System for optimizing T-shaped empennage layout structure dynamic scaling equivalent model

Abstract

The invention discloses a system for optimizing a T-shaped empennage layout structure dynamic scaling equivalent model, and belongs to the technical field of aircraft structure optimization. The system comprises an input and parameter definition module, a finite element model automatic construction module, an optimization algorithm execution module, and a result output and visualization module, and adopts a beam frame, an aviation laminate dimension shape frame and a glass fiber skin as standard structure templates, and the model is optimized on the basis of scale rule parameters such as size and density and first five-order modal target parameters. A refined finite element model is automatically generated, a hybrid optimization logic combining genetic algorithm global search and least square method local fine tuning is adopted, an optimization target is achieved, a design result is finally output, the dynamic fidelity and design efficiency of a scale model are greatly improved, manual dependence is reduced, and the design efficiency is improved. And reliable technical support is provided for aviation wind tunnel test and related structure design.

Description

Technical Field

[0001] This invention relates to a system for optimizing a scaled-down equivalent model of the dynamics of a T-tail configuration structure, belonging to the field of aircraft structural optimization technology. Background Technology

[0002] A T-tail is an aircraft tail configuration where the horizontal plane (tail and elevator) is mounted atop the vertical tail. This design protects the tail from engine exhaust, making it more widely used on aircraft with tail-mounted engines. Seaplanes and amphibious aircraft also typically employ T-tails to keep the horizontal plane as far away from the water as possible, reducing internal vibration and noise. The T-tail is a specialized aerodynamic structure. At lower angles of attack, this configuration allows the horizontal tail to avoid wing-draft influences, resulting in higher control efficiency. Furthermore, this configuration facilitates the opening of the rear fuselage and is beneficial for cargo transport. Therefore, many military transport aircraft designs incorporate T-tails.

[0003] However, the T-tail configuration, due to its high horizontal stabilizer layout, results in complex dynamic characteristics, particularly flutter and deep stall, posing a threat to flight safety. This is because, on the one hand, the horizontal stabilizer has a large mass and is connected to the top of the vertical stabilizer; therefore, even small movements of the vertical stabilizer can lead to significant in-plane and out-of-plane movements of the horizontal stabilizer. On the other hand, since the weight of the horizontal plane is located at the top of the vertical stabilizer, the resulting moment arm places a large load on the vertical stabilizer. Therefore, research on T-tails is of great significance for the development of large transport aircraft.

[0004] Due to the extreme operating conditions (load, temperature) faced by large transport aircraft, full-scale aircraft testing is impossible. Even when prototype testing is possible, the testing is expensive, time-consuming, and difficult to control. Therefore, it is necessary to design a scaled-down equivalent model with similar structural dynamic characteristics to the aircraft, and then conduct wind tunnel tests to simulate real aerodynamic loads. In particular, the T-tail configuration, due to its high horizontal stabilizer, has complex dynamic characteristics, making wind tunnel testing even more necessary to accurately verify its dynamic characteristics. Simulating the structural dynamic characteristics of the T-tail is the key to the design model. Summary of the Invention

[0005] To address the design requirements of existing wind tunnel scaled-down equivalent models of large transport aircraft with T-tails, a model optimization system is provided that can perform multi-objective optimization for frequencies and mode shapes. This system enables efficient and reliable design of scaled-down equivalent elastic models of aircraft, resulting in structural dynamic characteristics that more closely resemble those of a real aircraft and more reliable wind tunnel test data.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a system for optimizing the dynamic scaled-down equivalent model of the T-tail wing layout structure, including an input and parameter definition module, a finite element model automated construction module, an optimization algorithm execution module, and a result output and visualization module. These modules work together to achieve fully automated optimization from constraint input to production implementation.

[0007] The input and parameter definition module is used to receive the target parameters of the T-tail fin mode, the scaling rule parameters and the basic structural parameters, and generate the initial design constraint set.

[0008] The finite element model automated construction module, based on the initial design constraint set, calls the preset structural template library to automatically generate a three-dimensional digital model and the corresponding refined finite element model, and completes mesh generation and boundary condition application;

[0009] The optimization algorithm execution module adopts a hybrid optimization logic that combines genetic algorithm and least squares method to perform global search and local fine-tuning of the design variables of the finite element model, iterating until the preset convergence condition is met.

