Research On Optimization Method For The Major Box Of Wing Flutter Model

During the flight, aircraft flutter is always appearing, which can lead to the destruction To avoid the happening of the flutter, study should be done for the flutter characteristics of the plane. The wind tunnel test is a reliable method to verify vibration characteristics of the plane which is on the moving load. A lot of research examples show that the precision of test data is applied to the similarity of the model with the original structure. So how to design flutter model, including shape elastic, mass similarity and stiffness similarity, is a problem worth studying.Research shows that topological optimization, size optimization and shape optimization is a effective method to solve the flutter model design. Because of the internal complexity of the wing structure, it is easy to make subtle structure for the design of flutter model. The subtle structure is bad for the manufacturing of flutter model. And because of bad elastic characteristics, the subtle structure can cause nonlinear problems. The mayor box is the main load-carrying structure of wing flutter model. The mass and stiffness similarity of the box is the base of resolving the similarity of wing flutter model. Combined with the topology optimization and size optimization, based on the shape similarity, a optimization method of box to achieve structural and elastic similarity is proposed. Work is as follows:First, with APDL,the finite element model of the box framework is built. Then using flexibility as target and exits state of beams or ribs as design variables, optimization model of framework is established. Based on the analysis of sensitivity for the beam or rib with the objective, by using MATLAB software, modify program of the structure is made to conduct the structural optimization.Second, when topology configuration of the framework is fixed on, the characteristic parameters of the beam are regarded as design variables and the structural optimization of the framework is conducted with different segmentations. The parameters of framework is modified in multi-island genetic algorithm method and sequential quadratic programming method. After optimization, the arithmetical compliment of the parameter is removed and the thickness of framework is adjusted at integer multiples of each layer thickness.Third, the skin structure of the box is established on the framework structure, thus the optimization is developed as follows:the thickness of the skin is used as a design variable and the flexibility of the box is used as a design target. After optimization, the arithmetical compliment of the parameter is removed and the thickness of the skin is adjusted at integer multiples of each layer thickness.At last, the counterweight is added on the corresponding place of the box for the similarity with the mass distribution of the real structure. The optimization is developed with frequency as a design target and counterweight as a design variable.This optimization method is applied to a box model. Results show that the proposed approach is feasible to achieve given natural frequency and natural vibration. In other words, both of the mass and stiffness similarity with the real structure are ensured. Meanwhile the method improve the efficiency of the structure, so it provides a useful reference for the design of wing flutter models.