Research On Matrix Performance Test And Simulation Analysis Of Corrugated Composite Material Skin

The research status of variant wing and its skin structure both at home and abroad was first introducedin this paper. A type of corrugated skin structure in composite material was proposed for meetingrequirements of continuous smooth and large deformation. Necessary calculation expressions wereestablished, such as density of glass fiber composites, volume fraction of each phase, elastic modulus andso on. According to calculated results, finite element simulation analysis of corrugated skin stretching andthree point bending test was made by the use of MSC. Patran/Nastran. Skin testing samples of flat andcorrugated type were prepared for tensile tests and three point bending test. The test results and simulationresults were compared and analyzed. The error between the simulation and the test was less than10%,which proved the exactness of simulation model and large deformation ability of corrugated skin.And then, on the basis of simulation, a strain sensor testing system with four channels was designedand calibrated considering the special configuration design of the skin. The strain of corrugated skin wastested at each wave crest and trough with the tensile load. Compared with simulation results, the maximumerror of was33%.Finally, a SMARC corrugated skin structure with active deformation was proposed. Two types of testsamples in this structure were prepared and tested in tensile and driving experiments with differentenvironment temperature. Results of tensile test show that the tensile deformation performance of SMARCskin samples is better than the ordinary type corrugated skin samples. But SMARC corrugated skinsamples are easily affected by temperature. When the temperature is higher, the stiffness of skin samples issmaller. In the driving test, if the drive displacement is no more than2mm, the skin samples have lineardeformation with good recovery property and repeat performance. When the displacement is more than2mm, the deformation of displacement is nonlinear due to the influence of temperature on the matrix. ARBF neural network model was established based on the data of driving test, where the current strengthparameters were regarded as the input and output parameters were the driver displacement. Theapproximation curve in the RBF neural network model is more accurate and close to the practical drivingcharacteristics. In addition, the biggest relative prediction error is less than6%which provides for activecontrol of intelligent skin with a certain theoretical and experimental basis.