[0010] The results output and visualization module verifies the dynamic characteristics of the optimized model, outputs design documents, verification reports and optimization process records, and displays the verification results through visualization charts.

[0011] Furthermore, the templates in the preset structural template library adopt a combined structural form of beam frame, aviation shelf shaped frame and fiberglass skin, including standardized components such as beam frame, shaped wooden frame, balsa wood front and rear edges, foam top cover and fiberglass skin. The material properties and functional correspondence of each component are preset as follows: the beam frame bears the structural rigidity, and the aviation shelf shaped frame and fiberglass skin realize the shape.

[0012] Furthermore, the modal target parameters include the target frequencies and target mode shapes of the first 5 modes, and support the import of FEA model result files in ANSYS and Nastran formats; the scaling rule parameters include size scaling, density ratio, and dynamic pressure ratio, and the system automatically calculates the mass ratio and frequency ratio based on the scaling rule parameters; the structural foundation parameters include the material properties of elastic modulus and density of aluminum alloy, glass fiber, and aerospace laminate, and support custom component material configuration.

[0013] Furthermore, the refined finite element model adopts a hybrid modeling method of 1D beam elements, 2D surface elements and 3D solid elements; the boundary conditions are configured by default as a 6-DOF fixed constraint at the front end of the rear fuselage beam, and user-defined constraint scenarios are supported.

[0014] Furthermore, the specific implementation of the hybrid optimization logic is as follows: a genetic algorithm is used to initialize and generate a population of design variables, including structural parameters such as the width, height, web thickness, and flange thickness of the channel beam. The population range is configured according to preset upper and lower limits. A fitness function is constructed using the weighted sum of frequency error and modal confidence (MAC) value. Elite individuals are selected through selection, crossover, and mutation operations. The least squares method is used to call the Nielsen sensitivity analysis formula to calculate the sensitivity of elite individuals. Based on the calculation results, the design variables are fine-tuned. Dynamic sensitivity analysis is performed, and the sensitivity formula of frequency to structural parameters is:

[0015] ;

[0016] In the formula, and For the first Eigenvalues ​​and eigenvectors of order 1 It is the generalized stiffness matrix of the structure. It is the structural generalized mass matrix; For the first There are 1 design variable, where T is the matrix transpose.

[0017] Furthermore, the preset convergence condition is that the change in optimal fitness over 10 consecutive generations is ≤1e. -6 The default criteria for verifying the dynamic characteristics are: frequency error of the first 5 modes ≤ 2%, modal confidence MAC value ≥ 95%. If the default criteria are not met, the system supports readjusting the constraint parameters and restarting the optimization.

[0018] Beneficial Effects: This system comprises an input and parameter definition module, an automated finite element model construction module, an optimization algorithm execution module, and a result output and visualization module. Using beam frames, aerospace layered slab frames, and fiberglass skin as standard structural templates, and based on scaling parameters such as dimensions and density, along with the target parameters for the first five modal modes, it automatically generates a refined finite element model. Employing a hybrid optimization logic combining global search using a genetic algorithm and local fine-tuning using the least squares method, and supported by core algorithms such as sensitivity analysis formulas and MAC formulas, it achieves optimization targets of ≤2% error in the first five modal frequencies and ≥95% MAC value. The final design output significantly improves the dynamic fidelity and design efficiency of the scaled-down model, reduces reliance on manual intervention, and provides reliable technical support for aerospace wind tunnel testing and related structural design. By optimizing the beam frame section parameters of the scaled-down equivalent model of the T-tail and adjusting the model stiffness to meet the constraints of target frequencies and mode shapes, the design process of the scaled-down equivalent model of the T-tail dynamic is greatly accelerated. This system not only provides insights for multidisciplinary optimization of other similar structures but also serves as a reference for subsequent T-tail characteristic analysis and wind tunnel model design. Attached Figure Description

[0019] Figure 1This is a three-dimensional internal structure diagram of the present invention.

[0020] Figure 2 This is a three-dimensional overall structural diagram of the present invention.

[0021] Figure 3 This is a schematic diagram of the detailed finite element model of the structure of the present invention.

[0022] Figure 4 This is a framework diagram of the system of the present invention.

