWorks uses the Tcl / Tk language as the standard development language and provides it with a rich set of API (Application Program Interface) command functions in the form of Tcl. Based on Tcl commands to call API functions, the software can be driven to perform corresponding operations, realizing the customization process of the secondary development of Altair HyperWorks.

[0075] In this embodiment, the Tcl script commands are used to implement the import of geometric models, mesh generation, boundary and load condition settings, and the initialization of parameters such as panel thickness and penalty factor for the pre-processing automation of Altair HyperMesh, including the following steps:

[0076] S1.1. Use the GUI operations of HyperMesh and generate the corresponding Tcl script commands according to the Command file (a standard ASCII file that records all GUI operation histories);

[0077] S1.2. Use the Tcl script commands to import the geometric model of the stiffened plate structure to be analyzed into HyperMesh. This step is the prerequisite for finite element analysis;

[0078] S1.3. Perform mesh generation operations on the imported geometric model to ensure that the model is suitable for numerical simulation. This step is crucial for ensuring the accuracy of the analysis results;

[0079] S1.4. Set the boundary conditions and apply loads according to the actual working conditions. These conditions will directly affect the calculation results of the structural response;

[0080] S1.5. Set the design variables (such as panel thickness), objective functions (such as minimizing strain energy), and constraint conditions (such as material volume fraction constraints or convergence tolerances) required during the optimization process to prepare for subsequent topology optimization;

[0081] S1.6. Set the initial panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting the optimization process;

[0082] S1.7. Through loop judgment, determine whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max and whether the penalty factor p is less than or equal to the maximum allowable penalty factor p max ;

[0083] Among them, by cyclically judging whether the panel thickness h is less than or equal to the maximum allowable value h of the panel thickness max and whether the penalty factor p is less than or equal to the maximum allowable value p of the penalty factor max , the following steps are included:

[0084] S1.71. Judge whether the panel thickness h is less than or equal to the maximum allowable value h of the panel thickness max . If it is satisfied, go to step S1.72; if not, end the pre-processing automation process;

[0085] S1.72. Judge whether the penalty factor p is less than or equal to the maximum allowable value p of the penalty factor max . If it is satisfied, output the ".fem" file supported by Altair OptiStruct based on the Tcl command, name it in the form of h-p, and then go to step S1.73; if not, initialize p = p1 based on the Tcl command, modify h = h + Δh, and then return to step S1.71, where Δh represents the step size of the panel thickness;

[0086] S1.73. Modify p = p + Δp based on the Tcl command and return to step S1.72, where Δp represents the step size of the penalty factor.

[0087] S2. After the pre-processing automation is completed, perform Altair OptiStruct solution automation (the Altair OptiStruct solution automation process is as Figure 3 shown), adopt the SMP parallel mode and in-core memory management mode to batch process multiple ".fem" files, perform finite element analysis, and find the optimal design variables through a loop optimization process;

[0088] The Altair OptiStruct solver can batch process ".fem" files with high performance, thus well solving the problem of solving a large number of stiffened plate topology optimization examples;

[0089] In this embodiment, after the pre-processing automation is completed, perform Altair OptiStruct solution automation, adopt the SMP parallel mode and in-core memory management mode to batch process multiple ".fem" files, including the following steps:

[0090] S2.1. Call the solver to batch input all ".fem" files and adopt the SMP parallel mode and in-core memory management mode to improve the calculation efficiency;

[0091] S2.2. Set the initial panel thickness h = h1 and the penalty factor value p = p1;

[0092] S2.3. Determine whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max , if satisfied, proceed to the next step, otherwise, end the automated solution process;

[0093] S2.4. Select the corresponding ".fem" file for analysis based on the current panel thickness h and the penalty factor p.

[0094] Furthermore, finite element analysis (FEA) is an engineering simulation technique used to predict the behavior of structures or materials under various physical conditions. It works by decomposing a complex system into a finite number of smaller, simpler parts - called "finite elements", and then using mathematical models to simulate the behavior of each element, and then synthesizing the responses of these elements to understand the performance of the entire system;

[0095] Conduct finite element analysis and find the optimal design variables through an iterative optimization process, including the following steps:

[0096] S2.5. Set the design variables (panel thickness and penalty factor), the objective function (in this embodiment, the objective function is the minimization of strain energy (corresponding to the maximization of structural stiffness), set in the optimization problem through a Tcl script), and the constraint conditions (in this embodiment, the total volume of the optimized material does not exceed a specified ratio of the initial volume, and the convergence tolerance is specified in the solver parameters) for the finite element analysis;

