The Research On Structural Shape Reconstruction Using Modal Superposition Method

Large flexible structure is widely applied in aerospace field, such as solar panels, large aircraft wings and satellite planar antenna, etc. These aerospace components with the characteristics of small rigidity, light quality and large area need holding a certain precision in the long working. But they are vulnerable to deformation and low frequency vibration by slightly disturbed in outer space, which has caused great potential safety hazard to the spacecraft. Therefore, the research of flexible structure shape reconstruction has far-reaching significance for aerospace components health monitoring.For structural shape reconstruction, the generally method is the measurement of physical information such as strain, acceleration using attachable sensor. Then using a reconstruction algorithm to establish transformation relationship from measured physical information to global displacements, and it is possible to reconstruct the structural shape using the transformation relationship. In this thesis, the plate flexible structure model is studied based on spacecraft health monitoring. The structural shape reconstruction is researched by combining the sensing network based on fiber Bragg grating sensor, displacement estimation based on mode shape superposition and computer graphic processing technology.Firstly, the mechanical properties of flexible structure based on finite element method are researched. In the ANSYS finite element analysis software, the finite element model of target structure is created, and strain mode shapes and displacement mode shapes are obtained through modal analysis. Through statics analysis and harmonious response analysis, it is helpful to understand the structural deformation characteristics and structural deformation value such as stain and displacement are obtained. The data is the foundation of the contrastive analysis with numerical simulation results for this structural shape reconstruction method.Secondly, the theoretical derivation and numerical simulation for structural shape reconstruction is performed. Modal coordinates are firstly gained by using a small amount of measured strain and strain modal superposition. Then the global displacements are reconstructed through the linear combination of displacement mode shapes using the modal coordinates. Therefore, a transformation relationship from measured discrete strain data to global displacements is built. The numerical simulation for structural shape reconstruction is performed through C#.net combining with the interface technology of ANSYS. The simulation results show this structural shape reconstruction method has good reconstruction effects both for simple deformation and complex deformation.Meanwhile, the research on sensor layout optimization in sensing network is performed and some important parameters are discussed. In this thesis, the error of structural shape reconstruction is selected as optimal criteria for sensor layout. Optimization analysis for sensor position is performed using simulated annealing optimization algorithm, and the effect of structural shape reconstruction is significantly improved by the optimal sensor layouts. Then the effect of number of sensor and sensor direction is analyzed respectively. Finally, the multi-objective optimization analysis for sensor position and direction is performed, and Multi-objective optimization results show that the reconstruction effect is more significant. In addition, the related parameters are researched of excitation frequency, excitation amplitude, number of mode shapes and condition number.Finally, experimental platform is constructed and experimental verification and analysis for shape reconstruction is performed. In the experiment, the discrete FBG sensors pasted on flexible structure surface are used to measure strain data. Then the displacements are reconstructed using the shape reconstruction algorithm based on the modal superposition method and the real-time display for structural shape is realized through the visualization platform. Experiment analysis includes static deformation and dynamic deformation. In order to evaluate the experiment result, the displacements at particular point measured by laser displacement sensor are used as reference displacements. The experiment results show that the reconstruction displacements have a good agreement with the reference displacements both in static and dynamic experiments. Therefore, the effect of shape reconstruction based on modal superposition method is verified.