[0023] In the diagram: 1-beam frame, 2-wooden frame, 3-lightweight wood front and rear edges, 4-foam top cover, 5-fiberglass skin. Detailed Implementation

[0024] The present invention will be described below with reference to the accompanying drawings and specific embodiments:

[0025] Figure 1-4 The overall structural diagram of the present invention is shown. A scaled-down equivalent model optimization system for a T-tail fin layout structure determines initial conditions, including constraints such as the frequency and mode shape of the target mode of the T-tail fin, through initial design. A detailed structural model is constructed, and the detailed structural form of the scaled-down T-tail fin model is designed and determined. A three-dimensional digital model is drawn, and the internal detailed structure is shown below. Figure 1 As shown, the overall structure is as follows Figure 2 As shown;

[0026] The detailed structural form of the scaled-down T-tail model employs a beam frame structure to bear the model's stiffness, with aerospace-grade laminated frames and fiberglass skin for shaping. A detailed finite element dynamic model of the T-tail is established, such as... Figure 3 As shown.

[0027] The system is deployed in a three-layer architecture consisting of an input layer, a core function layer, and an output layer. Each module communicates through a standardized JSON format data interface. The input and parameter definition module and the result output and visualization module provide a graphical interactive interface based on PyQt. The finite element model automatic construction module and the optimization algorithm execution module run in the background and provide real-time status feedback through inter-process communication.

[0028] During the implementation of the input and parameter definition module, users input or import three types of core parameters through the interactive interface, and the system automatically completes format verification and parsing. Modal target parameters can be manually input with the target frequencies and target mode shape vectors of the first five modes, or FEA result files in ANSYS or Nastran format can be directly imported. The system automatically extracts the target parameters by parsing the eigenvalues ​​and eigenvectors in the file. Scaling rule parameters require input of the size scaling ratio, density ratio, and dynamic pressure ratio. The system automatically calculates the mass ratio and frequency ratio based on dynamic similarity theory. Structural foundation parameters provide a standard property library for aluminum alloy, fiberglass, and aerospace laminates by default. Users can customize or add other material types through the parameter editing interface. The system verifies the validity of the parsed parameters. After successful verification, it automatically generates an initial design constraint set containing parameter types, numerical ranges, and constraint priorities, and stores it in XML file format.

[0029] The automated finite element model construction module, based on the initial design constraint set, calls a pre-set structural template library to complete model construction. The templates in the library adopt a combined structural form of beam frame, aerospace-grade layered frame, and fiberglass skin, including standardized components such as beam frame, layered wooden frame, balsa wood front and rear edges, foam top cover, and fiberglass skin. Each component has pre-defined fixed material properties and functional correspondences. The beam frame provides structural stiffness, while the aerospace-grade layered frame and fiberglass skin provide the external shape. The module uses a hybrid modeling method of 1D beam elements, 2D surface elements, and 3D solid elements. It determines the optimal element size based on mesh sensitivity analysis results and automatically completes mesh generation, ensuring the number of nodes meets the requirements for refined modeling. The default boundary condition is a fixed 6-DOF constraint at the front end of the rear fuselage beam. Users can also customize constraint scenarios through an interactive interface to adapt to different experimental needs. After modeling is completed, the system automatically generates a 3D numerical model and a corresponding refined finite element model, providing a foundation for subsequent optimization calculations.

[0030] The optimization algorithm execution module employs a hybrid optimization logic combining genetic algorithms and least squares methods. The specific implementation process is as follows: First, a population containing design variables is generated using a genetic algorithm. These design variables include structural parameters such as the width, height, web thickness, and flange thickness of the channel beam. The population range is configured according to preset upper and lower limits. Then, a fitness function is constructed using a weighted sum of frequency error and modal confidence (MAC) values. The top 10% of elite individuals are selected through selection, crossover, and mutation operations. Next, the least squares method is used to calculate the sensitivity of the elite individuals. The Nielsen sensitivity analysis formula used is:

[0031] ;

[0032] In the formula, and For the first Eigenvalues ​​and eigenvectors of order 1 It is the generalized stiffness matrix of the structure. It is the structural generalized mass matrix; For the first There are 1 design variable, where T is the matrix transpose.

[0033] The design variables are fine-tuned locally based on the sensitivity calculation results to accelerate convergence. During the optimization process, the optimal fitness change ≤ 1e-6 over 10 consecutive generations is used as the preset convergence condition, and the calculation stops after iterating until this condition is met.

