Neural-Network Vibration Control of Rings With Light-Activated Shape Memory Polymer Actuators

Abstract

The light-activated shape memory polymer (LaSMP) is sensitive to ultraviolet light with specific wavelength. It is featured with dynamic stiffness. In this study, LaSMP is used to control the vibration of thin ring shells induced by external loading. Firstly the variation of LaSMP’s Young’s modulus is modeled. The mathematical model reflects the influence of light intensity, the decay coefficient and thickness of LaSMP. Besides, the model is suitable for LaSMPs with different reaction orders. Then, with Lamé parameters and the radii of rings the governing equations of the flexible ring laminated with LaSMP actuators are established. Love operators of LaSMP actuators are derived. Based on the mode expansion method, the modal forces of external loading and LaSMP actuators are given. The modal participation factors are analyzed with the modal forces. As the variation of Young’s modulus to the light intensity is nonlinear, the control effect of LaSMP actuators to common harmonic excitation is not perfect. Because the neural network control is effective to identify complex models, it is introduced to adjust the profile of light intensity. In the case study, the model of LaSMP’s Young’s modulus is validated with the experimental data. Then the forced harmonic responses of the ring are studied. For the mode n=2, the modal participation factor is reduced by 47.7% with the control of LaSMP actuator. To further enhance the control effect, the phase shift method is applied. With π/6 phase shift, the modal participation factor is reduced by 80.8%. With the neural network control method, the modal participation factor is cut down by 98.1%. The study shows that LaSMP actuator provides a new choice to control the forced vibration of flexible rings. It is also possible to apply LaSMP actuator to vibration control of other thin shell structures.