[0097] S2.6. Run the finite element analysis to evaluate the structural performance under the current design variables. The process of running the finite element analysis to evaluate the structural performance under the current design variables first involves defining the structural model based on the design variables (such as panel thickness, penalty factor, etc.), and applying the boundary conditions and loads of the actual working conditions. Then, discretize the structure into multiple finite elements, establish a mathematical model for each element based on physical laws (such as mechanical laws), forming a series of algebraic equations. By solving these equations, calculate the responses (such as displacements) of each node, and then deduce the stress, strain distribution and other performance indicators of the entire structure, so as to evaluate whether the performance of the structure under specific conditions meets the design requirements;

[0098] S2.7. If the analysis result converges, proceed to step S2.12, if not, continue the optimization process;

[0099] S2.8. Conduct a sensitivity analysis of the design variables to understand the influence degree of each design variable on the final result;

[0100] Among them, conducting a sensitivity analysis of the design variables to understand the influence degree of each variable on the final result includes the following steps:

[0101] S2.81. Select the adjoint variable algorithm as the sensitivity analysis method. The adjoint variable algorithm is a mathematical method used to calculate the sensitivities of the objective function and constraints with respect to the design variables. By solving the adjoint equation, it can efficiently obtain this sensitivity information to support the optimization design process. The adjoint variable algorithm can calculate the sensitivities of all design variables to the objective function and constraints in an additional solution process, which significantly reduces the computational amount compared with the direct difference method, especially when the number of design variables is large.

[0102] The adjoint variable algorithm is as follows:

[0103]

[0104] where C represents the constraint condition or constraints; x e represents the design variable; U represents the adjoint vector, which is obtained by solving the adjoint equation. The introduction of the adjoint vector is to simplify the derivative calculation process, enabling the sensitivities of all design variables to be obtained with fewer additional calculations; K represents the stiffness matrix, which is a matrix describing the rigidity of the structure. In finite element analysis, it is determined according to the geometric shape, material properties, and boundary conditions of the structure. represents the partial derivative of the stiffness matrix K with respect to the design variable x e , which reflects the change in the stiffness matrix when a certain design variable is changed; T represents the transpose operation.

[0105] In the optimization of stiffened plates, the sensitivities of different design variables (such as panel thickness and element density) to the objective function (strain energy) may vary significantly. The direct impact of panel thickness change on stiffness is more significant. The element density (topological variable) indirectly affects the material distribution through the penalty factor, and the sensitivity may be lower. The optimization process may be overly biased towards high-sensitivity variables, resulting in ineffective optimization of other variables. The update step size of variables with low sensitivity is small, leading to slow convergence or getting stuck in a local optimum. By adjusting the weight coefficient, the optimization algorithm can approach the global optimum solution faster. For example: assign a higher weight to key variables (such as panel thickness) and quickly converge to the optimal thickness value.

[0106] To address the problem of unbalanced influence of design variables on the objective function, a weight coefficient α is introduced in the adjoint variable algorithm to dynamically adjust the sensitivities of different design variables:

[0107]

[0108] S2.82. Use the built-in function of OptiStruct to initiate the sensitivity analysis.

[0109] S2.83. After the analysis is completed, extract the sensitivity data from the output file generated by OptiStruct. These data describe the magnitude of the impact of each design variable on the objective function and constraints.

[0110] S2.84. Analyze and understand the sensitivity values of each design variable to identify the variables that have a significant impact on the objective function and constraints.

[0111] S2.9. Based on the results of the sensitivity analysis, construct an approximation model for quickly predicting the performance of different design schemes. The approximation model is a simplified mathematical model that quickly predicts the performance of different design schemes based on limited high-precision simulation or experimental data, thereby significantly reducing the calculation time and cost while ensuring a certain level of accuracy. The process of constructing an approximation model for quickly predicting the performance of different design schemes first involves understanding the impact of each design variable on the objective function based on existing data or through sensitivity analysis, and then using this information to establish a simplified model, such as a response surface model or a surrogate model. This approximation model can significantly reduce the calculation time while maintaining sufficient accuracy, enabling the rapid evaluation of the expected performance under a large number of design variable combinations, thus supporting a more efficient optimization decision-making process.