[0034] The results output and visualization module verifies the dynamic characteristics of the optimized model. The default verification criteria are: frequency error ≤2% for the first 5 modes and modal confidence (MAC) value ≥95%. If these criteria are not met, the system allows users to readjust constraint parameters and restart the optimization process. Upon successful verification, the module outputs design files, a verification report (including a frequency error comparison table, MAC value comparison graph, and mode shape comparison diagram, clearly presenting the optimization effect), and an optimization process record (covering iteration count-fitness curves and design variable change trend graphs for easy tracing of the optimization logic). Simultaneously, the module displays the verification results through visual charts and supports exporting the complete verification report as a PDF, providing comprehensive data support for subsequent engineering applications.

[0035] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A system for optimizing a scaled-down equivalent model of the T-tail fin layout structure dynamics, characterized in that, It includes modules for input and parameter definition, automated finite element model construction, optimization algorithm execution, and result output and visualization. The input and parameter definition module is used to receive the target parameters of the T-tail fin mode, the scaling rule parameters and the basic structural parameters, and generate the initial design constraint set. The finite element model automated construction module, based on the initial design constraint set, calls the preset structural template library to automatically generate a three-dimensional digital model and the corresponding refined finite element model, and completes mesh generation and boundary condition application; The optimization algorithm execution module adopts a hybrid optimization logic that combines genetic algorithm and least squares method to perform global search and local fine-tuning of the design variables of the finite element model, iterating until the preset convergence condition is met. The results output and visualization module verifies the dynamic characteristics of the optimized model, outputs design documents, verification reports and optimization process records, and displays the verification results through visualization charts.

2. The system for optimizing the scaled-down equivalent model of the T-tail wing layout structure according to claim 1, characterized in that, The templates in the preset structural template library adopt a combined structural form of beam frame, aviation shelf shaped frame and fiberglass skin. It includes standardized components such as beam frame, shaped wooden frame, balsa wood front and rear edges, foam top cover and fiberglass skin. The material properties and functional correspondence of each component are preset as follows: the beam frame bears the structural rigidity, and the aviation shelf shaped frame and fiberglass skin realize the shape.

3. The system for optimizing the scaled-down equivalent model of the T-tail wing layout structure according to claim 1, characterized in that, The modal target parameters include the target frequencies and target mode shapes of the first 5 modes, and support the import of FEA model result files in ANSYS and Nastran formats; the scaling rule parameters include size scaling, density ratio and dynamic pressure ratio, and the system automatically calculates the mass ratio and frequency ratio based on the scaling rule parameters; the structural foundation parameters include the material properties of elastic modulus and density of aluminum alloy, glass fiber and aerospace laminate, and support custom component material configuration.

4. The system for optimizing the scaled-down equivalent model of the T-tail wing layout structure according to claim 1, characterized in that, The refined finite element model adopts a hybrid modeling method of 1D beam elements, 2D surface elements and 3D solid elements; the default boundary condition is a fixed constraint of 6 degrees of freedom at the front end of the rear fuselage beam, and user-defined constraint scenarios are supported.

5. The system for optimizing the scaled-down equivalent model of the T-tail wing layout structure according to claim 1, characterized in that, The specific implementation of the hybrid optimization logic is as follows: the genetic algorithm initializes and generates a population of design variables, including structural parameters such as the width, height, web thickness, and flange thickness of the channel beam, and the population range is configured according to preset upper and lower limits; A fitness function is constructed by weighting the frequency error and modal confidence (MAC) values, and elite individuals are selected through selection, crossover, and mutation operations. The least squares method is used to apply the Nilsson sensitivity analysis formula to calculate the sensitivity of elite individuals. Based on the calculation results, the design variables are fine-tuned, and dynamic sensitivity analysis is performed. The sensitivity formula of frequency to structural parameters is as follows: ； In the formula, and For the first Eigenvalues ​​and eigenvectors of order 1 It is the generalized stiffness matrix of the structure. It is the structural generalized mass matrix; For the first There are 1 design variable, where T is the matrix transpose.

6. The system for optimizing the scaled-down equivalent model of the T-tail wing layout structure according to claim 1, characterized in that, The preset convergence condition is that the change in optimal fitness over 10 consecutive generations is ≤1e. -6 The default criteria for verifying the dynamic characteristics are: frequency error of the first 5 modes ≤ 2%, modal confidence MAC value ≥ 95%. If the default criteria are not met, the system supports readjusting the constraint parameters and restarting the optimization.

Images