[0112] Among them, based on the results of the sensitivity analysis, constructing an approximation model for quickly predicting the performance of different design schemes includes the following steps:

[0113] S2.91. After performing the sensitivity analysis, use the sensitivity analysis results to construct or update the approximation model. The approximation model is a simplified model used to quickly predict the objective function values for different design variable combinations.

[0114] S2.92. Clearly define the optimization objectives and constraints, and determine the design variables and their boundary conditions (such as the upper and lower limits of the panel thickness h).

[0115] S2.93. Select the gradient descent method as the optimization algorithm and configure the specific parameters of the gradient descent method, such as the step size and convergence criterion. The gradient descent method is an optimization algorithm that iteratively updates the parameters in the opposite direction of the gradient of the objective function to minimize or maximize the value of the objective function.

[0116] S2.94. Use the gradient descent method to start iterative searching for the best design variable combination based on the approximation model. For each iteration, use the approximation model to predict the objective function value according to the current design variable combination.

[0117] S2.10. Optimize according to the approximate model to find the combination of design variables that makes the objective function reach the optimum. In the process of optimizing according to the approximate model, first, use the approximate model to quickly evaluate the objective function values of different combinations of design variables. Then, through an optimization algorithm (gradient descent method), iteratively search the design space and gradually adjust the design variables to find the combination that makes the objective function reach the optimum. This process takes advantage of the fast prediction ability of the approximate model, effectively reducing the computational cost and accelerating the optimization process;

[0118] S2.11. Update the design variables and return to step S2.6 to re - execute the finite element analysis until the best design variables are found or the convergence criteria are met;

[0119] S2.12. After each iteration, check whether both the panel thickness h and the penalty factor p have reached their maximum values. If not, adjust the panel thickness h and the penalty factor p (judge whether the penalty factor p is less than or equal to the maximum allowable value of the penalty factor p max . If satisfied, return to step S2.4. If not, initialize p = p1 based on the Tcl command and modify h = h + Δh, and repeat the above process until all analyses within the preset parameter range are completed.

[0120] S3. Import the Altair OptiStruct solution results into HyperView for Altair HyperView post - processing automation (the Altair HyperView post - processing automation process is as Figure 4 shown), load and save the final ribbed panel topology structure element density map;

[0121] In this embodiment, different from Altair HyperMesh discussed above, the all - capable and general post - processor Altair HyperView cannot obtain GUI operation commands through a Command file. Its secondary development relies only on the function structure tree (the function structure tree is as Figure 5 shown). It is necessary to extract the handle of the object according to this function structure tree to implement software operations. All commands are centered around the object and strictly access the handles of the specified object level by level according to the function structure tree. The general syntax format of the command is:

[0122] {parent_object_handle_name}{command}{object_handle_name};

[0123] Among them, parent_object_handle_name represents the name of the parent object handle; command is the API command to obtain the handle of the accessed child object; object_handle_name is the name of the handle of the accessed child object.

[0124] Import the solution results of Altair OptiStruct into HyperView for Altair HyperView post - processing automation, including the following steps:

[0125] S3.1. Initialize the panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting the post - processing.

[0126] S3.2. Determine whether the panel thickness h is less than or equal to the maximum allowable value h max of the panel thickness. If it is satisfied, continue with step S3.3; otherwise, end the post - processing automation process.

[0127] S3.3. Before starting the processing, release all the previously existing object handles (Handle) to ensure that subsequent operations will not be interfered with.

[0128] Furthermore, load and save the final ribbed plate topology structure unit density map, including the following steps:

[0129] S3.4. Use Tcl script commands to load the analysis result file generated by Altair OptiStruct (the analysis result file is in the.fem format), as well as the corresponding unit density file, which contains the result information of the topology optimization.

[0130] S3.5. Set the currently processed model through Tcl commands (referring to the target dataset model set through Tcl script commands in Altair HyperView for subsequent operations (such as showing the structural response in a specific state or generating a unit density distribution map, etc.). It refers to the ribbed plate structure model after topology optimization, which is the result obtained through finite element analysis based on the previously defined design variables (such as panel thickness, penalty factor), objective function, and constraint conditions) so that subsequent operations can be carried out for the correct dataset.

[0131] S3.6. Select or define specific working conditions according to actual needs, such as load conditions, boundary conditions, etc., for showing and analyzing the structural response in a specific state.

[0132] S3.7. Obtain and record how many iterations have been carried out in total during the entire optimization process, which helps to understand the development trend of the optimization process.

[0133] S3.8. Display the element density distribution diagram of the stiffened plate structure obtained in the last iteration to visually present the optimized design effect;

[0134] S3.9. Take a screenshot of the finally displayed element density diagram and save it with the name in the format of "panel thickness - penalty factor" for convenient subsequent reference and report use.

[0135] In this embodiment, a further description is given to an Altair HyperWorks automated batch processing method provided by the present invention for the conceptual design of stiffened plates:

[0136] Adopt a two-dimensional eccentric stiffened plate model and conduct topology optimization design with minimum strain energy for it. Its geometric model is as Figure 6 shown, where B = 200, L = 200, H = 20. The four sides of the stiffened plate are simply supported and subjected to a surface pressure of 1. The elastic modulus of the material E = 2.06×10 5 , Poisson's ratio v = 0.3, and material density ρ = 7.85×10 9 . The mesh elements adopt four-node two-dimensional plate elements with a size of 1×1. The material volume fraction volfrac is 0.3. The convergence tolerance w0 = E -4 to obtain a convergent and 0-1 distributed stiffener topology layout. Based on the present invention, optimize the panel thickness h, where the minimum value of h is E -4 to ensure that the panel is subjected to a complete surface pressure during the topology optimization process. The maximum value of h is 4, and the increment is 0.2. Optimize the penalty factor (p) for each panel thickness h, and the value range is [1, 5], with an increment of 0.05, totaling 1701 numerical examples. Each numerical example needs to perform preprocessing - solving - postprocessing operations. Based on the present invention, after 1701 processes, search for the best penalty factor value under each panel thickness and obtain the best panel thickness value on this basis, as Figure 7 shown, where the best panel thickness h opti , and the corresponding best penalty factor value p opti = 2.7. The topology layout distribution of the stiffened plate under the best panel thickness is as Figure 8 shown, with a clear layout and an obvious 0-1 distribution. Thus, it proves the effectiveness of the present invention in the field of conceptual design of stiffened plates.

[0137] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the descriptions in the specification are only preferred examples of the present invention and are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed.

Claims

1. An Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates, characterized in that, It includes the following steps: S1. Use Tcl script commands to achieve geometric model import, mesh generation, boundary and load condition setting, and initialize parameters for Altair HyperMesh pre-processing automation; S2. After the pre-processing automation is completed, perform Altair OptiStruct solution automation. Use the SMP parallel mode and in-core memory management mode to batch process files, perform finite element analysis, and find the optimal design variables through a loop optimization process; S3. Import the Altair OptiStruct solution results into HyperView for Altair HyperView post-processing automation, load and save the final ribbed panel topology unit density map.

2. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 1, characterized in that: In the above S1, using Tcl script commands to achieve geometric model import, mesh generation, boundary and load condition setting, and initialize parameters for Altair HyperMesh pre-processing automation includes the following steps: S1.

1. Use the GUI operation of HyperMesh and generate corresponding Tcl script commands according to the Command file; S1.

2. Use Tcl script commands to import the geometric model of the ribbed panel structure to be analyzed into HyperMesh; S1.

3. Perform mesh generation operations on the imported geometric model; S1.

4. Set boundary conditions and apply loads according to the actual working conditions; S1.

5. Set the design variables, objective function, and constraint conditions required during the optimization process; S1.

6. Set the initial panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting the optimization process; S1.

7. Determine by loop whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max and whether the penalty factor p is less than or equal to the maximum allowable penalty factor p max .

3. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 2, wherein: In S1.7, by circularly judging whether the panel thickness h is less than or equal to the maximum allowable value h of the panel thickness max and whether the penalty factor p is less than or equal to the maximum allowable value p of the penalty factor max , it includes the following steps: S1.

71. Determine whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max . If satisfied, proceed to step S1.72; if not, end the pre - processing automation process. S1.

72. Determine whether the penalty factor p is less than or equal to the maximum allowable value of the penalty factor p max , if satisfied, output the ".fem" file supported by Altair OptiStruct based on the Tcl command, name it in the h-p form, and then proceed to step S1.

73. If not satisfied, initialize p = p1 based on the Tcl command, modify h = h + Δh, and then return to step S1.71, where Δh represents the step size of the panel thickness; S1.

73. Modify p = p + Δp based on the Tcl command and return to step S1.72, where Δp represents the step size of the penalty factor.

4. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 3, characterized in that: In the above S2, after the pre-processing automation is completed, perform Altair OptiStruct solution automation. Use the SMP parallel mode and in-core memory management mode to batch process files, including the following steps: S2.

1. Call the solver to batch input all ".fem" files and use the SMP parallel mode and in-core memory management mode; S2.

2. Set the initial panel thickness h = h1 and the penalty factor value p = p1; S2.

3. Determine whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max , if satisfied, proceed to the next step, otherwise, end the automated solution process; S2.

4. Select the corresponding ".fem" file for analysis according to the current panel thickness h and penalty factor p.

5. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 4, characterized in that: In the above S2, perform finite element analysis and find the optimal design variables through a loop optimization process, including the following steps: S2.

5. Set the design variables, objective function, and constraint conditions for finite element analysis; S2.

6. Run finite element analysis to evaluate the structural performance under the current design variables; S2.

7. If the analysis result converges, proceed to step S2.

12. If it does not converge, continue the optimization process; S2.

8. Perform sensitivity analysis on the design variables to understand the influence degree of each design variable on the final result; S2.

9. Based on the results of the sensitivity analysis, construct an approximate model to predict the performance of different design schemes. S2.

10. Optimize according to the approximate model to find the combination of design variables that makes the objective function reach the optimum. S2.

11. Update the design variables and return to step S2.6 to re - execute the finite element analysis until the best design variables are found. S2.

12. After each iteration, check whether both the panel thickness h and the penalty factor p have reached their maximum values. If not, adjust the panel thickness h and the penalty factor p, and repeat the above process until the analysis within all preset parameter ranges is completed.

6. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 5, characterized in that: In S2.8, perform a sensitivity analysis on the design variables to understand the influence degree of each variable on the final result, including the following steps: S2.

81. Select the adjoint variable algorithm as the sensitivity analysis method, and in view of the problem that the influence of design variables on the objective function is unbalanced, introduce a weight coefficient α into the adjoint variable algorithm to dynamically adjust the sensitivities of different design variables. S2.

82. Use the built - in function of OptiStruct to start the sensitivity analysis. S2.

83. After the analysis is completed, extract the sensitivity data from the output file generated by OptiStruct. S2.

84. Analyze and understand the sensitivity values of each design variable.

7. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 6, characterized in that: In S2.81, the adjoint variable algorithm is: Among them, C represents the constraint condition; x e represents the design variable; U represents the adjoint vector; K represents the stiffness matrix; represents the partial derivative of the stiffness matrix K with respect to the design variable x e ; T represents the transpose operation; In view of the problem that the influence of design variables on the objective function is unbalanced, introduce a weight coefficient α into the adjoint variable algorithm to dynamically adjust the sensitivities of different design variables:

8. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 7, characterized in that: In S2.9, based on the results of the sensitivity analysis, construct an approximate model for quickly predicting the performance of different design schemes, including the following steps: S2.

91. After performing the sensitivity analysis, use the sensitivity analysis results to construct an approximate model. S2.

92. Define the optimization objectives and constraints, and determine the design variables and the boundary conditions of the design variables. S2.

93. Select the gradient descent method as the optimization algorithm and configure the parameters of the gradient descent method. S2.

94. Use the gradient descent method to start iterative search for the best combination of design variables on the basis of the approximate model. For each iteration, use the approximate model to predict the objective function value according to the current combination of design variables.

9. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 8, characterized in that: In S3, import the Altair OptiStruct solution results into HyperView for Altair HyperView post - processing automation, including the following steps: S3.

1. Initialize the panel thickness h = h1 and the penalty factor value p = p1 as the basic parameters for starting the post - processing. S3.

2. Determine whether the panel thickness h is less than or equal to the maximum allowable panel thickness h max . If it is satisfied, proceed to step S3.3; otherwise, end the post-processing automation process S3.

3. Release all previously existing object handles.

10. The Altair HyperWorks automated batch processing method applicable to the conceptual design of stiffened plates according to claim 9, characterized in that: In S3, load and save the final unit density map of the stiffened plate topology, including the following steps: S3.

4. Use the Tcl script command to load the analysis result file generated by Altair OptiStruct and the corresponding unit density file. S3.

5. Set the currently processed model through the Tcl command. S3.

6. Select the working condition according to the actual needs. S3.

7. Obtain and record the total number of iterations carried out during the entire optimization process. S3.

8. Display the unit density distribution map of the stiffened plate structure obtained in the last iteration. S3.

9. Take a screenshot of the finally displayed unit density map and save it with a name in the format of "panel thickness - penalty factor".

Images